Peripheral nerve injury-induced plasticity of spinal

voltage-dependent calcium channels:

an electrophysiological study of dorsal

horn neurones in the rat.

By

Elizabeth A. Matthews

A thesis submitted to the University of London for the degree of

Doctor of Philosophy.

Department o f Pharmacology

University College London

Gower Street

London

WCIE 6BT

ProQuest Number: U643290

All rights reserved

INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted.

In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed,

a note will indicate the deletion.

uest.

ProQuest U643290

Published by ProQuest LLC(2015). Copyright of the Dissertation is held by the Author.

All rights reserved.This work is protected against unauthorized copying under Title 17, United States Code.

Microform Edition © ProQuest LLC.

ProQuest LLC 789 East Eisenhower Parkway

P.O. Box 1346 Ann Arbor, Ml 48106-1346

ABSTRACT

Neuropathic pain can involve exaggerated pain responses often

accompanied by sensory deficits with some patients responding poorly to traditional

analgesics. Dysfunctional mechanisms, and thus potential drug targets, are

conceivably located on peripheral nerves and central neurones. Of interest here are

alterations in spinal neuronal excitability contributing to the plasticity of

transmission and modulating systems, manifest as altered nociception. Counterparts

of clinical symptoms can be studied in animal models. Spinal nerve ligation,

employed here, involves unilateral tight ligation of L5/6 spinal nerves o f the sciatic

nerve and reproducibly induces mechanical and cold allodynia. In vivo

electrophysiology was subsequently used to record evoked dorsal horn neuronal

responses to electrical and natural stimuli. Activation of voltage-dependent calcium

channels (VDCCs) is critical for neurotransmitter release and neuronal excitability,

and blockers are antinociceptive in behavioural and clinical studies. Here, effects of

N-, P/Q-, T- and L-type VDCC blockers (m-conotoxin G VIA, co-agatoxin I VA,

ethosuximide and nifedipine, respectively) on evoked neuronal responses were

investigated in an attempt to unravel their physiological and pathophysiological roles

in sensory transmission. Spinal co-conotoxin G VIA produced prolonged inhibitions

of the evoked neuronal responses in neuropathic and control rats, in a manner

increased after neuropathy. Spinal co-agatoxin I VA exerted inhibitory actions but to a

lesser extent. Minimal inhibitions were achieved with spinal ethosuximide and

nifedipine. Mechanisms of neuropathic pain are diverse; indicating that combination

therapy could be beneficial. The anticonvulsant gabapentin, effective in the clinical

treatment of neuropathic pain, has been demonstrated to bind the VDCC «26 subunit.

Here, gabapentin inhibited neuronal responses in neuropathic, but not normal, rats, in

line with neuropathy-induced VDCC plasticity. In the presence of gabapentin,

morphine exhibited greater inhibitions, again more pronounced after neuropathy.

These studies indicate a prominent role for N-type VDCCs in sensory transmission,

which is increased after neuropathy, suggestive of a key role in the increased central

excitability.

ACKNOWLEDGEMENTS

My thesis acknowledgements I find somewhat reminiscent o f an Oscar

acceptance speech - it may he somewhat long-winded and has moved me to tears, but

I will only regret not mentioning all that I want, so excuse the ramblings. And

besides, the Dickenson lab is famed for its glitz and glamour (that’s not just the

décor), and to have been part of that I am so very lucky. So thank you Tony (or is it

Austin Powers or Sporty Spice?) for giving me this opportunity. From you I have

learned so much, and that’s not just the science bit. Throughout my thesis not once

have you doubted my abilities, even when I times I did. Moreover, the time, support

and guidance you have donated to my cause has taught me to have a little more faith

in myself. In addition you’ve also come in very handy for champagne receptions and

lavish meals, and never have I met a man with a bigger shoe collection than myself

Thank you also to the other players in the Dickenson cast. That includes

old-faithfuls, has-beens and the newer replacements (I mean that in the nicest

possible way!). Louise, Mark, Wahida, Rie, Kate, Sarah, Katie, Lucinda, Lars, Vesa

and Idil. I am grateful to you all for science, neurones, tequila and dancing...you

really are a talented lot.

My parents have a lot to answer for. You have always made sure that I

know you’ll be proud of me no matter what I accomplish. Mum, thank you for being

a fantastic role model, I have learned so much from your strength and independence.

You’re the first person I fum to in good times and bad and I hope all the worry is

worth it at times like this! Dressmaking, hairdressing and now rats feet, thanks for

being so nifty with a pair o f scissors. I’ll make a scientist out o f you yet! Dad, there

aren’t many fifty-somethings who decide its time to get a degree. You have an

enthusiasm for learning and with no understanding o f in vivo electrophysiology I

know you’ll read this thesis anyway.

I’d also like to thank ‘KKK’ - Kate, Kath and Kylie. I’ve only known Kylie

to sport a white hood in public, and that’s where the similarity ends. Kate, what can I

say? Inspirational! A truly supportive, loyal and fun friend. At times I feel you know

me better than a do myself. Kath, you might be my 111’ sis, but you’ve become very

'wise in your old age, so much so the roles have been reversed for my benefit at

times. As for Kylie, happy music serves as good medicine and welcomed distraction.

Living in London wouldn’t have been as fun as it has been without the rest

of my ‘adopted’ family. Wahida, if anyone has taught me how to pull yourself

together and get on with things, it’s you .. .and just how do you manage to make a

Woolworths' carrier bag look so stylish? Vic and Bex, my longest standing friends.

From undergraduate days o f cheap wine and podium dancing to successful

professionals with slightly more expensive taste in wine and even better podium

dancing, you’ve been fantastic mates in every sense. That also goes for Julia, UCL

wouldn't have been the same without you. Beth and your fantastic ‘isms’, you’ve

always made me laugh. As have you Tom...trust me, the moonwalking comes with

practice. Finally to Rich, Matt and Rog, the family would be rather too oestrogen-

dominated if you guys weren’t around.

CONTENTS

Abstract

Acknowledgements__________________________________________________ 3

Contents ______________________________________________________________ 5

Chapter 1 Introduction________________________________________________9

1.1 Neuropathic Pain in the Clinic___________________________________________ 11

1.1.1 Incidence and Causes of Neuropathic Pain____________________________________ 11

1.1.2 Symptoms of Neuropathic Pain_____________________________________________ 11

1.1.4 Pharmacological Based Therapeutic Approaches to Neuropathic P ain_____________ 13

1.2 Animal Models of Neuropathic Pain______________________________________ 15

1.2.1 Models of Peripheral Mononeuropathy_______________________________________ 16

1.2.1.1 Chronic Constriction Injury____________________________________________ 18

1.2.1.2 Partial Sciatic Nerve Ligation __________________________________________19

1.2.1.3 Selective Spinal Nerve Ligation_________________________________________ 19

1.2.2 Models of Peripheral Polyneuropathy and Central Neuropathic Pain_________________20

1.2.2.1 Diabetic Neuropathy_________________________________________________ 20

1.2.2.2 Central Neuropathic Pain______________________________________________ 20

1.3 Mechanisms of Neuropathic Pain_________________________________________ 21

1.3.1 Peripheral Processes_____________________________________________________ 22

1.3.1.1 The Primary Afferent Fibres___________________________________________ 22

1.3.1.2 Neurochemical Changes of Primary Afferent Fibres_________________________23

1.3.1.3 Anatomical Changes of Peripheral Nerves_________________________________ 27

1.3.1.4 Primary Afferent Sprouting____________________________________________ 28

1.3.1.5 Ectopic Activity_____________________________________________________ 29

1.3.1.6 Ion channel Plasticity_________________________________________________ 30

1.3.1.7 Pathologic Intemeuronal Communication_________________________________ 32

1.3.1.8 Sympathetically-Activated Sensory System________________________________ 33

1.3.2 Central Processes________________________________________________________ 35

1.3.2.1 Dorsal Horn Neuronal Response Characteristics____________________________35

1.3.2.2 Significance of Spinal Cord Laminar Organization__________________________36

1.3.2.3 Abnormal Dorsal Horn Reorganization___________________________________ 39

1.3.2.4 Excitatory Neuropeptide and Amino Acid Receptor Systems__________________42

1.3.2.5 Inhibitory Receptor Systems___________________________________________ 45

1.4 Voltage-Dependent Calcium Channels____________________________________53

1.5 Aims of this Study______________________________________________________62

Chapter 2 Methods__________________________________________________ 63

2.1 Spinal Nerve Ligation Model of Neuropathy________________________________64

2.1.1 Surgery ___________________________________________________________ 64

2.1.2 Behavioural Testing______________________________________________________ 64

2.1.2.1 Sensitivity to Punctate Mechanical Stimuli________________________________ 65

2.1.2.2 Sensitivity to cooling stimuli___________________________________________ 65

2.2 Spinal Cord Electrophysiology___________________________________________ 65

2.2.1 Surgery________________________________________________________________65

2.2.2 Electrophysiological Recordings____________________________________________ 66

2.2.2.1 Neurone Isolation___________________________________________________ 66

2 2 .2.2 Neurone Characterization _________________________________________ 68

2.3 Pharmacological Studies_________________________________________________69

2.3.1 Drug administration______________________________________________________ 69

2.3.2 Drugs Employed_________________________________________________________70

2.4 Statistical Analysis______________________________________________________ 71

Chapter 3 Behavioural Consequences and Dorsal Horn Neurone Response

Characteristics Following Spinal Nerve Ligation__________________________ 72

3.1 Introduction___________________________________________________________ 73

3.2 Methods______________________________________________________________ 743.2.1 Behavioural Assessment__________________________________________________ 74

3.2.2 Determination of Dorsal Horn Neurone Response to Electrical Stimuli_______________ 75

3.2.3 Determination of Dorsal Horn Neurone Responses to Natural Stimuli_______________ 75

3.2.4 Determination of Dorsal Horn Neurone Receptive Field Size______________________ 75

3.3 Results________________________________________________________________763.2.1 Behavioural Assessment__________________________________________________ 76

3.2.2 Response Characteristics of Dorsal Horn Neurones to Electrical Stimuli _____________ 78

3.2.3 Response Characteristics of Dorsal Horn Neurones to Natural Stimuli_______________ 80

3.2.4 Receptive Field Characteristics of Dorsal Horn Neurones______ - _82

3.4 Discussion_____________________________________________________________ 85

3.4.1 Behavioural Consequences of SNL__________________________________________ 85

3.4.2 Dorsal Horn Neuronal Consequences of SNL___________________________________88

Chapter 4 Voltage-Dependent Calcium Channels: The Ca^2 Family__________ 96

4.1 Introduction___________________________________________________________ 97

4.2 Methods_______________________________________________________________99

4.3 Results 99

4.3.1 Behavioural Assessment___________________________________________________99

4.3.2 Spinal Cord Electrophysiology: Neurone Characterization_______________________ 100

4.3.3 Spinal Cord Electrophysiology: Effects of co-conotoxin G VIA____________________ 100

4.3.3 Spinal Cord Electrophysiology: Effects of co-agatoxin IVA_______________________ 107

4.4 Discussion___________________________________________________________ 112

4.4.1 Role of Spinal N-type VDCCs (Cay2.2, a ie )__________________________________ 113

4.4.2 Role of Spinal P-type VDCCs (Cay2.1, Œia) __________________________________ 119

4.4.3 Supporting Evidence and Implications________________________________________121

Chapter 5 Voltage-Dependent Calcium Channels: The Ca^ Family_________ 126

5.1 Introduction__________________________________________________________ 127

5.2 Methods_____________________________________________________________ 129

5.3 Results______________________________________________________________ 1295.3.1 Behavioural Assessment__________________________________________________129

5.3.2 Spinal Cord Electrophysiology: Neurone Characterization_______________________ 130

5.3.3 Spinal Cord Electrophysiology: Effects of ethosuximide_________________________ 130

5.4 Discussion____________________________________________________________ 135

Chapter 6 Voltage-Dependent Calcium Channels: The Ca^l Family_________143

6.1 Introduction__________________________________________________________ 144

6.2 Methods______________________________________________________________145

6.3 Results_______________________________________________________________1456.3.1 Behavioural Assessment_________________________________________________ 145

6.3.2 Spinal Cord Electrophysiology: Neurone Characterization________________________146

6.3.3 Spinal Cord Electrophysiology: Effects of nifedipine____________________________146

6.4 Discussion_____________________________________________________________151

Chapter 7 Gabapentin and Morphine: A Combined Approach to Inhibition of Ca

Influx____________________________________________________________ 158

7.1 Introduction____________________________________ 159

7.2 Methods______________________________________________________________ 161

7.3 Results________________________________________ 161

7.3.1 Behavioural Assessment__________________________________________________161

7.3.2 Spinal Cord Electrophysiology: Neurone Characterization_______________________ 162

7.3.3 Spinal Cord Electrophysiology: Effects of Morphine Alone______________________ 162

7.3.4 Spinal Cord Electrophysiology: Effects of Gabapentin Alone______________________163

7.3.5 Spinal Cord Electrophysiology: Effect of Morphine in the Presence of Gabapentin 168

7.4 Discussion___________________________________________________________ 170

7.4.1 Effect of Morphine Alone________________________________________________ 170

7.4.2 Effect of Gabapentin Alone_______________________________________________ 176

7.4.3 Effect of Gabapentin and Morphine Co-administration: Clinical Implications_________ 183

Chapter 8 Final Discussion__________________________________________189

References________________________________________________________ 195

Publications 242

CHAPTER 1

INTRODUCTION

Pain is an unpleasant sensory and emotional experience associated with

actual or potential tissue damage. It is subject to various influencing factors,

including the physical, psychological and social experiences o f the patient, and is

more complicated than a simple fixed stimulus-response relationship. On

encountering a potentially harmful or ‘noxious’ stimulus the final perception o f pain

is only established once the higher cortico-limbic brain centres receive the

information. This pathway involves a substantial degree o f encoding and modulation

that can occur in the peripheral nervous system, the dorsal horn of the spinal cord

and supraspinal relay and modulating centres, including the thalamus, midbrain and

brain stem.

Under normal physiological conditions pain is useful and serves as a

warning to protect the organism. In such circumstances pain is termed ‘acute’, in that

the undesirable sensations created by injury persist for only days or, in some cases,

weeks. The pain transmission pathways are, however, subject to dysfunction such

that pain can become exaggerated or ‘chronic’, persisting after resolution of the

injuring modality, and outweighing its biological purpose. Chronic pain is often

associated with expanded receptive fields; thus the amplitude of response to a given

stimulus can increase (hyperalgesia) and pain can be elicited by normally innocuous

stimuli (allodynia). Pain can also spread and become non-segmental. Such

maladaptive pain states can arise from damage to the nervous system and when

resulting from nerve injury are termed ‘neuropathic’.

Acute pain can be successfully controlled by a variety of pharmacological

agents, yet chronic pain is a major clinical concern and often responds poorly to

traditional therapeutic approaches such as non-steroidal anti-inflammatory drugs

(NSAIDs) and opioids. The development o f effective analgesics is made possible by

extensive research in the clinical and pre-clinical setting. Exploitation of continually

developing pharmacological, immunohistochemical, molecular and genomic

techniques, furthers knowledge o f the molecular and cellular mechanisms that

underlie the pain systems, along with its pathophysiological changes. Much o f the

understanding surrounding pain transmission comes from the extensive study into

cutaneous sensory input, and although non-cutaneous pain is also very important it is

10

not the focus of this thesis.

1.1 N e u r o p a t h ic P a in in t h e C l in ic

1.1.1 I n c id e n c e a n d C a u se s o f N e u r o p a t h ic P a in

Neuropathic pain is a generic term which refers to a group o f painful

disorders characterized by pain due to dysfunction or disease of the peripheral and/or

central nervous system (CNS). It has been estimated that in the UK 1% of the

population experience some form of neuropathic pain, which rises to probably 50%

in the over 65 age bracket (Bowsher, 1991). Persistent pain syndromes offer no

biological advantage and cause suffering and distress and since relatively few

neuropathic pain patients adequately benefit from pharmacological interventions,

they are therefore worthy o f investigation. Disorders that frequently result in

neuropathic pain are from diverse backgrounds (Ralston, 1998). Peripherally they

range from metabolic disorders, such as diabetes mellitus (diabetic neuropathy),

infections, such as herpes zoster (post-herpetic neuralgia) and human

immunodeficiency virus, to nerve compression due to entrapment or invading

tumours, as well as neuroma formation arising from amputation or nerve transection.

Centrally mediated neuropathic pain can result from disorders of the spinal cord such

as multiple sclerosis and trauma or tumours o f the spinal cord, brainstem and

hemispheric injuries.

1.1.2 S y m p t o m s o f N e u r o p a t h ic P a in

Irrespective o f the heterogeneous aetiologies o f neuropathic pain the clinical

picture exhibits common characteristic symptoms. Paradoxically the coexistence of

sensory deficits and increased sensations is often manifest. As might be expected

from peripheral nerve damage, impaired axonal conduction properties can arise

resulting in areas showing partial or complete loss o f sensation. This so-called

‘negative’ symptom, may or may not correspond to the somatic area of pain showing

‘positive’ phenomena, commonly localized to the territory innervated by the

damaged nerve.

11

One of the most commonly used terms to describe the positive symptoms

experienced is allodynia. This applies to the perception of pain evoked by a normally

non-noxious stimulus. Even very gentle somatosensory stimuli, such as bending of

hairs due the wearing of clothes or a breeze, can be painful. The cutaneous region of

allodynia usually shows no visible signs of injury. Allodynia can be further defined

in terms o f the evoking stimulus, such that ‘dynamic’ (or tactile) allodynia arises

from a moving stimulus in contrast to the ‘static’, ‘pressure’ and ‘punctate’,

allodynias, produced by large or fine-tipped probes, respectively. The term

hyperalgesia is often confused or misinterpreted as allodynia, but this describes an

enhanced pain perception, such that there is an increased pain response to, or

lowered threshold to, a normally noxious stimulus. Allodynia describes a change in

quality o f sensation, whereas hyperalgesia is indicative of a quantitative change, and

both of these represent stimulus-evoked pain. In contrast, pain perception emanating

from the innervation area of the injured nerve can occur in the absence of an external

stimulus and is termed ‘spontaneous’. Spontaneous pain can be continuous (ongoing,

non-relenting), or paroxysmal (episodic) and is often described as having ‘tingling,

crawling, cramping, burning or stabbing’ components.

Abnormal sensations are described as paraesthesias, such that the quality of

sensation is different to that experienced in the absence o f a neuropathic syndrome,

and these can be spontaneous or stimulus-evoked. In particular, dysesthesia relates to

unpleasant yet non-painful sensations that are qualitatively unrelated to the stimulus,

whereas hyperesthesia describes non-painful heightened sensations. Other features of

neuropathic pain include after-sensation and summation, whereby repetitive

stimulation with a mildly noxious stimulus, such as pin prick, evokes a delayed pain

(hyperpathia) and a progressive worsening of pain, respectively.

In an individual neuropathic pain patient not all o f the various types o f pain

exist, rather a combination o f these phenomena will be present in many conditions.

For example, in a study o f 63 patients with post-herpetic neuralgia, 49 showed some

form of mechanical allodynia or hyperalgesia, but only 16 exhibited dynamic,

pressure and punctate allodynia, the other 33 patients showed only one or two of the

forms in various combinations. Furthermore, 18 of the 63 patients reported thermal

allodynia, 13 reported cold allodynia and only 6 displayed both (Pappagallo et al.,

12

2000). The variety o f aberrant sensations observed in neuropathic pain indicates that

there are either multiple pathophysiological processes or common processes

affecting various components o f the peripheral and central nervous system that

nonetheless cause common symptoms o f allodynia and hyperalgesia, possibly

explaining its complicated clinical management.

1.1.4 P h a r m a c o l o g ic a l B a s e d T h e r a p e u t ic A p p r o a c h e s t o N e u r o p a t h ic

P a in

Traditionally, the treatment o f neuropathic pain is guided by a

disease/trauma based criteria dependent of the type o f nerve damage. It is now

becoming apparent that a symptom based approach may be more appropriate, as

underlying dysfunctional mechanisms often differ in their molecular, cellular and

anatomical basis (discussed in section 1.3). Since hyperexcitability, peripherally or

centrally, is an essential part o f many types of neuropathic pain, treatments have

focused on drugs that might reduce it. However, again, neuronal hyperexcitability is

not simply generated by one mechanism alone, but rather it is a culmination of

several different alterations to various parts of the sensory pathway.

Interpretation o f clinical evidence in order to determine effective analgesic

drugs for use in defined neuropathic pain states has proven difficult. However, over

the last decade numerous controlled trials have highlighted possibilities for new

therapeutic strategies as well as helping to uncover possible links between different

mechanisms and specific pain symptoms and pathologies (Woolf et al., 1998). One

of the most useful methods employed to assess analgesic efficacy is ‘NNT’ (numbers

needed to treat), which refers to the number o f patients that have to be treated before

one patient experiences more than 50% pain relief. One limitation to this method

concerns drugs with side-effect profiles that prohibit optimal dosing ’with respect to

efficacy. Side-effects are a major concern and thus although NNT may not truly

reflect efficacy they do bear relevance on the usefulness o f a drug.

Traditional analgesics, such as opiates and NSAIDs, are rarely used for the

treatment o f neuropathic pain, since they have been shown to exert limited

13

effectiveness, although there is considerable debate over the former. The current

mainstay treatments are therapeutic agents utilized for other neurological disorders,

namely depression and epilepsy, which have been observed to exhibit analgesic

properties in chronic pain states. Tricyclic antidepressants have been demonstrated in

controlled clinical studies effective in various neuropathic pains, with clinical

profiles different to those observed for depression (Kingery, 1997). These include

post-herpetic neuralgia and diabetic neuropathy with overall NNTs o f 2.3 and 3,

respectively (McQuay et al., 1996). Tricyclic antidepressants, some o f which act by

enhancing ai-adrenergic transmission by inhibition o f noradrenaline and serotonin

re-uptake, and newer antidepressants, which are serotonin-selective, show no

advantage over other treatments (McQuay et ah, 1996), and some are purported to

possess NMDA (N-methyl-D-aspartate) receptor blocking activity (Eisenach et al.,

1995). Side-effects, such as cardiac arrhythmias, postural hypotension and sedation

limit their use (Rowbotham et al., 1998).

Neuropathic pain and epilepsy both share neuronal hyperexcitability as a

common underlying mechanism. There are established anticonvulsants that target the

generation of neuronal hyperexcitability via non-specific block o f Na" channels and

some o f these have been proven effective in the treatment o f various forms o f

neuropathic pain. For example, carbamazepine has been shown to be effective in

diabetic neuropathy, yet its NNTs range from 2 - 9.4; thus is not always beneficial

(McQuay et al., 1995). Lamotrigine is a newer anticonvulsant displaying

subtype/conformational Na^ channel selectivity and is proving to have potential in

the treatment o f neuropathic pain (Eisenberg et al., 1998). Of great interest, is the

novel anticonvulsant gabapentin (GBP), which although originally designed as a y-

aminobutyric acid (GABA) analogue, does not interact with the G ABA system or at

the traditional Na" channel site o f other antiepileptics (Taylor et al., 1998). Its mode

of action remains somewhat elusive, with VDCCs being possible targets, nonetheless

GBP has been proven an effective analgesic in randomized, double-blind studies of

diabetic neuropathy (Backonja et al., 1998) and post-herpetic neuralgia (Rowbotham

et al., 1998) with NNTs o f 3.7 and 3.2, respectively. Advantages o f GBP include its

well-tolerated side-effect profile and its lack of adverse drug interactions, and thus it

represents a useful add-on therapy.

14

There are other clinically licensed drugs that show potential for the

treatment o f neuropathic pain. These include NMDA antagonists, such as ketamine

and memantine, and otz-adrenergic agonists, such as clonidine although their use is

hindered by adverse side-effects (Sang, 2000). The widely accepted notion of

‘opioid-resistant neuropathic pain’ has been re-addressed and neuropathic pain may

only show reduced opioid sensitivity, overcome with increasing dose, but again this

approach is not without side-effects (Portenoy et al. 1990; Jadad et a/., 1992). In

these cases utilization o f such drugs with other therapies may be o f benefit.

Additionally, there are emerging novel, but as yet unlicensed, drugs for the treatment

of neuropathic pain, such as the N-type VDCC blocker SNX-111, currently

undergoing phase III clinical trials, which shows promise (Mathur et al., 1998).

1.2 A n im a l M o d e l s o f N e u r o p a t h ic P a in

The development o f several animal models o f neuropathic pain has been

critical in elucidating its complex causal mechanisms. Pre-clinical models can be

used to provide an unanaesthetized, intact system, with preserved neuronal pathways

and spinal/supraspinal interactions, which realistically relates to the integrative

process o f nociception. In addition to aiding the understanding o f dysfunctional

mechanisms, behaviourally, animal models serve to define o f the pharmacology of

neuropathic pain states and can be used to assess the therapeutic benefit o f existing

drugs as well as the analgesic potential of novel drugs. Likewise, at a neuronal level,

in vivo electrophysiological techniques can be used to investigate pathological

changes involved in neuropathic pain, which is also useful for pharmacological

assessment and identification of possible new drug targets.

These approaches do provide useful predictive information regarding the

treatment o f neuropathic pain, especially in the case of novel therapies before

transfer into clinical setting. However, animal studies cannot provide descriptive

information about the quality o f pain, which may bear some relevance to underlying

mechanisms, and thus the therapeutic approach applied clinically. Furthermore,

differences in receptor types and distributions between animals used and humans

may also complicate the extrapolation of pre-clinically obtained data to the human

15

situation. The various animal models also show differences between each other in the

extent of behavioural symptoms manifest and underlying nervous system changes, as

is the case in human patients. It is important to appreciate this when making

comparisons to human neuropathic pain states, as in both scenarios the type of injury

and observed symptoms or behaviours often reflect differences in causal mechanisms

which can result in differential responsiveness to pharmacological interventions.

Thus, attention should be paid not to the model per se, but perhaps to these

individual characteristics.

Animal models of neuropathic pain can broadly be divided into peripheral

mononeuropathy, peripheral polyneuropathy and central neuropathic pain and these

will now be summarized.

1.2.1 M o d e l s o f P e r ip h e r a l M o n o n e u r o p a t h y

There are three main established rat models o f peripheral mononeuropathy

that produce partial denervation of one hindpaw. These are the chronic constriction

injury model (CCI) (Bennett & Xie, 1988), the partial sciatic nerve ligation model

(PSL) (Seltzer et aL, 1990), and the selective spinal nerve ligation model (SNL)

(Kim & Chung, 1992), which are represented diagrammatically in Figure 1. In

contrast to models of complete denervation, such as spinal section and dorsal root

rhizotomy (Zeltser & Seltzer, 1994), some sensory input into the spinal cord is

maintained allowing successful behavioural testing. Each produces behavioural signs

reminiscent o f the clinical picture, ipsilateral to nerve injury. These include tactile

allodynia, indicated by hindpaw withdrawal to cold or innocuous mechanical stimuli

such as von Frey filaments, and hyperalgesia, indicated by a shortened withdrawal

latency to mechanical and thermal stimuli. Often these evoked pain behaviours are

associated with other nocifensive manners, such as licking and biting of the injured

hindpaw in the presence or absence of an external stimulus, which in the latter case is

indicative of ongoing pain.

16

spinal cord

\/

dorsal root ganglia

sciatic nerve

saphenous nerve

receptive field

Figure 1. Animal models o f neuropathic pain. SNL (spinal nerve ligation) involves tight ligation o f L5 6 spinal nerves; CCI (chronic

constriction injury) involves placing 4 loose ligatures around the common sciatic nerve; PSL (partial sciatic nerve ligation) involves tight

ligation o f 1/3 to 1/2 o f the thickness o f the common sciatic nerve. For SNL the L4, L5 and L6 spinal nerves are exposed by removal o f part

o f the L6 transverse process. (Modified from Taho et al. (1999).

1.2.1.1 Chronic Constriction Injury

CCI involves loosely tying four ligatures o f 4-0 or 5-0 chromic gut suture

unilaterally around the sciatic nerve at the level of the popliteal fossa (back of the

knee joint), which is proximal to its branching into the L4, L5 and L6 spinal nerves

(Bennett & Xie, 1988). Loose ligation is performed such that the superficial

epineurial vasculature is maintained, however this ‘looseness’ imparts a degree of

experimental variability into the model. Depending on the extent of constriction

exerted by the ligatures, varying proportions of fibre types within the sciatic nerve

innervating the hindpaw could be injured (Basbaum et al., 1991; Carlton et al.,

1991). Nonetheless, behavioural studies following surgery reveal that CCI rats show

signs o f allodynia and hyperalgesia to mechanical and thermal (both hot and cold)

stimuli. These evoked responses are apparent within 1-2 days and maintained for a

post-operative period o f up to 3 months, and are accompanied by spontaneous

nociceptive behaviours, such as contact avoidance foot lifts exerted ipsilateral to

nerve injury in the absence o f external stimulus, indicative o f on-going pain. The

ipsilateral hindlimb also displays a characteristic ‘guarding’ posture. Instead o f fully

extending toes mediating an even spread o f weight over the entire hindpaw, the toes

are held clasped together, often raised probably to avoid contact with the floor, and

weight bearing shifts predominantly to the uninjured contralateral side. Autotomy

has also been reported, but to varying extents, following CCI (Bennett & Xie, 1988;

Attal et al., 1990) and is always seen in the complete denervation models. This self-

mutilation may reflect the animal’s response to pain or unpleasant sensation (Coderre

et al., 1986) by endeavouring to rid itself o f the painful hindpaw, but this is not a

common clinical observation and not considered relevant to the human condition.

The CCI-induced behavioural abnormalities appear to have some sympathetic

component (discussed in section 1.3) since sympathectomy has been shown to reduce

thermal allodynia and to a lesser extent mechanical hyperalgesia (Perrot et a l, 1993).

However this is not consistently reported and CCI is considered to produce relatively

sympathetic-independent pain. It has been suggested the clinical correlates o f CCI

are most likely peripheral nerve injuries and complex regional pain syndrome (see

Wallace, 2001).

18

1.2.1.2 Partial Sciatic Nerve Ligation

PSL involves ipsilaterally passing a 8-0 silk suture through the sciatic nerve

and tying it such that one third to a half of the nerve thickness is tightly ligated

(Seltzer et al., 1990). In accordance with the experimental variability encountered

using loose ligatures in the CCI model, PSL also inconsistently injures varying

proportions and types of sciatic nerve afferent fibres between animals. However,

reliable behavioural characteristics are exhibited, similar to those produced by CCI,

such as long-lasting spontaneous and evoked nociceptive behaviours, including

tactile allodynia and thermal hyperalgesia which are relieved to a certain extent by

sympathectomy (Shir & Seltzer, 1990). Although not documented in the original

paper, cold allodynia has since been described (Kim et ah, 1997). ‘Guarding’ posture

is displayed, but autotomy is not reported. Again, like CCI, the clinical correlates are

suggested to be peripheral nerve injuries and complex regional pain syndrome (see

Wallace, 2001).

1.2.1.3 Selective Spinal Nerve Ligation

SNL involves unilaterally tying a tight ligature, using 3-0 silk suture, around

two (L5 and L6) o f the three spinal nerves (L4 - L6) that comprise the sciatic nerve,

inbetween their conjunction to form the sciatic nerve but proximal to their DRG

(Kim & Chung, 1992). This results in a highly reliable tactile allodynia, thermal

hyperalgesia and cold allodynia, and to a lesser more variable extent, spontaneous

pain behaviours. Rats also show ‘guarding’ positioning o f the injured hindpaw, but

no autotomy. Stable allodynia is established within 2 days o f surgery and maintained

for 2 - 3 months. In comparison to CCI and PSL, the magnitude of mechanical

allodynia, assessed by withdrawal frequency to innocuous mechanical stimuli (von

Frey filaments), is greatest and most reliable in the SNL model. Conversely CCI,

which displays the least reliable mechanical allodynia, shows more substantial

spontaneous nociceptive behaviours than observed after SNL and PSL (Kim et ah,

1997). Sympathectomy is reported to relieve neuropathic pain behaviours induced by

SNL (Kim & Chung, 1992; Choi et ah, 1994; Kim et ah, 1997), and this has been

demonstrated to be notably greater than that achieved for CCI and PSL (Kim et ah,

1997).

19

In comparison to CCI and PSL, the selective tight ligation o f SNL

minimizes experimental variability to some extent, however individual rats anatomy

may differ in the contribution that each spinal segment makes to the sciatic nerve.

Furthermore the SNL model preserves segregation of injured and uninjured primary

afferent input proximal to the formation o f the sciatic nerve such that the

corresponding spinal segments and DRG for each spinal segment contain either

injured (L5 and L6) or intact (L4) fibres. SNL represents a model o f partial

denervation but in contrast to CCI and PSL, where partial denervation damage to the

common sciatic nerve results in all three spinal segments and their DRG containing a

combination of damage and undamaged fibres, spinal input from damaged afferents

can be selectively handled. The clinical correlates o f SNL are suggested to be nerve

root injury and plexus avulsion injury but may well also relate to other syndromes

where nerve compression occurs.

1 .2 .2 M o d e l s o f P e r ip h e r a l P o l y n e u r o p a t h y a n d C e n t r a l N e u r o p a t h ic

P a in

1.2.2.1 Diabetic Neuropathy

This model was developed to study the neuropathic pain state o f peripheral

diabetic neuropathy observed in some patients with diabetes mellitus. Subcutaneous

injection o f streptozotocin in rats induces a hyperglycaemia and glucosuria that

results in a peripheral polyneuropathy, whereby animals show allodynia, not reversed

by sympathectomy (Ahlgren & Levine, 1993), and decreased nociceptive threshold

to paw pressure (Malmberg et a l , 1993) within one week.

1.2.2.2 Central Neuropathic Pain

There are two main animal models of central neuropathic pain that involve

the generation o f spinal ischaemia. This can be induced transiently by occlusion of

the descending aorta (Marsala & Yaksh, 1994), or permanently by localized Argon

laser irradiation o f the spinal cord following injection of a photosensitive dye (Hao et

ah, 1992a). Animals exhibit dynamic tactile, and cold allodynia. This model is

proposed to be representative o f spinal cord injury, such as that induced by

20

ischaemia, trauma or radiation. Additionally there is another central neuropathic pain

model not involving ischaemia but rather inhibition o f inhibitory glycine or GABAa

receptors (discussed in section 1.3) by intrathecal injection o f strychnine or

bicuculline, respectively (Yaksh, 1989; Sherman & Loomis, 1994). This only results

in prominent tactile allodynia localized to the spinal dermatomes affected by the

spinal injection (hindlimbs and lower back). This is also suggested to be a model o f

human spinal cord injury.

The next section reviews many of the mechanisms, suggested or otherwise

proven, to underlie many o f the symptoms experienced, or exhibited in the case of

animal models, in neuropathic pain. Much of the information has been obtained or

consolidated with observations made in the above mentioned pre-clinical models.

1.3 M e c h a n is m s o f N e u r o p a t h ic P a in

In the normal situation, the perception o f pain requires that information

concerning a noxious stimulus at a peripheral cutaneous site must be conveyed to the

higher centres o f the brain. The sensory pathway begins with primary afferent

neurones. The primary afferent neurone comprises a cell body (within the dorsal root

ganglia (DRG)) and a stem process that extends both to the target tissue, via a

peripheral nerve, and to the dorsal horn o f the spinal cord. It is at the level of the

dorsal horn that afferent spatial and temporal characteristics o f the stimulus are

encoded and subject to excitatory and inhibitory modulation before projection to

higher brain centres via ascending tracts. The qualitative changes in sensation allude

to long term reorganization of sensory pathways and evidence from clinical and pre-

clinical studies reveals that changes in both peripheral nerves and central neurones

may underlie the causal mechanisms and altered drug sensitivities displayed by

neuropathic pain.

I will now go on to discuss the peripheral and central processes involved in

the transmission of ‘physiological’ pain alongside probable and established

dysfunctional mechanisms encountered within this pathway that result in

‘pathophysiological’ neuropathic pain.

21

1.3.1 P e r iph e r a l P r o c esse s

1.3.1.1 The Primary Afferent Fibres

Primary somatosensory afferents can be classified into essentially Ap-, AÔ-

and C-fibre groups utilizing anatomical, histochemical, functional and physiological

criteria. In skin they are present in proportions o f about 20, 10 and 70%, respectively,

and are differentially sensitive to noxious and non-noxious stimuli. Physiologically,

the degree o f axonal Schwann cell m yelination results in differing conduction

velocities and also permits discrimination by diameter size reflected by their cell

bodies in the DRG (Table 1). Heavily myelinated and thus large diameter, rapidly

conducting AP-fibres normally only transmit modes of non-noxious, low intensity

mechanical stimuli such as touch, vibration, and pressure. Thinly myelinated,

intermediately sized and conducting A6-fibres convey both innocuous and noxious

information. Type I Ad-fibres are specifically activated by noxious high-intensity,

mechanical stimuli, whereas the less common type II respond preferably to noxious

heat. AÔ-fibres more responsive to cooling are also reported. C-fibres lack

myelination and are therefore the sm allest and slow est conducting afferents

specialized for transm ission o f noxious stimuli. Polymodal C-fibres respond to

mechanical, heat and chemical activation. Additionally C-fibres preferentially

responsive to either chemical, warming, cooling or low -threshold mechanical

pressure stimuli have also been described.

Table 1. Properties o f the sensory cutaneous primary afferent fibre types Estimated

conduction velocities (Schouenborg, 1984; Fox, 1999) and DRG cell body diameters

(Scroggs & Fox, 1992a) refer to those o f adult rats.

AP-FIBRE AÔ-FIBRE C-FIBRE

Conduction velocity (m/sec) 7 - 7 5 . 2 - 7

Myelination Heavy Light

DRG cell body diameter (pm) 45- 51 . :. 33 -38 '.,3

Stimulus conveyed Non-noxious Non-noxious/

noxious

Noxiou^', t '

22

Histochemically, C-fîbre afferent neurones can be further divided into IB4

(plant-derived isolectin) -positive and TrkA (NGF (nerve growth factor) receptor) -

positive populations. Functional differences may exist between these two nociceptor

populations relating to particular facets of pain transmission (Stucky & Lewin, 1999)

and differential sensitivities to trophic factors, critical not only during development

but also for maintenance o f afferent phenotypes in the adult nervous system. These

differences may underlie fibre alterations that occur in persistent pain states (Silos-

Santiago et al., 1995; Fundin et a l, 1997; Molliver et a l , 1997).

1.3.1.2 Neurochemical Changes of Primary Afferent Fibres

Within the sensory pathway, neurotransmission between nociceptive

primary afferent neurones and secondary neurones in the dorsal horn of the spinal

cord is dependent on neurotransmitters. Neurotransmitters are produced in the DRG-

located cell bodies o f primary afferent neurones. From here they are axonally

transported to central terminals and stored in presynaptic vesicles awaiting

depolarization-coupled release. Translocation to peripheral terminals can also occur

and this is important in neurogenic inflammation (Holzer, 1988). Neurotransmitters

comprise excitatory amino acids (EAAs), such as glutamate and aspartate, associated

with fast excitatory neurotransmission and neuropeptides, such as calcitonin gene-

related peptide (CGRP) and substance P, that can exert longer-lasting actions.

Physical separation o f EAAs and peptide transmitters exists such that EAAs are

stored in visually small, clear vesicles, whereas neuropeptides are stored in larger,

dense vesicles, and this allows differential release in response to varying axonal

activity patterns. In order to discuss alterations in the different primary afferent

neurochemical phenotypes it is first necessary to understand the normal situation.

The majority o f studies concerning neurotransmitter distribution patterns have been

conducted in the rat, and this appears to be similar in most mammals studied

(Merighi et a l , 1990; Bonfanti et a l , 1991). There is little information regarding

neurotransmitters involved in human nociception, however data derived from

primates may approximate.

23

A/3-fibres

Neurotransmitter types located in the primary afferents of AP-fibres are

unclear. Although at least 50% of DRG neurones stain positive for glutamate in

primates (rather more in rats), most glutamate-positive cells are of small diameter.

Co-localization of glutamate and aspartate has been reported in the DRG, but these

amino acids appear to distribute to different populations o f afferent fibre terminals in

the dorsal horn (Merighi et a i , 1991), therefore aspartate may be found within AP-

fibres (Sorkin & Carlton, 1997). Peptides tend not to be localized in large diameter

DRG cell bodies, although correlations between peptide immunoreactivity and

neuronal conduction velocity has revealed a small proportion o f AP-fibres are

positive for CGRP (McCarthy & Lawson, 1990). AP-fibres do not express substance

P.

AS-fibres

Aô-fibre DRG cell bodies are positive for glutamate, aspartate and a wide

range of peptides. Notably a third contain CGRP and a fifth contain substance P and

co-localization is probable (McCarthy & Lawson, 1989). The C-fibre phenotype is

much more clearly defined and may equally well describe the AÔ-fibre phenotype

originating from high threshold cutaneous mechanoreceptors.

C-flbres

C-fibre DRG cell bodies give rise to central terminals that contain both

small clear and large dense vesicles, which highlights the co-transmission of EAAs

and neuropeptides. As concerns the EAAs, terminals containing glutamate are more

abundant than those for aspartate, and these amino acids are not found together

(Merighi et ah, 1991). 50% of small diameter C-fibre DRG neurones do not contain

peptides and these correlate to the IB4-staining population. Those expressing TrkA

on cell bodies and central terminals (Averill et al., 1995) however do stain positive

for CGRP and substance P (Molliver et al., 1995). Co-localization o f CGRP with

somatostatin and/or tachykinins has been demonstrated (Merighi et ah, 1988;

Merighi et al., 1991) and see (Sorkin & Carlton, 1997; Wilcox & Seybold, 1997).

24

Alongside transmitters already mentioned are a number o f possible others. The

presence o f various peptide neurotransmitters is subject to positive and negative

modulation following peripheral nerve injury, as will now be discussed.

Neurochemical Down-Regulation

In rat and monkey DRG approximately 45% and 75%, respectively, o f

neurones are CGRP-immunoreactive (McCarthy & Lawson, 1990; Wilcox &

Alhaider, 1990) and these are primarily nociceptive primary afferents, as already

mentioned. Likewise 40% of monkey DRG neurones synthesize substance P mRNA

(see Wilcox & Seybold, 1997) and this is nociceptive fibre specific. Following nerve

injury there is a marked reduction in the expression o f both these peptides in the

DRG (Villar et a l , 1989; Nothias et al., 1993; Cameron et a l , 1997). This is

confirmed by a reduction of substance P mRNA in the DRG in models of CCI,

sciatic nerve section and nerve crush (Nielsch et al., 1987; Nothias et a l , 1993;

Marchand et al., 1994). A reduction in these peptides and their mRNA can be

explained by the transient loss of TrkA receptors on the peripheral terminals of

TrkA-positive, peptidergic primary afferents and disrupted axonal transport of NGF

to the DRG, a necessary requirement for CGRP and substance P expression.

Furthermore, the development of spontaneous activity within the DRG after nerve

injury (Kajander et al., 1992) may result in a depletion o f presynaptic terminal stores

of peptide due to increased activity-dependent vesicular release.

Neurochemical Up-Regulation

Other peptides appear to upregulated in sensory neurones following

peripheral nerve damage and the majority o f these changes probably serve to

enhance pain transmission. Interestingly, in contrast to the substance P decrease in

nociceptive fibres, substance P’s previously quiescent expression in Ap-fibres is

induced after nerve injury (Noguchi et ah, 1994; Miki et ah, 1998), although overall

the levels o f substance P mRNA are downregulated. This fibre switch to a

nociceptive mode from a previously tactile-specific function, may have relevance to

touch-evoked pain. Vasoactive intestinal polypeptide (VIP) is not normally

expressed in lumbar ganglia, in contrast to the sacral and cranial spinal nerves

25

(Honda et al., 1983; Yaksh et a l , 1988; Helke et a l , 1990), however it does appear

to be upregulated in ipsilateral DRG after sciatic nerve transection (Shehab &

Atkinson, 1986). NGF may act to prevent expression of VIP (Mulderry, 1994), the

inhibitory effect o f which would be removed on the decrease o f NGF after nerve

injury. It is important in glycogenolysis and thus its upregulation may reflect active

processes o f nerve regeneration, however it appears to mediate a central excitatory

role in pain transmission (Wiesenfeld-Hallin et a l , 1992), see also (Dickinson &

Fleetwood-Walker, 1999) for review. In intact animals, galanin is expressed in a

small number o f DRG neurones (Lawson et a l , 1993), preferentially small diameter

cells, although its nociceptive/non-nociceptive role is somewhat unclear, it generally

enhances nociception. Following peripheral nerve damage, its expression is

upregulated in small diameter neurones and induced in larger cells (Noguchi et a l ,

1993; Ma & Bisby, 1997) and now exerts inhibitory actions in the dorsal horn

(Wiesenfeld-Hallin et a l , 1992; Luo & Wiesenfeld-Hallin, 1995). Neuropeptide Y,

not normally expressed in sensory neurones, is also induced after nerve injury

(Wakisaka et a l , 1992; Verge et a l , 1993) in large diameter neurones. This is

paralleled by an increase o f corresponding Yz receptors on peripheral afferents and

central AP-fibre terminals. Centrally it may produce antinociceptive effects, yet

peripherally it may be pro-nociceptive (Munglani et a l , 1996).

Nerve injury induces a complex pattern of neurochemical and receptor

alterations in primary afferent fibres, often in a fibre type-specific manner. Some of

the changes observed are purely targeted to the development o f neuropathic pain,

however some o f those apparently associated with regeneration of damaged nerves

may also contribute to enhanced nociception. Although a general consensus prevails,

the extent and specifics of such changes maybe subject to slight variation in different

pre-clinical models. For example, levels o f substance P depletion were less profound

after nerve crush compared to section (Nielsch et a l , 1987; Nielsch & Keen, 1989).

Not all changes may be permanent, indeed the reduction of NGF and decreased TrkA

receptor expression after L5 ligation returned between 2 days and 2 months,

respectively. Furthermore, endogenous return o f the trophic mediators has been

correlated with the return o f substance P levels (also demonstrated with exogenous

NGF application), again relating to extent o f nerve damage.

26

Aside from neuropeptides, certain alterations to other neuromodulators also

exist in primary afferent fibres, such as nitric oxide (NO). NO is a neuronal

messenger found throughout the CNS and is produced on demand from L-arginine

by nitric oxide synthase (NOS). Evidence suggests that it may be linked to events

downstream o f NMDA receptor activation, underlying central sensitization

(discussed in section 1.3.2), (Meller et al., 1992; McMahon et al., 1993) as well as

exerting peripheral actions and thus maybe important in chronic pain states.

Following some types o f peripheral nerve damage, small DRG neurones show an

increase in NOS mRNA (Luo et al., 1999), which may be attributable to NGF loss,

and NOS inhibitors have shown experimental antinociceptive effects (Yoon et al.,

1998). However, changes in NOS do not parallel the behavioural allodynia (Luo et

oL, 1999).

However convincing these nerve injury-induced histochemical changes are,

the relationship o f these to the onset and nature of neuropathic pain behaviours and

the clinical relevance they might serve still remains to be conclusively proven.

1.3.1.3 Anatomical Changes o f Peripheral Nerves

Peripheral nerve damage usually results in the formation o f a ‘neuroma’,

whereby the distal end o f the remaining primary afferent seals and swells, due to

continued axonal transport. Disruption to Schwann cell distribution and density also

occurs along the damaged axon and at the neuroma producing areas o f

demyelination. Nerve injury, much in the same way as damage to non-neuronal

tissue mediates an inflammatory response, can induce infiltration o f macrophages

and immunocompetent cells to damaged nerves and DRG, and it is thought that

phagocytotic action of macrophages is responsible for the reduction in myelin. Loss

o f this electrical insulation results in altered axonal conduction properties and

abnormal excitability may ensue (see Garry & Tanelian, 1997).

As might be expected after nerve lesion there is substantial loss o f sensory

axons distal to the site of damage accompanied by sensory neuronal cell death in the

DRG. Anatomical and electrical studies report this to be largely a myelinated A-fibre

27

loss of up to 85%, with little reduction in C-fibre numbers. Since axonal transport is

critical for the transport of cellular products neuronal death is likely a result of target-

derived trophic factor deprivation (see Garry & Tanelian, 1997).

1.3.1.4 Primary Afferent Sprouting

A characteristic of the sensory nervous system is its ability to regenerate

after injury. Indeed, substantial deafferentation induced by peripheral nerve damage

can initiate the reinnervation of target tissue, involving growth cone formation and

subsequent elongation o f the injured nerve. Axonal and Schwann cell-derived

growth-enhancing substances are required for this process. These include NGF,

known to be necessary for the survival of many nerve cells, and apoliprotein E,

involved in remyelination. Regeneration is not always lucrative and its success can

depend on fibre type and extent o f injury. It appears that large, myelinated Ap-fibres

are least likely to regenerate, in comparison to C-fibres, and this is increasingly

hindered with increasing severity of nerve injury (Navarro et al., 1994; Navarro et

al., 1997). In some cases regeneration fails to be initiated and sensory loss in the

target area results (Koerber et al., 1989). In other situations inappropriate targeting of

regenerating fibres can lead to innervation of low-threshold cutaneous receptors by

nociceptive afferent neurones.

Another complication is the phenomena o f ‘collateral sprouting’, whereby

uninjured sensory afferents o f adjacent uninjured nerves, or the damaged nerve itself,

expand into the deafferented area once occupied by the lesioned nerve. This is

mediated by all fibre types, yet smaller primary afferents, in comparison to heavily

myelinated axons, appear to have the potential to expand over larger areas (Navarro

et a l , 1994). This may impact upon the development o f hyperalgesia as opposed to

the return of normal sensation after nerve injury. Hyperalgesia observed after CCI, a

model showing marked myelinated fibre degeneration, was prevented by saphenous

nerve section to prevent collateral sprouting, one week after nerve injury (Ro &

Jacobs, 1993). Models of nerve crush or transection also demonstrate collateral

sprouting which may be related to pain behaviours in the rat (Markus et al., 1984;

Brenan, 1986; Kingery & Vallin, 1989).

28

In both cases of axonal regeneration and collateral sprouting erroneous

reinnervation o f deafferented target tissue could contribute to aberrant pain

sensations and abnormal central spatial representation o f tactile information in

neuropathic pain sufferers.

1.3.1.5 Ectopic Activity

Damage to peripheral nerves results in alterations in their electrical

transduction properties and ‘ectopic’ activity arises whereby sites different to the

normal transduction elements o f peripheral terminals of the primary afferent receptor

generate impulses. This can be both spontaneous and stimulus-dependent in nature

and can arise from afferent axonal sprouts iimervating the neuroma, axonal areas of

demyelination and DRG of A- and C-fibres (Kajander & Bennett, 1992; Xie et aL,

1995; Study & Krai, 1996). Ectopic discharge emanating from the DRG is slow and

erratic (Kajander et aL, 1992), which contrasts to the high frequency rhythmic firing

from injured afferents. The contribution o f DRG-derived ectopic activity to the total

amount appears to depend on the type of nerve injury model, and ranges from very

little (Wall & Devor, 1983) to accounting for nearly all ectopic activity in the CCI

model (Kajander et aL, 1992). The injured peripheral nervous system provides

several locations at which numerous pathophysiological changes can occur

contributing to this aberrant activity.

Stimulus-dependent ectopic activity can arise in the injured or regenerating

nerve at the neuroma or at axonal sites, whereby novel or increased sensitivity to

thermal and chemical mechanical stimuli develops such that even blood vessel

pulsation can evoke pain. Gentle mechanical stimulation o f neuromas can produce

bursts o f firing that can extend beyond the stimulation period (Burchiel, 1984).

Algesic chemicals such as bradykinin, histamine, prostaglandins and leukotreines are

inflammatory mediators normally only active at peripheral nerve terminals, not along

axons, where they act as sensitizing agents. Subsequent to nerve injury axons

become responsive to exogenous application o f these substances, and this may be

important in the development o f hyperalgesia (Devor et aL, 1992). Endogenously,

peripheral nerve damage has been shown to bring about a local inflammatory

29

response (Tracey & Walker, 1995; Michaelis et aL, 1997) involving the release of

inflammatory mediators from immuncompetent cells and proliferating Schwarm

cells.

Spontaneous activity of primary afferent neurones occurs independently to a

stimulus, and in an intact peripheral nervous system this is rarely observed. After

injury, primary afferent axons develop spontaneous activity (Devor, 1991).

Differences in the characteristics o f spontaneous activity occur for its time o f onset

and proportions of fibres involved, and this is under the influence of primary afferent

fibre type, type of nerve injury and species. Onset has been observed 2 days after

nerve injury (Steel et aL, 1994), and the proportion o f nerves displaying such

spontaneous activity appears to increase over the postoperative period (Govrin-

Lippmann & Devor, 1978). Generally A-fibre mediated spontaneous activity peaks at

an earlier time point, which is then followed by C-fibres (Devor, 1991; Kajander &

Bennett, 1992), and after nerve section these were demonstrated to be 2 and 4 weeks,

respectively (Govrin-Lippmann & Devor, 1978). Furthermore, up to about 30% of

myelinated fibres tend to develop spontaneous activity, whereas only as little as 3%

of unmyelinated fibres do and 3 days after CCI this was shown to coincide with the

onset of painful behaviours (Kajander & Bennett, 1992). In the clinic, positive

correlations have been made between the occurrence o f spontaneous firing of

nociceptors innervating the painful region and neuropathic pain (Nystrom &

Hagbarth, 1981; Gracely et aL, 1992). In both human neuropathic pain states and

animal models, stimulus-independent pain is observed and spontaneous ectopic

activity seems a likely contributing candidate.

1.3.1.6 Ion ch annel Plasticity

The flow of Na" , K" and Ca ions through their respective channels in the

axonal membrane is critical for the normal functioning o f peripheral nerves

controlling excitability and the propagation of action potentials. It is logical that

altered electrical properties and excitability o f primary afferent fibres subsequent to

peripheral nerve damage is thus resultant from modifications to expression and

distribution o f such ion channel proteins. Na^ channels have attracted substantial

30

investigation and a multiple alterations have been highlighted follovdng nerve injury.

In the adult, non-pathological situation, DRG have been shown to express at

least seven types o f Na" channels encoded by different genes. Based on their

susceptibility to block by tetrodotoxin (TTX) they have been classified into TTX-

sensitive (low threshold o f activation; rapidly activating/inactivating) and TTX-

resistant (high threshold of activation; slower activation/inactivation) types. Of these,

the TTX-resistant SNS and NaN channels are specifically expressed upon

nociceptive small diameter neurones. Interestingly, after peripheral nerve damage

their gene expression is reduced (Oku et aL, 1988; Novakovic et aL, 1998; Dib-Hajj

et aL, 1999), yet translocation of SNS to the site of injury is enhanced (Novakovic et

aL, 1998). The lack o f myelin at surrounding the neuroma allows insertion o f

translocated Na" channels into the axonal membrane, and indeed increased axonal

Na^ channel density correlates with the materialization o f ectopic spontaneous

discharge, also demonstrated to occur at demyelinated areas (Burchiel, 1980).

‘Knock-down’ of SNS protein function, utilizing antisense oligodeoxynucleotides,

has been shown to prevent CCI-induced neuropathic pain behaviours in rats, yet SNS

‘knockout’ mice retain neuropathic pain symptoms, which may be due to

developmental compensatory increases (Porreca et aL, 1999). In contrast, the

expression o f a silent TTX-sensitive a-III embryonic Na" channel emerges in C-

fibres. Re-expression of a-III is thought to mediate the rapidly repriming Na" current

displayed by DRG neurones following nerve injury (Black et aL, 1999) and its

kinetics permit repetitive firing in injured neurones that likely contribute to ectopic

activity. A similar pattern o f TTX-sensitive channel increase and TTX-resistant

decrease is also mirrored in damaged A^-fibres (Rizzo et aL, 1994).

Ectopic neuronal discharge can be reduced by systemic and topical

administration o f Na^ channel blockers. Anticonvulsants, such as carbamazepine,

antiarrhythmics, such as mexilitine and local anaesthetics such as lidocaine non-

selectively block Na" channels in use-dependent manner. Studies employing these

agents show that differential Na" channels types expressed at the neuroma and within

the DRG are responsible for ectopic activity and pre-clinical and clinical data reports

a reduction in neuropathic pain behaviours or symptoms following Na^ channel

31

blockade (see Garry & Tanelian, 1997).

1.3.1.7 Pathologic Interneuronal Communication

Under normal physiological conditions single primary afferent fibres

function independently o f one another in the peripheral nervous system until

convergence occurs in the dorsal horn. This is made possible by the myelination of

A-fibres, and the presence o f Schwann cell processes that separate unmyelinated C-

fibres. However, as already discussed, nerve injury can result in a disruption of

insulating Schwann cell density and distribution. At areas o f axonal demyelination

and neuromas apposition of denuded axons can allow the ‘cross-excitation’ of an

impulse in one fibre to an adjacent fibre via ‘ephatic communication’ (see Garry &

Tanelian, 1997).

Another non-synaptic mode o f communication between neurones is

‘crossed-afterdischarge’ whereby the activity o f a group of neurones alters the

endogenous firing activity of their neighbours (Amir & Devor, 1992). This occurs in

non-pathological settings in primarily large diameter neurones with limited effects

due to minor depolarizations. Suggested mechanisms include diffusible factors, such

as ATP (adenosine triphosphate), and an increase in extracellular concentrations.

Again, due to nerve injury-induced demyelination, neurones and neuromas become

more susceptible to such substances and crossed-afterdischarge may also enhance

any ectopic activity ongoing in damaged nerves as well a recruit silent neurones

(Amir & Devor, 1997).

Both these non-synaptic modes o f intemeuronal communication have the

potential to mediate abnormal interactions between different fibre types. Low-

threshold Ap-fibres may directly activate nociceptive C-flbres, such that an

innocuous stimulus could evoke a painful sensation and could thus contribute to

mechanical allodynia.

32

1.3.1,8 Sympathetically-Activated Sensory System

In the intact peripheral nervous system, the afferents of sensory primary

afferents are functionally discrete from the efferent sympathetic system and its

innervation of the DRG usually ceases within the encompassing vasculature, where it

controls blood supply. After damage to peripheral nerve both anatomical and

physiological studies have demonstrated atypical sensory-sympathetic coupling such

that the sympathetic nervous system may be involved in ensuing sensory changes

that result from neuropathy.

Anatomically, nerve injury can induce sympathetic sprouting resulting

formation o f ‘basket’ structures that surround DRG soma, such that close contact

between sympathetic terminals are formed (Chung et aL, 1997). In particular, those

corresponding to A(3-fibres are innervated, however transient basket formation

around smaller neurones may also occur (Chung et aL, 1996; Ramer & Bisby, 1997;

Ramer et aL, 1997). This process appears to rely upon NGF and cytokines tliat are

produced by proliferating Schwann and immunocompetent cells following peripheral

nerve injury. The resultant effect o f this novel sympathetic innervation is

noradrenergic modulation of DRG activity such that spontaneous or evoked firing is

dramatically enhanced, and to a lesser extent non-active A^-fibres may become

active. Via Ap-fibres this may directly evoke allodynia, and via facilitation of a

constant C-fibre input into the dorsal horn it may contribute to central sensitization.

Physiologically, primary afferents show increased responsiveness to

exogenous and endogenous stimulation. Application of noradrenaline to the neuroma

of humans has been shown to produce burning sensations and increased electrical

discharges. Likewise in pre-clinical studies systemic administration of noradrenaline

activated and sensitized intact C-fibres o f the partially injured nerve and electrical

stimulation of the appropriate sympathetic nerve trunk caused increased activity of

the neuroma. Subtype-specific alterations in density and/or responsiveness o f 0C2-

adrenergic receptors appear to contribute to the increased sympathetic sensitivity

(Cho et aL, 1997; Birder & Perl, 1999). Following sciatic nerve section, type A

receptors are upregulated and type C downregulated in DRG corresponding to A-

fibres, paralleled by alterations in their gene expression observed after SNL.

33

Furthermore peripheral nerve injury altered the activity o f a population o f tt2

adrenoceptors such that their activation reduced currents and increased

excitability of small DRG neurones (see Abdulla & Smith, 1997).

Numerous clinical and pre-clinical studies, some utilizing surgical or

chemical intervention, have confirmed a sympathetically-maintained component of

some, but not all, types o f neuropathic pain. In human neuropathic pain patients the

occurrence o f sympathetic-attributable symptoms, such as skin vasomotor and

sweating anomalies and dystrophic changes in skin, hair, nails and bone can be

observed and their neuropathic pain is worsened by sympathetic stimulation (see

Taylor, 2001). More convincing is the ability o f guanethidine, which depletes

noradrenergic terminals and thus all sympathetic neurotransmitters, and adrenergic

antagonists to produce long-lasting relief from sympathetically-maintained pain in

the majority o f patients (Torebjork et aL, 1995). In the animal model o f SNL,

surgical sympathectomy performed after or before nerve ligation alleviated allodynia

or prevented allodynia and hyperalgesia, respectively (Kim et aL, 1993). In SNL and

other nerve injury models similar effects have been observed using guanethidine

(Shir & Seltzer, 1991; Kim et aL, 1993), however reversal o f neuropathic pain

behaviours has not been successful in L5 ligated rats. Such controversy in the

literature may well be explained by differences between pre-clinical models since

types o f peripheral nerve damage and animal strains may provide sources of

variability in the onset and time course o f sympathetic sprouting, thus requiring

appropriate timing of sympathectomy for successful results. Alternatively, there may

not be a sympathetic component to some types o f injury. Allodynia is observed in the

absence o f sympathetic sprouting (Marchand et aL, 1999) and a positive correlation

between the extent o f sprouting and sympathetic dependence was shown not to exist

(Kim et aL, 1998). Furthermore, not all neuropathic pain patients have benefited

from sympatholytic procedure (Kingery, 1997) and for those cases which do, painful

symptoms can return. The use of sympathectomy as a treatment o f neuropathic pain

is therefore debatable and not always suitable for every situation.

34

1 .3 .2 C e n t r a l P r o c e sse s

As we progress along the sensory transmission pathway from the periphery

to the brain, primary afferent fibres must now terminate and make their first synapse

with the CNS. This occurs in the dorsal horn o f the spinal cord, a site where the

peripheral input undergoes anatomical convergence and neurotransmitter system

modulation, before projection to higher brain centres via ascending tracts. The spinal

cord exhibits both excitatory and inhibitory systems mediated by a diverse array of

neurotransmitters and receptors and the balance of the two is critical in determining

the stimulus-response relationship. Nerve injury-induced changes to this part o f the

pain pathway will now be discussed.

1.3.2.1 Dorsal Horn Neuronal Response Characteristics

In addition to the altered peripheral processes mediated by primary afferent

fibres, peripheral nerve damage can also lead to long-term changes in the CNS which

contribute to the dysesthesias associated with neuropathic pain. After SNL the

receptive fields of dorsal horn neurones are enlarged (Suzuki et aL, 2000b), and

implies that a greater number of spinal neurones would be recruited for a given

stimulus. Nociceptive dorsal horn neurones also display de novo spontaneous activity

and reduced mechanical stimuli thresholds in parallel with lowered magnitudes of

responses (Chapman et ah, 1998b). Furthermore, experimental nerve injury creates a

loss o f afferent input into the spinal cord (Castro-Lopes et aL, 1990). These

observations fit well with the clinical neuropathic pain profile of both allodynia and

hyperalgesia together with sensory deficits, yet how altered peripheral and central

neuronal responses contribute to these symptoms still remains to be thoroughly

defined. What is clear is that in the presence of dramatically reduced afferent input,

dorsal horn neuronal responses manage to sustain their response magnitudes,

indicative of central neuronal compensatory mechanisms which may explain the

positive symptoms o f neuropathic pain despite loss o f normal sensory input (see

Dickenson et aL, 2001).

The increased barrage o f impulses arising from primary afferent fibres

mediated by the discussed peripheral mechanisms is important in eliciting central

35

sensitization and hyperexcitability o f dorsal horn neurones (W oolf & Mannion,

1999). Ongoing peripheral activity results in sustained neurotransmitter release and

activation o f neurotransmitter systems within the spinal cord. Plasticity in the

modulatory systems and functional connectivity together contribute to a persistent

pain state. Appreciation o f these alterations first requires some description of the

spinal cord organization and termination patterns of the primary afferent fibres.

13.2,2 Significance of Spinal Cord Laminar Organization

The spinal cord accommodates axons within its white matter and neurones

and their processes within its grey matter. The grey matter is cytoarchitecturally

subdivided into 10 horizontal laminae, and this serves to roughly group functionally

related cells (Figure 2). Laminae I-VI receive sensory input and constitute the dorsal

horn. Laminae I is the outer marginal layer. Ventral to this is laminae II - the

substantia gelatinosa, further divided into outer (IIo) and inner (Hi) layers. Laminae

III and IV are the nucleus proprius, V and VI (only defined in lumbo-sacral and

cervical enlargements) are the deep dorsal horn layers. Laminae VII-IX contain

motor neurones and compose the ventral horn, and lamina X surrounds the central

canal (see Sorkin & Carlton, 1997).

Primary afferent fibres collectively enter the dorsal horn via dorsal roots and

either terminate in the dorsal horn o f the segment o f entry or send rostro-caudal

projections out o f this segment, projections that terminate several segments away or

as far as the dorsal colunrn nuclei. Such A^-flbre afferent collaterals are in the

medially-running dorsal colunms, whereas AÔ- and C-flbre afferent collaterals run

laterally in Lissauer’s tract. Each primary afferent dorsal horn terminal has clearly-

defined laminae targets creating a dorso-ventral organization which reflects both

receptive field and stimulus modality (Molander & Grant, 1986). This identification

was made possible by the use o f the axonal tracers horse radish peroxidase (HRP)

and wheat germ agglutinin, which selectively label large myelinated and small

unmyelinated fibres, respectively.

36

peripheralendings

dorsal root ganlgia

high intenity noxious stimuli

I/O jTeavH îT^elinated fa s t conducting

intensitynon-noxious

stimuli

thinly myelinated interm ediate conducting

unmyelinated slow conducting

Figure 2. Organization o f the spinal cord laminae in lumbar segments and the termination patterns o f cutaneous primary afferent fibre

input. Noxious stimuli evoked afferent input mediates the release o f peptide neurotransmitters SF (substance P) and CGRP (calcitonin

gene-related peptide) from C-fibre terminals in the superficial laminae o f the dorsal horn. Non-noxious and noxious input converges upon

WDR (wide dynamic range) neurones, which then project to the brain via ascending tracts.

AP-fibre Termination Pattern

Ap-fibre afferents, mediating non-noxious information, enter the spinal cord

via the medial dorsal root. They descend to the deep laminae either medially through

or curve round the superficial medial edge o f the dorsal horn to enter at a more

ventral location where the majority terminate in laminae III-V. Little or no

projections are made directly into laminae I and II. In laminae III and IV Ap-fibres

terminate onto non-nociceptive, low-threshold second order neurones, which project

into the spinocervical (SCT) and spinothalamic (STT) tracts and the postsynaptic

dorsal column (PSDC). In lamina V AP-fibres also terminate upon wide dynamic

range (WDR) neurones, which also receive C-fibre input (see Sorkin & Carlton,

1997).

Aô-fihre Termination Pattern

AÔ primary afferent fibres differ in their termination pattern in a manner

dependent on their peripheral terminal receptor type. Those having high-threshold

mechanoreceptors distribute ipsilaterally to laminae I, IIo and V, with some

projections to contralateral lamina V. Here they terminate upon predominantly WDR

and nociceptive, high-threshold second order dorsal horn neurones which project to

the STT, spinomesencephalic (SMT) and spinohypothalamic (SHT) tracts and to the

parabrachial nucleus (PEN). Aô-fibres that innervate hair cells terminate in lamina

III (see Sorkin & Carlton, 1997).

C-fihre Termination Pattern

C-fibre primary afferent fibres terminate in the superficial dorsal horn.

Furthermore, peptidergic TrkA-positive nociceptors distribute to laminae I and IIo,

whereas the non-peptidergic IB4-positive population distribute to Hi. Lamina I

contains nociceptive-specific (mechanical, heat and cold) and WDR dorsal horn

neurones, the majority of which project to the SCT, STT and spinoreticular tract

(SRT) as well as the periaqueductal grey (PAG), PEN and nucleus submedius. In

contrast, the cells o f laminae II only project locally to surrounding segments and not

to ascending tracts, thus they serve as interneurones and relay inputs to deeper

38

laminae, such as the WDR neurones in V. Morphologically they can be distinguished

into excitatory ‘stalk’ cells found in IIo and the inhibitory GAB A neurotransmitter

containing ‘islet’ cells found in Hi. Furthermore neurones of IIo receive input from

high-threshold and thermoreceptive afferents and those o f Hi receive low-threshold

mechanical information (see Sorkin & Carlton, 1997).

Anatomically, and thus functionally, there exists a considerable amount of

convergence within the dorsal horn of the spinal cord, and this is largely mediated by

the WDR neurones located predominantly in lamina V (also IV, VI and some

superficially). Non-noxious and noxious primary afferent inputs, directly from Ap-

fibres and via a multi-synaptic pathway from AÔ- and C-fibres, all terminate upon

these multireceptorial cells. WDR neurones display a dynamic response over a wide

stimulus range such that they can encode stimulus intensity. WDR cells of lamina V

then project to STT, SMT, SRT, SCT, SHT and PSDC. The supraspinal targets of

these ascending tracts are all important in the transmission o f nociceptive

information. In particular the STT terminates in the thalamus, which is a very

important in encoding type, temporal pattern, intensity and topographical

information regarding pain before relay to the cortex. Via interactions with cortical

and limbic regions it is also responsible for sensory-discriminative and emotional

facets o f pain perception (see Craig & Dostrovsky, 1997, for further discussion).

1.3.2.3 Abnormal Dorsal Horn Reorganization

Damage to peripheral nerve can impact upon the anatomical organization of

primary afferent termination patterns within the dorsal horn o f the spinal cord such

that abnormal changes arise that may underlie some o f the experienced neuropathic

pain symptoms and characteristics. As might be expected it is likely that peripheral

nerve injury results in loss of primary afferent input into the dorsal horn, indeed after

peripheral nerve transection a substantial loss of C-fibre terminals has been observed

in lamina II (Castro-Lopes et aL, 1990). Further to this deafferentation, nociceptive

input to dorsal horn neurones is reduced, which could underlie the occurrence of

sensory deficits experienced by neuropathic pain patients (Fields et aL, 1998).

39

Contrary to this, deafferentation has the ability to induce primary afferent

regeneration and repair activity, as observed in the periphery. Whilst this can be

beneficial in restoring original circuitry, more often than not in the development of

neuropathic pain, aberrant re-wiring may underlie the positive neuropathic pain

symptoms such as allodynia. It has been demonstrated by the use of cholera toxin B-

conjugated (CB)-HRP tracing that the central projections o f intact AP-fibres within

the deep dorsal horn, that convey non-noxious information, may make abnormal

dorsal-oriented sprouts into lamina II after axotomy (Woolf et aL, 1992). This is also

evident after SNL (Lekan et aL, 1996) and CCI (Nakamura & Myers, 1999).

Alongside the previously mentioned induction o f substance P expression in AP-

fibres, sprouting may permit innocuous input to reach the superficial dorsal horn, a

region synonymous with pain transmission, whereby their terminations may activate

nociceptive-specific spinal cord neurones (see W oolf et aL, 1995). This appears a

rational mechanism for touch-evoked pain however it is difficult to prove clinically.

New evidence has come to light, which may support a role for anatomical

sprouting in the dorsal horn. After nerve crush-induced neuropathic pain innocuous

stimulation o f the target tissue evoked c-fos, an immediate early gene product

expressed after noxious neuronal activation, in superficial dorsal horn neurones and

activated PBN neurones, a major supraspinal termination site in the rat for

nociceptive-specific superficial dorsal horn neurones (Bester et aL, 2000).

Innocuously activated Ap-fibres, could activate nociceptive specific neurones in the

superficial dorsal horn if their termination pattern extended dorsally. Additionally or

alternatively, activation of superficially terminating, low-threshold mechanoreceptive

C-fibres could be involved (Vallbo et aL, 1999). Also, low-threshold stimulation has

been demonstrated to evoke synaptic potentials in lamina II neurones 3 weeks after

nerve injury (Kohama et aL, 2000), an area which normally only responds to high

threshold stimuli.

AP-fibre sprouting could possibly be triggered by the physical absence of

normal C-fibre afferent terminals in lamina II after their degeneration. (Doubell et

aL, 1997), yet dorsal rhizotomy produced lamina II vacancy was shown not to induce

Ap-fibre sprouting (Mannion et aL, 1998). More likely is the involvement of trophic

40

factors and chemoattractants, yet this is still yet to be proven. Suggestions include a

possible upregulation of growth related proteins, such as GAP-43 (Woolf et aL,

1990), and injured C-fibre terminal secretion o f an Ap-fibre attractant (Doubell et

aL, 1997; Mannion et aL, 1998). AP-fibre sprouting after axotomy has been

demonstrated to be inhibited by application o f NGF, but since large diameter DRG

cells to not express the appropriate TrkA receptor, the familiar explanation of

‘disrupted axonal transport o f peripherally-derived NGF’ seems unfeasible as a

causal mechanism.

Certain doubts surround the phenomena o f AP-fibre sprouting and its

functional contribution to neuropathic pains. Some discrepancies have come to light

concerning the specificity of CB-HRP utilized to identify Ap-fibres and their

sprouting. It has been proposed that after axotomy, small nociceptive neurones

express novel cell membrane glycoconjugates that mediate the uptake of CB, evident

as an altered neuronal labelling profile after nerve injury (Tong et aL, 1999). This

might imply that CB-HRP labelling in superficial laminae following peripheral nerve

damage may not represent Ap-fibre sprouting. However others do not describe

injury-induced alterations in DRG neuronal cell size distribution o f CB-HRP

(Bennett et aL, 1996). As to the functional relevance of sprouting, it cannot be the

sole mediator o f allodynia for several reasons. Electrophysiological studies have

shown that established behaviourally antiallodynic drugs have little impact upon the

Ap-fibre-evoked response (Chapman et aL, 1998a; Suzuki et aL, 1999). Furthermore,

the time scale o f functional reorganization of AP-fibre terminals does not correlate

with the rapidity with which allodynia can be manifest. Allodynia is observed 2 days

after nerve injury (Bennett & Xie, 1988; Shir & Seltzer, 1990; Kim & Chung, 1992),

yet CB-HRP labelling (W oolf et aL, 1995) and low threshold evoked neuronal

activity (Kohama et aL, 2000) is not observed in lamina II until one and three weeks,

respectively. Structural reorganization as a basis for allodynia cannot be responsible

for the induction of allodynia yet it may have relevance to its maintenance since

sprouting demonstrated by CB-HRP labelling is maximal two weeks after nerve

injury and still apparent at six months (Woolf et aL, 1995).

41

1,3.2.4 Excitatory Neuropeptide and Amino Acid Receptor Systems

Substance P

Substance P is the predominant excitatory neurotransmitter peptide and acts

at the N K l receptor to mediate slow depolarizations and sustained excitatory

postsynaptic potentials (EPSPs), which are important in the recruitment o f NMDA

receptor for activation, as will be discussed later. NK receptor activation, via a G-

protein mechanism, results in the release o f Ca from intracellular stores and protein

kinase C (PKC)-mediated NMDA receptor phosphorylation (a further recruiting

action). Downstream effects o f elevated intracellular Ca " involve enhancement of

neuronal excitability, including NO synthesis that is implicated in the pro­

nociceptive actions o f substance P on dorsal horn neurones (Radhakrishnan et aL,

1995). Substance P containing neurones are localized to the superficial dorsal horn,

in particular lamina II, where its mRNA is also found (Gibson et aL, 1981; Hunt et

aL, 1981; Westlund et aL, 1990). Rhizotomy studies have demonstrated that, as

already discussed, nociceptive primary afferent fibres give rise to about half these

substance P containing terminals, whereas intrinsic dorsal horn neurones give rise to

the remaining half. NK l receptors are found on spinal cord neurones in the

superficial laminae and on deeper cells with dendrites located superficially (Todd et

aL, 2000). Substance P is only released in response to noxious and persistent

stimulation (Duggan et aL, 1988), mainly owing to the fact that C-fibres, not Ap-

fibres, express substance P. The actions o f substance P may be enhanced after nerve

injury. Upregulation o f NK-1 receptors occurs in the superficial dorsal horn

(Abbadie et aL, 1996), neurokinin antagonists attenuate spinal sensitization after

sciatic nerve section (Luo & Wiesenfeld-Hallin, 1995), and AP-fibres phenotypically

transform allowing them to synthesize substance P, as already discussed (Noguchi et

aL, 1995). In contrast, a decrease in spinal substance P itself is observed after nerve

injury (Na et aL, 2001). In addition to this, mechanical hyperalgesia has been shown

to be attenuated in NK-1 receptor knockout mice (Mansikka et aL, 2000).

Calcitonin Gene-Related Peptide

CGRP is another important excitatory peptide although CGRP-containing

terminals are less abundant than those for substance P. CGRP is found in lamina I, II,

42

V and X, but in contrast to substance P, unmyelinated and small diameter myelinated

primary afferent fibres are its only source. Little CGRP is generated in intrinsic

spinal cord neurones. Activation of CGRP dorsal horn receptors probably serves to

enhance the actions of substance P, but its role is comparatively less well defined due

to a lack o f useful antagonists, and may involve positive coupling to adenylate

cyclase (AC). CGRP, like substance P, produces slow membrane depolarizations and

increased intracellular Ca^ levels by both Ca " influx and release from internal

stores (Miletic & Tan, 1988; Oku et aL, 1988; Wimalawansa, 1996).

Excitatory Amino Acids

The EAAs comprise glutamate and aspartate, and as already mentioned are

found in both small and large primary afferent fibres (Battaglia & Rustioni, 1988;

Maxwell et aL, 1990). Within the spinal cord, glutamate-containing fibres and

terminals are found in laminae I-IV, the majority arising from primary afferent

neurones, however some are intrinsic. Aspartate-containing terminals are found in

laminae I-III and coexistence studies indicate that these are separate to the glutamate

population (Merighi et aL, 1991). The EAAs are considered to be fast

neurotransmitters and their substantial localization to the superficial dorsal horn

implies an important role in pain transmission. Glutamate is predominantly found co­

localized with substance P (Battaglia & Rustioni, 1988; Merighi et aL, 1991)

ensuring that both are probably released from afferent nociceptive fibres upon

noxious stimulation.

The glutamate receptors are subdivided into the slow, G-protein-coupled

metabotropic receptors and the fast, ligand-gated, ionotropic receptors. lonotropic

receptors are further characterized by synthetic agonists into AMP A (which

contribute to fast transmission o f acute innocuous and noxious stimuli), kainate and

NMDA receptors (see Ozawa et aL, 1998). All three are prominently localized to the

superficial dorsal horn, with some found in deeper laminae which may correspond to

receptors on WDR neurones (Bonnot et aL, 1996; Yung, 1998). The NMDA receptor

has been the most extensively studied glutamate receptor owing to its ubiquitous

CNS expression. It is an ionotropic receptor coupled to a cation channel, which

under resting conditions is quiescent due to intra-channel Mg " block. Only upon the

43

binding of glutamate, in addition to the co-agonist glycine, and sufficient membrane

depolarization to relieve the Mg^ block, is the complex activated and Ca " influx

occurs.

Prolonged peripheral input into the spinal cord via nociceptive C-flbres is

known to enhance dorsal horn neuronal responses to subsequent afferent input - an

experimental phenomenon known as ‘wind-up’ used to study central sensitization

and spinal cord hyperexcitability (Dickenson, 1995). The NMDA receptor is critical

for wind-up and thus is important in events that enhance and prolong sensory

transmission (Seltzer et aL, 1991; Dickenson, 1995). Central sensitization is

characterized by reduced thresholds to stimuli and enlarged receptive fields

(McMahon et aL, 1993) and at the level of the spinal cord NMDA receptor activity is

heavily implicated. The inherent properties of the NMDA receptor/channel complex

described permit its role in the development o f wind-up and central sensitization.

Under conditions o f a repetitive nociceptive afferent input, sustained release of

glutamate, substance P and CGRP activate AMP A and neuropeptide receptors, the

actions of which summate and relieve Mg " block, thus recruiting NMDA receptors

into glutamatergic transmission. Ca influx through the NMDA receptor itself

mediates further membrane depolarizations recruiting more NMDA receptors ... and

so on. These convergent events result in amplified activation of the NMDA receptor.

Increased and sustained Ca " influx results, which sets in motion signal transduction

cascades culminating in secondary modifications of ionic currents via receptor/ion

channel phosphorylation, as well as long term alterations in receptor/neurotransmitter

expression via gene transcription processes. Central sensitization is normally kept

under control by inhibitory controls, yet in neuropathic pain central sensitization can

persist and become pathological.

After nerve injury there is evidence to show an ipsilateral enhancement of

glutamate release (Kawamata & Omote, 1996) and an upregulation o f glutamate

receptors (Harris et aL, 1996; Croul et aL, 1998). Administration of NMDA receptor

antagonists in models o f neuropathic pain have proved effective in reducing

hyperalgesic (Mao et aL, 1993) and allodynic (Yaksh, 1989) behaviours, as well as

dorsal horn neuronal responses (Suzuki et aL, 2001). Likewise, the hypothesized

44

roles for NMDA receptors in central sensitization and its link to the development of

neuropathic pain have been proven in clinical neuropathic pain states with the use of

antagonists, but with limiting side-effects (Sang, 2000).

1.3.2.5 Inhibitory Receptor Systems

Opioids

The endogenous inhibitory opioid system has been manipulated to provide

pain relief since the discovery o f opium, o f which morphine is the main active

component. Cloning and isolation techniques have established three classical opioid

receptor types, namely 5, k, each showing considerable sequence homology to one

another (Dickenson & Suzuki, 1999). The ]li opioid receptors (and Ô, k and ORLl

receptors) have a seven transmembrane domain receptor structure and are coupled to

G-proteins. Upon extracellular binding o f the appropriate ligand a conformational

change in the receptor is elicited that inhibits cAMP formation, via pertussis (PTX)

sensitive G-proteins, that subsequently opens channels (|X, ô and ORLl

receptors). The resultant neuronal hyperpolarization leads to a decrease in the

opening o f VDCCs. In the case of K opioid receptor activation this reduction of Ca "

influx is direct. The |X, Ô and K receptors are endogenously activated by the

endorphin, enkephalin and dynorphin peptides, respectively, and thus mediate

membrane hyperpolarization, reducing neuronal excitability. Manipulation of these

by exogenous agonists is thus extremely valuable as a means o f exerting analgesia

(Table 2). A recent addition to the opioid receptor family is the opioid-like receptor

(ORL-1) (Wick et aL, 1994; Peluso et aL, 1998). Endogenously activated by the

peptide nociceptin (or orphanin FQ), this receptor shows substantial sequence

homology to the others, yet it only displays a low affinity for the universal opioid

receptor antagonist, naloxone. Its physiological function and pain modulation role

and is not yet well defined, but under investigation (Carpenter & Dickenson, 1998;

Darland et aL, 1998).

45

Table 2. Agonists and antagonists o f the opioid receptors.

.TWif Enkephalins ^DynoiphinsEndogenous

agonist

Endorphins ociceptm/

Synthetic

agonist

Morphine

DAMGO

-DPDPE f c . None '4

-::Nalokone,.

Nor-BNI"^

Antagonist Naloxone

CTAP

Naloxone

Naltrindole

Abbreviations: DAMGO ([D-Ala^,N-Me-Phe'^,Gly-of] enkephalin); DPDPE ([D-

Pen^,D-Pen^] -enkephalin); CTAP (D-Phe-Cys-Tyr-D-Trp-Arg-Thr-Pen-Thr-NH2);

Autoradiographic and immunohistochemical techniques have demonstrated

that within the spinal cord, opioid receptors are mostly located in the superficial

dorsal horn (laminae I and II), with a smaller population in deeper layers (Besse et

aL, 1990b; Rahman et aL, 1998). The contribution o f p, 5, and K receptors to the

total opiate binding throughout the spinal cord is estimated at 70%, 24% and 6%,

respectively (Besse et aL, 1990a,b; Backonja et aL, 1994), at a predominantly (>

70%) presynaptic location relating to the central terminals o f only small diameter

nociceptive primary afferents. Opioid receptors are synthesized in small diameter

DRG cells bodies and transported centrally and peripherally. This implies that the

main m echanism o f spinal opioid analgesia, w hether it be endogenous or

exogenously mediated, is via activation o f presynaptic opioid receptors, which act to

selectively decrease transm itter release from nociceptive afferents and thus

nociceptive transmission leaving innocuous-evoked activity intact. Indeed, spinally

applied m orphine can reduce substance P and CGRP release after noxious

stimulation (Go & Yaksh, 1987) and excitatory but not inhibitory lamina II synaptic

transmission in is inhibited by a presynaptic opioid mechanism (Kohno et aL, 1999).

The remaining 30% of opioid receptors are located postsynaptically on

interneurones, and the dendrites o f projections cells (Besse et aL, 1990a), visualized

functionally as agonist-stimulated receptor internalization (Trafton et aL, 2000).

Here, any opioid-mediated cell hyperpolarization will not exert nociceptive-specific

46

effects, and from electrophysiological studies a small inhibition of Ap-fibre evoked

responses can be observed (Dickenson & Sullivan, 1986). Since this inhibitory effect

is much less pronounced than that observed on the C-fibre evoked response, this

confirms that the predominant site o f action o f spinal opioids is via presynaptic

opioid receptors on central terminals o f nociceptive afferents (Ossipov et aL, 1997).

Activation o f the opioid system as a whole can mediate considerable

antinociception as demonstrated by various pre-clinical studies (Mao et aL, 1995;

Ossipov et aL, 1995a; Ossipov et aL, 1995b) and its widespread clinical

manipulation (McQuay et aL, 1992). Endogenous release o f the endorphins and

enkephalins from intrinsic spinal neurones is stimulated only by high-intensity

stimuli (Cesselin et aL, 1984; Lucas & Yaksh, 1990) and under normal conditions

the opioid antagonist naloxone does not produce hyperalgesia (Yaksh, 1989),

suggestive o f no opioid-mediated tonic control. However, the opioid system displays

a substantial amount o f plasticity in persistent pain states. Whilst such changes are

beneficial in the presence of inflammation (Ossipov et aL, 1997), neuropathic pain

due to peripheral nerve damage more often than not displays reduced sensitivity to

opioids. This is evident both pre-clinically (Mao et aL, 1995; Ossipov et aL, 1995a)

and clinically (Portenoy et aL, 1990; Jadad et aL, 1992), and is surrounded by much

controversy. Mechanisms by which opioid sensitivity may be reduced after nerve

injury include a reduction o f spinal opioid receptors (Porreca et aL, 1998), non­

opioid receptor-expressing Ap-fibre-mediated allodynia, increased cholecystokinin

antagonism (Nichols et aL, 1995) and NMDA-mediated dorsal horn neuronal

hyperexcitability, likely requiring a greater opioid inhibitory counteraction (see

Dickenson, 1997). These possibilities will be discussed further in chapter 7.

It should be briefly mentioned that opioid receptors are distributed

throughout the CNS especially in areas other than the spinal cord also concerned

with nociceptive processing, such as the thalamus and supraspinal midbrain and

brainstem structures - the PAG and the RVM (rostroventromedial medulla).

Exogenous application of morphine into these sites elicits antinociceptive effects by

increasing the activity of inhibitory descending controls that terminate in the dorsal

horn (Heinricher, 1997). The pathway that descends from the RVM contains mostly

47

enkephalin, serotonin, GAB A and glycine containing fibres. The pharmacology of

these modulatory systems in the dorsal horn will now briefly be mentioned.

Inhibitory Amino Acids

GAB A is the major inhibitory neurotransmitter in the CNS and its extensive

distribution permits it to tonically control spinal cord excitability. The majority o f

G ABA terminals arise from intemeurones, though some are from descending tracts

(Millhorn et aL, 1987), and are most concentrated in the superficial dorsal horn

(laminae I-III) (Hunt et aL, 1981; Magoul et aL, 1987). GABA-ergic terminals can

also contain other substances, such as glycine (Todd et aL, 1996), galanin (Simmons

et aL, 1995), enkephalin (Todd et aL, 1992) and neuropeptide Y (Rowan et aL, 1993)

in different populations, and they contact mainly nociceptive A5- and C-fibre

terminals (Bernardi et aL, 1995). There are two receptors for GAB A in the spinal

cord, namely G ABA a and G A BA b , both o f which are found pre- and

postsynaptically on nociceptive afferents (Desarmenien et aL, 1984). GABAa is a

ligand-gated chloride channel that can also be modulated by barbiturates and

benzodiazepines (Schofield, 1989). GABAr is a G-protein-coupled receptor linked to

the inhibition o f Ca influx via and Ca channel mediated actions. Activation of

presynaptic G ABA receptors, most likely GABAr, has been demonstrated to inhibit

neurotransmitter release (Bourgoin et aL, 1992; Malcangio & Bowery, 1994) and

postsynaptically, probably GABA a , they prevent EPSPs via neuronal

hyperpolarization (Alvarez et aL, 1992).

Under normal conditions the GABA-ergic control o f nociceptive

transmission is near maximally activated, indeed block o f G ABA receptors can

mediate nociceptive responses to previously innocuous stimuli (Dickenson et aL,

1997a). Loss o f GABA-ergic control upsets the balance o f excitations and inhibitions

within in the spinal cord favouring the generation o f hyperexcitability and therefore

plasticity within this system is implicated in pathophysiological pain states. CCI and

sciatic nerve transection have been shown to result in a decrease o f spinal G ABA

content and immunoreactivity (Castro-Lopes et aL, 1993; Ibuki et aL, 1997) and

subtype-selective alterations in GABA receptor expression appear to exist (Castro-

Lopes et aL, 1995). GABAr agonists have been shown more effective than GABAa

48

agonists in ischaemia-induced rat spinal cord injury (Hao et ah, 1992b; Xu et ah,

1992), yet electrophysiologically midazolam, a positive allosteric modulator of

GABAa receptor function, was shown to be more effective in inhibiting nociceptive

activity after SNL (Kontinen & Dickenson, 2000).

Glycine is another inhibitory amino acid and is found in neurones within the

spinal cord localized to the superficial dorsal horn mostly in coexistence with GABA

(Todd & Sullivan, 1990). It acts via a strychnine-sensitive receptor that is a ligand-

gated chloride ion channel to inhibit dorsal horn neurones and spinal strychnine

mediates allodynia and nociceptive behaviour in rats (Yaksh, 1989; Sivilotti &

Woolf, 1994), thus its role in nociceptive transmission appears similar to that of

GABA. Little is known regarding changes to the glycinergic system after nerve

injury, however increased glycine levels have been reported (Satoh & Omote, 1996).

Furthermore, in the dorsal horn of neuropathic pain patients the ratio o f aspartate to

GABA and glycine is increased, suggestive of an impaired excitatory/inhibitory

balance (Mertens et ah, 2000). In contrast glycine is a co-agonist at the NMDA

receptor, but spinal levels are such that this affects its activation.

Monoamines

The majority of nociceptive pathways (including SRT and STT) are under

the control o f bulbospinal projections originating in brainstem nuclei. The nucleus

raphe magnus (NRM) gives rise to serotonin (or 5-HT (5-hydroxytryptamine))

containing projections, the locus coeruleus and lateral tegmentum cell systems

contain noradrenaline, and these are released in the spinal cord selectively in

response to high-intensity stimulated afferent input (Dickenson et ah, 1997b).

The role o f descending serotonergic pathways in the modulation o f pain

transmission is somewhat complicated by the multiplicity o f its target 5-HT

receptors, activation o f which exert both pro- and antinociceptive actions (see

Wilcox & Seybold, 1997) dependent on the differing effector mechanisms and

neuronal locations. Dense 5-HT labelling is found in laminae I and II, the majority

arising from supraspinal sites, in addition to a small intemeuronal population (see

49

Wilcox & Seybold, 1997). Stimulation o f the nucleus NRM displays a biphasic,

excitatory followed by inhibitory, influence on dorsal horn neurones (Zhuo &

Gebhart, 1997). Importantly, these descending pathways display plasticity after nerve

damage. Chronic pain patients show changes in spinal levels o f 5-HT and deep

laminae terminating serotonergic terminals sprout into lamina II (Wang et a l, 1991;

Marlier et a l , 1992) such that nociceptive transmission is enhanced. Blockade of

serotonin re-uptake by tricyclic antidepressants may account for their noted analgesic

effects, as previously mentioned. In brief, 5-HTib and 5-HTm mediate selective

inhibition of nociceptive neurones and are implicated in headache (Goadsby, 2000).

5-HT2C andiA are positively coupled to phospholipase C and suppress K" current. 5-

HT3 increase intracellular Ca " either directly by activation of a cation channel that

triggers the opening o f VDCCs or via induction o f phospholipase C. The localization

o f 5-HT2A, 2c and 3 on excitatory intemeurones or primary afferent terminals mediates

the pro-nociceptive actions of 5-HT or descending facilitation. 5-HTia are found

densely localized in the superficial dorsal horn, like 5-HT3, and increase currents

and decrease Ca^ currents via negative coupling to AC thus inhibition of these

receptors on inhibitory intemeurones contribute to central sensitization and allodynia

by disinhibition o f WDRs (see Millan, 1997).

Noradrenaline containing spinal cord terminals are localized to laminae I, II

and V, and arise purely from supraspinal sites (see Sorkin & Carlton, 1997). Within

the spinal cord there are two (%2-adrenergic receptors. (%2C have a presynaptic location

in primary afferents and 0C2A are found in secondary dorsal horn neurones (Nicholas

et a l , 1993). They inhibit dorsal horn neurones by presynaptic inhibition of

neurotransmitter release (Pang & Vasko, 1986; Go & Yaksh, 1987), probably via a

PTX-sensitive Gi/o-protein coupling and postsynaptically by hyperpolarization,

mediated by coupling to inwardly rectifying channels. Administration of the a-

adrenoceptor antagonist phentolamine has been shown not to produce hyperalgesic

effects (Yaksh, 1989), and thus like the endogenous noradrenergic system is unlikely

to exert ongoing modulation o f nociception. There is little evidence so far but

changes after nerve injury enhance the analgesic effect o f (X2-agonists (Suzuki,

personal communication).

50

Adenosine

The purines, adenosine, produced by ATP metabolism, alongside ATP

itself, are implicated in the modulation o f nociception, although centrally the role o f

the adenosine receptor system is more defined (Sawynok, 1998; Dickenson et al.,

2000). Adenosine-like immunoreactivity is detected in lamina II of the dorsal horn.

Three differentially G-protein-coupled receptors for adenosine have been

characterised. The A1 receptor inhibits AC whereas A2 and A3 stimulate AC. Via

mainly a postsynaptic location, the A1 receptor predominates the antinociceptive

effects o f spinally applied adenosine and adenosine analogues, exerted through

channel mediated hyperpolarization (Sawynok, 1998; Dickenson et al., 2000).

Additionally those receptors found on the terminals o f primary afferent fibres inhibit

Ca current and neuropeptide release. Manipulation o f the adenosine system with

analogues has been shown to be effective against neuropathic pain in both clinical

and pre-clinical studies, in a manner that may be enhanced after nerve injury (Suzuki

et al., 2000a). Furthermore, adenosine is able to inhibit NMDA receptor-mediated

hyperexcitability (Dickenson et al., 2000). Interactions between adenosine and

serotonergic modulation are apparent (Sawynok, 1998), yet the mechanisms behind

such effects remain to be uncovered.

Galanin

Galanin-containing dorsal horn cells are found superficially, localized to

intemeurones, projection neurones and primary afferent terminals, sometimes in

coexistence with substance P, CGRP, cholecystokinin (CCK) and NOS (see Sorkin

& Carlton, 1997). There are three cloned galanin receptors, namely GalRl, 2 and 3.

GalRl and 3 couple to the Gi G-protein and inhibit levels of cAMP, whereas GalR3

couples to Gq and activates phospholipase C. Both GalRl and 2 mRNA have been

demonstrated to be present in the spinal cord and DRG of rats (for review see

Branchek et al, 2000). In normal animals galanin release seldom occurs in response

to noxious and innocuous stimulation (Morton & Hutchison, 1989) and low levels

are pro-nociceptive (Wiesenfeld-Hallin et al., 1992). As already mentioned, nerve

injury upregulates galanin synthesis and high doses appear to be antinociceptive

(Wiesenfeld-Hallin et al., 1992). Thus, the spinal administration o f galanin has

complex but predominantly inhibitoiy/antinociceptive effects, which are enhanced in

51

models of neuropathy (see Xu et a l , 2000, for review).

Neuropeptide Y

The Y2 receptors for neuropeptide Y normally occur in small diameter DRG

neurones and their activation inhibits Ca currents probably via a Go coupling.

Receptor activation can block the release of substance P from cultured neurones

(Walker et a l , 1988) and spinal neuropeptide Y has been reported to exert potent

analgesic effects (Hua et a l , 1991). Spinal neuropeptide Y largely arises in

descending neurones, but as previously mentioned, nerve injury induces the normally

quiescent expression of neuropeptide Y in sensory primary afferents, which is also

paralleled by distributional shift of Y% receptor expression to include Ap-fibre

terminals. Thus the neuropeptide Y receptor system appears to display nerve injury-

induced changes that are beneficial.

Cannabinoids

There are two identified cannabinoid receptors, CBi and CB2. CBi is found

in the spinal cord where its mRNA has been located to medium and large DRG cell

bodies (Hohmann & Herkenham, 1999) and at both presynaptic and postsynaptic

spinal sites (Hohmann et a l , 1999). Its activation here has been demonstrated to

release dynorphins (Houser et a l , 2000) and may be involved in endogenous

inhibitory tone (Chapman, 1999; Drew et a l , 2000). Cannabinoid agonists are also

antihyperalgesic in the presence of inflammation (Martin et a l , 1999b; Drew et a l ,

2000) and nerve injury (Fox et a l , 2001), possibly mediated via actions in both the

CNS and in the periphery, and activity within this system may increased in the

presence o f persistent pain (Martin et a l , 1999b).

52

1 .4 V o l t a g e -D e p e n d e n t C a l c iu m C h a n n e l s

It appears that nerve injury resulting in a neuropathic pain state initiates

many changes to the peripheral and CNS that are accountable for the various

symptoms that manifest. To summarize these include changes in spinal cord

connectivity, loss of intrinsic modulatory systems and upregulation o f excitatory

process including the emergence of spontaneous activity and central sensitization.

Pivotal to many o f these alterations is the concentration o f intracellular Ca ions. In

addition to release from internal stores heavily implicated in 2° messenger systems

and gene induction, the major influx route is across the plasma membrane via the

NMDA-receptor/channel complex and VDCCs, in response to neurotransmitter

receptor activation and membrane depolarization. VDCCs are critical to the sensory

pathway in various aspects, but more specifically they permit the synaptic

transmission of sensory information from the periphery to the brain via the control of

depolarization-coupled neurotransmitter release.

Primary Structure o f VDCCs

VDCCs can be found in the membrane o f many cell types throughout the

body where they mediate Ca^ entry into the cell in response to membrane

depolarization. Originally solubilized and purified from skeletal muscle (Curtis &

Catterall, 1984), VDCCs are hetero-oligomeric complexes consisting o f various

combinations of an a\ subunit with auxiliary 0(26, P and y subunits (Figure 3).

Hydrophobicity plots, glycosylation and biochemical analysis o f the subunits has

revealed their primary structure. The a\ subunit forms the pore of the channel and is

predicted to have four domain repeats (I - IV), each having six transmembrane

segments (SI - S6) with a membrane-associated loop between S5 and S6 lining the

pore, and the 84 segment forming the voltage-sensor required for activation. The p

subunit is predicted to have no transmembrane regions and associates with « i

intracellularly, in contrast to the y subunit which has four transmembrane segments.

The 0C2 subunit is entirely extracellular with many glycosylation sites and is anchored

to the membrane through a disulphide bond with the 5 subunit. Using in vitro

expression systems, generally the a\ subunit is enough to form functional channels

and the auxiliary subunits serve to modulate its gating and current characteristics via

53

enhancing expression levels, shifting voltage-dependence of activation and

inactivation to more negative potentials and increasing the rate of inactivation. In

vivo, non-neuronal VDCCs comprise all five subunits, whereas neuronal channels

lack the y subunit (Ahlijanian et al., 1990; Witcher et al., 1993; Liu et al., 1996).

Several different classes o f VDCCs exist and these were originally

identified by their electrophysiological and pharmacological profiles into the L-, N-,

P-, Q-, R- and T-types, each subserving different functional roles relative to their

cellular location. Subsequent advances in molecular biology have identified ten

genes encoding the main pore-forming a i subunit, termed aiA to an, and a recently

devised nomenclature groups these into three families, Ca^l, 2 and 3, based upon

structural and functional characteristics (Ertel et al., 2000). Cayl and Cay2 comprise

the high voltage-activated (HVA) VDCCs, which allow Ca influx upon substantial

membrane depolarization, such as that mediated by an action potential, whereas Cay3

channels are the low voltage-activated (LVA) VDCCs, such that they permit Ca "

flux at resting membrane potentials (Table 3).

L-type VDCCs (CaD)

There are currently four Cayl channels (Cay 1.1, 1.2, 1.3, 1.4 or a iA , ic, id , if,

respectively) which show 75% sequence homology amongst themselves. These

comprise the L-type VDCCs and are selectively blocked by 1,4-dihydropyridines

(DHPs), phenylalkylamines and benzothiazepines. They can be located throughout

the body and functions include excitation-contraction coupling in skeletal and

cardiac muscle, hormone secretion in endocrine cells (Milani et al., 1990) and tonic

neurotransmitter release in the retina, a ic and am can be found in neurones, through

which Ca influx contributes to gene regulation and is important in the integration

o f synaptic inputs (Bean, 1989a). Immunohistochemical studies have demonstrated

the neuronal presence o f am and am subunits upon the cell bodies and dendrites in a

variety o f cell populations, including the spinal cord, where they are distributed

throughout the deep dorsal and ventral horns (Ahlijanian et al., 1990; Hell et a l ,

1993; Westenbroek et al., 1998).

54

COOH

* 2* COOH COOH COOH

COOH

extracellular

intracellular

Figure 3. Siihunit strucutre and composition o f a voltage-dependent calcium

channel. Predicted a helices are depicted as cylinders. As shown fo r domain III, 6

transmemhrane segments form a domain and the 4 domains comprise the pore-

forming a j subunit.55

Table 3. Classification, cellular localization and functional significance o f voltage-dependent calcium channels.

VDCC I Current type type Übiïïtiixc

Tissuelocalization

Subcellularlocalization

Spinal cord distribution

Specific Function blocker

M Skeletal muscle T-tubulesC a v l . l. ( H V A )

C a v l.2 Cardiac muscle(HVA) ^ Smooth muscle

Neurones PancreasCavl.3

» (HVA)

Cavl.4 L ._________ (HVA)

Endocrine tissue j Neurones M Retina

Membrane surface, t- tubule sarcolemna Cell somaCell soma and larger dendrites o f neurones

y deep dorsal and ventral horns

y deep dorsal and ventral horns

DHPs Excitation-contraction coupling

DHPs Excitation-contraction coupling,hormone secretion, gene regulation

DHPs Hormone secretion, generegulation

Tonic release o f neurotransmitters

Cav2.1^(HVA)

Cav2.2(HVA)^^

Cav2.3___________ (HVA)

Neurones

Neurones

Neurones

Nerve terminals and dendritesNerve terminals and dendrites

Soma, dendrites and terminals

y throughout

y mostly concentrated in superficial laminae

œ-agatoxinIVA(0-conotoxinGVIANone

N euro transmitter release

Neurotransmitter release

Neurotransmitter release

Ca,3.1 .T^ (LVA)

Ca.3.2 T(LVA) :

Ca.3.3 OCii r - «

Neurones Cardiac muscle Skeletal muscle Neurones Cardiac muscle Neurones

Soma and dendrites y low levels throughout None

Soma and dendrites

Soma and dendrites

y mostly restricted to Nonesuperficial laminae

y throughout, mainly in Nonelaminae III-IV

Pacem aking, gradual depolarization for multiple APs and oscillatory behaviour, depolarizes cells to threshold for other channels

Abbreviations: VDCC (voltage-dependent calcium channel); HVA (high voltage-activated; LVA (low voltage-activated); DHPs

(dihydropiridines); APs (action potentials).

LAo\

P/Q-type VDCCs (Cavl.l)

There are three members of the Cay2 family and these are purely localized

to neuronal populations and show 70% sequence homology amongst themselves, but

less than 40% with the Cayl channels. Cay2.1 encompasses both the P- and Q-type

currents (often referred to as P/Q-type). These are generated by alternative splicing

of the ttiA gene, differentially blocked, in a reversible manner, by (o-agatoxin IVA, a

48 amino acid peptide isolated from the funnel-web spider of the Agelenopsis genus

(see Olivera et al., 1994, for review). P-type current, first recorded in Purkinje

neurones (Linas et al, 1989), is specifically blocked over a 100 nM range, whereas

Q-type current, first recorded in cerebellar granule neurones (Randell & Tsien,

1995), is blocked at concentrations over 1000 nM.. P/Q-type VDCCs are localized to

nerve terminals and dendrites in a number of neuronal cell types throughout the

brain, as shown by antibody labelling and in situ mRNA hybridization (Bertolino &

Llinas, 1992; Mintz et a l , 1992; Turner et a l , 1992; Turner et al., 1993; Stea et al.,

1994), and P/Q-type current has been identified in spinal and DRG cells (Mintz et

al., 1992). The terminal localization o f P/Q-type VDCCs within neurones is

indicative of their well-documented importance in synaptic transmission. Ca " influx

through P/Q-type VDCCs in response to neuronal depolarization initiates the release

of fast neurotransmitters at synapses, as shown in synaptosome and in vitro neuronal

preparations (Turner et al., 1992). co-agatoxin IVA has been demonstrated to block

excitatory (Luebke et al., 1993; Castillo et al., 1994; Yamamoto et al., 1994) and

inhibitory (Takahashi & Momiyama, 1993) synaptic transmission. Thus, P/Q-type

VDCCs may be involved in the release of glutamate, aspartate, dopamine, serotonin,

noradrenaline, GAB A and glycine (Turner et al., 1992; Takahashi & Momiyama,

1993; Turner et al., 1993; Kimura et al., 1995; Miljanich & Ramachandran, 1995).

Furthermore, P/Q-type VDCCs can be modulated by various G-protein-coupled

pathways that are important for regulation of synaptic transmission. For example,

activation o f GABAb receptors mediates a reduction in P-type current and a decrease

in the rate o f activation (Mintz & Bean, 1993).

57

N-type VDCCs (Cavl.l)

Cay2.2 is the N-type (am ) current and is sensitive to block by the co-

conotoxins, which are 24 - 29 amino acid peptides isolated from the venom of fish-

hunting marine snails of the Conus genus (see Olivera et ah, 1994, for review). In

particular, co-conotoxin GVIA exhibits high affinity, irreversible block o f N-type

VDCCs via a competitive mechanism. Autoradiographical and electrophysiological

studies have shown N-type VDCCs to be widely expressed throughout the brain in

numerous cell types (Regan et al., 1991; Ishibashi et ah, 1995) and in the spinal cord,

most concentrated in laminae I and II o f the superficial dorsal horn where

nociceptive primary afferents synapse upon spinal pain transmission neurons (Kerr et

al., 1988; Gohil et al., 1994), and in DRG (Regan et al., 1991). In vitro studies have

demonstrated the requirement o f Ca^ influx through N-type VDCCs for

depolarization-coupled neurotransmitter release. Release of CGRP from rat spinal

afferents (Santicioli et al., 1992); sensory neuropeptides from nociceptive dorsal

spinal cord afferents (Maggi et al., 1990) and substance P from primary sensory

neurones in culture (Holz et al., 1988) have been shown to be co-conopeptide

sensitive. Glutamatergic synaptic transmission between DRG cells and spinal cord

neurones has also been inhibited by block of pre-synaptic N-type VDCCs (Gruner &

Silva, 1994). N-type VDCCs are also modulated through G-protein-coupled

pathways, more so than P-type, such that Ca current is reduced by a positive shift

in voltage-dependence of activation and a slowed rate of activation. Inhibition of N-

type VDCC current is observed following activation o f a number o f receptors,

including opioid, GABAb and neuropeptide Y (see Lynch III, 1997), and appears to

be mediated by the interaction of GPy at a number of sites on the am subunit

(notably the intracellular loop between domains I and II). Additionally, in certain

cells N-type current may be tonically inhibited (Kasai, 1991; Kasai, 1992).

Conversely, activation o f PKC by various neurotransmitter systems can overcome

the G-protein-mediated inhibition of N-type VDCCs (Swartz, 1993; Swartz et al.,

1993), via a mechanism that also appears to involve phosphorylation of the I-II linker

(Zamponi et al., 1997).

Further substantiating evidence o f a role for N- and P/Q-type channels in

neurotransmission comes from their direct interaction with the SNARE protein

58

complex at the synprint site in the intracellular II-III loop. This mediates the fusion

o f secretory vesicle and plasma membranes upon Ca " influx through VDCCs

permitting neurotransmitter release, which in N-type VDCCs has also been shown to

modulated by PKC and Ca^" /calmodulin protein kinase II (Yokoyama et al., 1997),

providing feedback inhibition. Whilst L-type VDCCs have been shown not to be

involved in fast synaptic transmission (Daniell et al., 1983; Suszkiw et al., 1986;

Pfrieger et al., 1992), recent studies show that these channels may also exhibit

SNARE protein complex interactions that also show feedback modulation. This has

importance in the endocrine secretion of hormones for which they are responsible

(Wiser et al., 1999; Yang et al., 1999).

R-type VDCCs (Cav2.3)

The third member of the Cay2 family is Cay2.3, the R-type (am) current,

resistant to all previously mentioned agents (Zhang et al., 1993a), thus its

distribution, function and modulation are somewhat less defined. However residual

HVA Ca " current after inhibition of L-, N-, P/Q-type current with their specific

blockers, is presumably mediated by R-type VDCCs and can be further distinguished

by its electrophysiological properties. R-type current is found on a number of

neuronal populations on the cell bodies, dendrites and terminals, and important for

Ca^^-dependent action potentials and neurotransmitter release (Ishibashi et al., 1995).

T-type VDCCs (Ca^S)

There are three Cay3 channels, Cay3.1, 3.2 and 3.3 (aio, am and an,

respectively), that mediate the T-type current, and these show 30% homology to

HVA channel forming subunits (Cribbs et al., 1998; Perez-Reyes et al., 1998; Lee

et al., 1999). T-type channels are kinetically distinct LVA VDCCs, such that they

permit Ca " flux at resting membrane potentials in skeletal (am) and cardiac muscle

(aiG and am ) and neuronal cells (am, am and an,). They activate at voltages near

the resting membrane potential, inactivate rapidly, deactivate slowly and have a

small single channel conductance (Huguenard, 1996). These unique gating properties

o f T-type channels permits their involvement in pacemaking, low amplitude

oscillations, neuronal bursting, synaptic signal boosting, Ca " entry promotion and

59

lowering threshold for high-threshold spike generation. This current appears to

play an important physiological role in near-threshold phenomena and regulation of

neuronal excitability. In situ hybridization studies on the rat brain have shown that

ociG, ociH and an have unique neuronal distributions, including the dorsal horn o f the

spinal cord and sensory ganglia (Talley et al., 1999). This is complemented by

reported T-type currents in primary sensory neurones (Carbone & Lux, 1984;

Kostyuk et al., 1992; Scroggs & Fox, 1992a; Todorovic & Tingle, 1998) and some

superficial rat dorsal horn neurones (Ryu & Randic, 1990). Unlike the HVA Ca

channels, the functional roles of T-type channels have been hindered by a scarcity of

specific pharmacological agents, as they only show preferential block by nickel ions.

Diversity o f VDCCs

The existence and expression of numerous genes encoding the VDCC pore-

forming a i subunit is not the only means o f generating VDCC diversity, the

auxiliary subunits also contribute by the expression o f different encoding genes and

the generation of splice variants. Currently there are four genes and various splice

variants o f the (3 subunit (see Hofmann et al., 1999, for review), three y subunit genes

(Eberst et al., 1997; Letts et al., 1998; Black & Lennon, 1999), and three genes

encoding for a 2Ô, again with numerous splice variants (Klugbauer et al., 1999;

Hobom et al., 2000). These different subunit types also show differential cellular

distributions and exert differential modulation upon different a i pore-forming VDCC

subunits. Thus, the subunit composition o f a VDCC has influence on its

electrophysiological and pharmacological properties as well as its subcellular

localization and interaction with regulatory proteins. Various combinations o f a i

subunits with the different auxiliary subunits contributes to the diversity o f VDCCs

observed amongst cell types.

Given their role in controlling neuronal excitability and neurotransmitter

release, the role o f VDCCs in the transmission o f pain has been o f great interest.

VDCC antagonists have been demonstrated to be antinociceptive in models of

inflammation, based on behaviour and in vivo electrophysiology, confirming a role

for Ca " influx into neurones in the spinal processing of nociceptive information (see

60

Vanegas & Schaible, 2000, for review). To varying extents N-, P/Q- and L-type

VDCC blockers have been shown to mediate antinociception in inflammatory

behavioural studies (Malmberg & Yaksh, 1994; Malmberg & Yaksh, 1995;

Bowersox et al., 1996; Sluka, 1997; Sluka, 1998). Enhanced neuronal responses

induced by inflammation are also inhibited by N- and P/Q-type blockers, but not L-

type, (Neugebauer et al., 1996; Diaz & Dickenson, 1997; Nebe et al., 1997; Nebe et

al., 1998). In contrast, investigations into the role o f VDCCs in sensory transmission

utilizing models of neuropathy have been limited, however a few behavioural studies

demonstrate that N-type VDCC block can reduce pain-related responses (Chaplan et

al., 1994; Xiao & Bennett, 1995; Bowersox et al., 1996; White & Cousins, 1998). In

contrast block of L- and P/Q-type Ca current in the presence o f neuropathy has

been shown to have little effect (Chaplan et al., 1994; White & Cousins, 1998).

61

1.5 A im s o f this Study

The experiments undertaken for this thesis were performed in order to

investigate the role of VDCCs in nociceptive transmission at the level of the spinal

cord, and to assess any alterations in dorsal horn neuronal excitability and possible

plasticity o f the different VDCCs induced by peripheral nerve damage that may

contribute to altered nociception. VDCC activation is critical for neurotransmitter

release and neuronal excitability, and blockers are antinociceptive in behavioural and

clinical studies, but as yet there remains little electrophysiological data regarding the

role of spinal VDCCs in the processing of neuropathic pain, an important relay site

of sensory information.

In order to achieve these goals I employed the SNL model of neuropathy

(Kim & Chung, 1992), confirmed by behavioural testing over a postoperative period

of two weeks, to induce a neuropathic state. After the establishment of neuropathy, in

vivo electrophysiological studies o f dorsal horn spinal neurones were made to

investigate the effects of various spinally delivered VDCC blockers on a wide range

of electrical and natural-evoked neuronal activity, in comparison to sham-operated

and naïve rats, co-conotoxin GVIA, (O-agatoxin IVA, ethosuximide and nifedipine

were used to inhibit Ca " flux through N-, P/Q-, T- and L-type VDCCs, respectively.

Since neuropathic pain has multiple causes and symptoms, dysfunctional

mechanisms are likely to be diverse. For these reasons I also aimed to investigate the

benefits o f combination therapy using clinically licensed drugs. The anticonvulsant

gabapentin has proved effective in the clinical treatment o f neuropathic pain.

Although its mechanism of action is undetermined it does bind to the VDCC azb

subunit and acts to inhibit excitation. Conversely morphine, not often beneficial in

the treatment o f neuropathic pain, acts to facilitate inhibitory systems, thus the

effects o f low doses of subcutaneous gabapentin and morphine in the same

experimental setting were also investigated. Due to the difficulties in conducting

combination studies in patients with neuropathic pain it was hoped that this approach

may provide a guide to potential improvements in its treatment.

62

CHAPTER 2

METHODS

63

2.1 Sp in a l N e r v e L ig a t io n M o d e l o f N e u r o p a t h y

2.1 .1 S u r g e r y

Male Sprague-Dawley rats (University College London Biological

Services), initially weighing 130 - 150g, were used for SNL surgery. All

experimental procedures were approved by the Home Office and follow the

guidelines under the International Association for the study o f Pain (Zimmermann,

1983). Selective tight ligation of spinal nerves L5 and L6, and a sham procedure

were performed as first described by Kim and Chung (1992). Under gaseous

anaesthesia with a mixture of halothane (3.5% for induction, 1.5% for maintenance)

and a 1:1 flow ratio o f NzOiOi, the rat was placed in a prone position. A midline

incision was made from L4 - S2 and the left paraspinal muscles were separated from

the spinous processes. The L6 transverse process and the sacrum were made visible

by scraping off attached ligaments. Part o f the L6 transverse process was then

removed to expose the L4 and L5 spinal nerves and L6 was identified lying just

under the sacrum. Using 6-0 silk thread, the left spinal nerves L5 and L6 were tightly

ligated distal to their dorsal root ganglion and proximal to their conjunction to form

the sciatic nerve. Hemostasis was confirmed, the wound sutured and the animal

recovered from anaesthesia. A sham operation was performed to produce a control

group, whereby the surgical procedure was identical to that of the experimental

group, but spinal nerve ligation was omitted.

2 .1 .2 B e h a v i o u r a l T e s t in g

For two weeks following surgery the rats were housed in groups of 4, in

plastic cages under a 12/12h day/night cycle and their general health monitored.

Successful reproduction o f the neuropathic model was confirmed by behavioural

testing (postoperative (PO) days 2, 3, 5, 7, 9, 12 and 14). Rats were placed in

transparent plastic cubicles on a mesh floor and allowed to acclimatize before

initiating any tests.

64

2.1.2.1 Sensitivity to Punctate Mechanical Stimuli

Foot withdrawals to trials of ascending von Frey filaments (bending forces

1, 5 and 9 grams: 9.9, 49.5 and 89.1mN respectively), considered non-noxious under

normal circumstances, were quantified. In a trial a single filament was applied 10

times to the plantar surface of the foot, through the mesh floor, for 2-3 seconds each

time. A period o f 3 - 4 minutes was left before commencing with the next filament.

The number o f foot withdrawals (scored out o f 10) to each von Frey filament was

measured on both the ipsilateral and contralateral hindpaws.

2.1.2.2 Sensitivity to cooling stimuli

Foot withdrawals to the application of a drop of acetone to the plantar

region of the foot were quantified. In a trial a drop of acetone was squirted through

the mesh floor via a syringe, 5 times, at 5 minute intervals. The number o f foot

withdrawals (scored out of 5) was measured on both the ipsilateral and contralateral

hindpaws. The acetone was applied with a gentle squirt so as not to evoke a

mechanical response.

2 .2 S p in a l C o r d E l e c t r o p h y s io l o g y

2 .2 .1 S u r g e r y

Subsequent to behavioural testing (PO days 14 - 17), the operated rats were

used for electrophysiological studies, as described previously (Dickenson &

Sullivan, 1986). Anaesthesia was induced by first placing the rats in a sealed box and

filling it with 3.5% halothane in a mixture of 66% N2O and 33% O2, until loss of the

righting reflex. The rat was then placed on a heating blanket on its back and

anaesthesia sustained at 3% via a nose cone. On confirmation o f areflexia, the

trachea was exposed and a cannula inserted though which anaesthesia was

maintained for the duration of the experiment. The rat was placed in a stereotaxic

frame and secured via ear bars and a tooth bar rest, and a rectal probe was inserted to

regulate body temperature (36.5 - 37°C) via an automatic feedback unit connected to

the heating blanket. Lumbar vertebrae LI - L3 were located by rib positioning and a

clamp, attached to the stereotaxic frame, was secured to the vertebrae rostral of LI.

65

This allowed a laminectomy to be performed removing most of LI - L3 to expose

segments L4 - L5 of the spinal cord, which receive afferent input from the toe region.

Another clamp was positioned caudal to the exposed site and to the frame in order to

maintain stability during electrophysiological recordings, and the level of halothane

was finally reduced to 1.7% such that the rats breathed spontaneously and areflexia

was maintained throughout the course of the experiment.

2.2 .2 E l e c t r o p h y s i o l o g i c a l R e c o r d in g s

Extracellular recordings of single convergent neurones, located deep within

the dorsal horn (> 500 pm) were made using a parylene coated tungsten electrode,

recorded and analyzed with a NeuroLog (Digitimer) system (Figure 4). The electrode

was held in a manoeuvrable headstage that was grounded by an earth lead to the

stereotaxic frame and steel table. To minimize interference and non-neuronal

electrical activity, the head stage received electrical signal from both the spinal cord

(A) via the electrode and the general surroundings (B) via a crocodile clip attached to

the rats skin, such that the latter was subtracted from the former. The resultant (A-B)

signal was then filtered and amplified by a series of NeuroLog modules and finally

fed through an audio amplifier to a speaker and oscilloscope to allow the neuronal

activity to be monitored. Adjustment o f the window discriminator permitted selective

data capture over background activity, differentiating cells by their amplitude, such

that only action potentials above a certain height triggered a counting pulse. This was

relayed to the oscilloscope (apparent as a dot over the spike), the latch counters

(showing cumulative count) and to the CED 1401 interface for analysis. The CED

1401 was further coupled to a Pentium computer and the data captured using Spike 2

software (Rate and Post-stimulus histogram functions).

2.2.2.1 Neurone Isolation

A single convergent dorsal horn neurone was isolated that either received

input from the toe region ipsilateral to the spinal nerve ligation or sham procedure, or

from either side o f the spinal cord in non-operated, naïve rats. The electrode, held in

its headstage, could be moved laterally and rostro-caudally outside of the spinal cord,

and dorso-ventrally through the cord using gross and micromanipulator manual

66

A

"b

B-signalfrom rat

Head Stage

A-signal from neurone

Earth

A-B is amplified, f i l te red and fed through audio

speakers and oscilloscope _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ I_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Action potentials above a cer ta in amplitude a re discriminated

he re and fed into th e CED

Frequency of stimulation (0.5Hz), duration of each pulse (2msec), amplitude of c u r ren t

applied and number of pulses in t h e tra in (16) a r e s e t here

Number o f action potentials evoked

during th e band s e t — here (9 0 -300msec)

a re displayed cumulatively here

. U

Pre-Amp»

IN

AC-DC ■'Amp

Filters

O UT-n

1----Background signal (B) is su b trac ted from signal from isolated cell (A)

AudioAmp

O U T -,

SpikeTrigger

OUT-n

Speaker

^ Period Generator

IN

O U T -I

Digital ; Width .

O U T^

S'

Pulsed : a B u f f e r !

O U T ^

Counter

IN

OUT

Delay J

OUT

y Latch " Counters

IN

BB00

Stimulusisolator

,^mputcr

O n

Figure 4. Schematic representation o f data capture, courtesy o f Dr. K. J. Carpenter.

controls. In parallel to slow descent of the electrode through the cord, the peripheral

hindpaw receptive field was tapped with the finger allowing location of the spinal

cord area receiving input specifically from the toe region. A neurone was selected for

an experiment if it responded to both noxious (pinch) and non-noxious (touch)

stimuli and if its action potential amplitude could be counted over background. Two

fine needles, attached to a stimulus isolator module, were then inserted into the

centre o f the cells receptive field, to allow transcutaneous electrical stimulation.

After an interval o f about ten minutes, the C-fibre threshold current was determined

by giving single, electrical pulses of amplitude 0.1 - 3.3 mA incrementally until a C-

fibre response was evoked (distinguished by latency after stimulus). Subsequently,

electrical stimuli were given at 3-times this C-fibre threshold, which even at a

possible maximum intensity of 9.9 mA was not observed to damage the hindpaw.

2.2.2.2 Neurone Characterization

An electrical test consisted of a train of 16 stimuli, o f a 2 ms wide pulse at

0.5 Hz at 3-times the threshold required to evoke a C-fibre response, which were set

using the period generator, digital width and pulse buffer modules respectively, and a

post-stimulus histogram was constructed. Electrically-evoked action potentials,

mediated by different fibre types, were separated on a latency basis, determined by

their known conduction velocities, into A^-fibres (0 - 20 ms), AÔ-fibres (20 - 90 ms),

C-fibres (90 - 300 ms). Action potentials arriving 300 - 800 ms after an electrical

pulse were classed as ‘post-discharge’, a result o f repeated stimulation leading to

wind-up neuronal hyperexcitability. The latch counter was set, via the delay width

module, to count cumulatively the C-fibre and post discharge activity evoked by

each o f the 16 stimuli (90 - 800 ms). These counts allowed calculation o f the 'input'

(non-potentiated response) and the 'excess spikes' (potentiated response, evident by

increased neuronal excitability to repeated stimulation).

Input = (action potentials (90 - 800 ms) evoked by first pulse) x (total number of

pulses (16)).

Excess spikes = (total action potentials (90 - 800 ms) after train of 16 stimuli) -

Input.

68

After establishment o f a cell’s response to electrical stimulation its response

profile to a range o f natural stimuli was determined. Any spontaneous activity

exhibited by a neurone in the absence of external stimulus to its receptive field was

recorded over a period of 10 minutes. The number o f action potentials evoked over a

10 second period to constant application of various natural stimuli was quantified in

response to both punctate mechanical (von Frey filaments 9 and 75 grams) and

thermal (constant water jet at 45 °C) stimuli applied to the centre o f the neurone’s

receptive field. The thermal response to 45 °C was determined by subtracting the

response to 32 °C (a non-noxious temperature so as to ascertain any mechanically-

evoked response from the water jet) from the response to 45 °C. All responses to

natural stimuli were normalized by the subtraction of any spontaneous activity

measured over a period of 10 seconds before the application of each stimulus.

2 .3 P h a r m a c o l o g ic a l S t u d ie s

2.3.1 D r u g a d m in is t r a t io n

The testing protocol, initiated every 10 minutes, consisted of an electrical

test followed by the natural stimuli, as described above. Stabilization o f the neuronal

responses was confirmed on consecutively obtaining at least three consistent pre­

drug responses (< 10% variation), for all measures. These values were then averaged

to generate pre-drug control values with which to compare the effect of drug

administration on subsequent evoked responses. Drugs were applied either spinally,

in a volume of 50 |Lil directly onto the exposed spinal cord, or sub-cutaneously in a

volume o f 250 p.1. After drug application the testing protocol was subsequently

continued for between 40 - 90 minutes per dose, to follow the effects until the

evoked neuronal responses peaked or plateaued, dependent on the time course of

action of the drug in use.

69

2.3.2 D r u g s E m p l o y e d

co-conotoxin GVIA was obtained from Sigma-Aldrich Company Ltd., Poole, Dorset,

U.K. The toxin was dissolved in 0.9% saline and stores in separate aliquots stored at

-20 °C.

co-agatoxin IVA was obtained from the Peptide Institute, Inc., Scientific Marketing

Associates, Bamet, Hertfordshire, U.K. The toxin was dissolved in 0.9% saline and

stores in separate aliquots stored at -20 °C.

Ethosuximide (2-Ethyl-2-methylsuccinimide) was obtained from Sigma-Aldrich

Company Ltd., Poole, Dorset, U.K. A stock solution was prepared in 0.9% saline and

stored at 4 °C.

Morphine sulphate was obtained from Evans Medical. A stock solution was

prepared in 0.9% saline and stored at 4 °C.

Gabapentin, a gift from Parke Davis, was dissolved in 0.9% saline and stored at 4

°C.

Naloxone hydrochloride was obtained from Sigma-Aldrich Company Ltd., Poole,

Dorset, U.K. A stock solution was prepared in 0.9% saline and stored at 4 °C.

Nifedipine was obtained from Sigma-Aldrich Company Ltd., Poole, Dorset, U.K.

Due to solubility difficulties in saline, this was dissolved in a mixture o f ethanol

(0.07%), cremophor (33.33%) and water (60%). The stock solution was then stored

protected from light at 4 °C.

70

2 .4 S t a t ist ic a l A n a l y sis

The effects of drugs on the electrically- and naturally-evoked responses

were calculated as maximum percentage changes from the averaged pre-drug value

for each neurone (% control) and the overall results for each dose were calculated as

means ± standard error o f mean (S.E.M.) of the normalized data. In most cases

graphs display the results expressed as mean% inhibition ± S.E.M. (% inhibition =

100% -% control, such that 100% inhibition represents no response, and 0%

inhibition represents control value). Statistical analysis o f maximal drug effects at

each dose compared to the averaged pre-drug value was determined by paired

Student’s Mest on raw data (number o f evoked action potentials). Comparison of

drug effects between different experimental groups (naïve, sham and SNL) were

made using an unpaired Student’s /-test on the normalized data. This comparison was

deemed valid if the evoked number o f action potentials determined for each neuronal

measurement by pre-drug characterization was found not to be statistically different

between groups using an unpaired Student’s /-test on raw data. The level of

significance was taken as P < 0.05. All statistical analysis o f data was performed

using the computer programme Statview 4.5 (Abacus Concepts). Additional analyses

used for particular experiments are described when necessary.

71

CHAPTER 3

BEHAVIOURAL CONSEQUENCES

AND DORSAL HORN NEURONE RESPONSE

CHARACTERISTICS FOLLOWING

SPINAL NERVE LIGATION

72

3.1 In tro du c tio n

The underlying pathophysiological mechanisms o f neuropathic pain, and

thus guidance towards improved therapeutic approaches, have been enlightened by

the introduction of animal models o f nerve injury. Early models of spinal transection

or dorsal rhizotomy resulted in complete denervation o f the hindpaw (Zeltser &

Seltzer, 1994). However, more recently models that produce partial denervation have

been developed more reminiscent of the clinical situation (see chapter 1.2). These

involve constriction of the common sciatic nerve with 4 loose ligatures in rats in the

CCI model (Bennett & Xie, 1988; Attal et a l, 1990), tight ligation o f 1/3 - 1/2 of the

sciatic nerve in rats in the PSL model (Seltzer et a l, 1990), and tight ligation o f one

or two o f the spinal nerves in rats in the SNL model (Kim & Chung, 1992) or

monkeys (Palecek et a l , 1992a). Common to each o f these neuropathies is the

behaviourally observed manifestation o f thermal hyperalgesia and mechanical

allodynia ipsilateral to nerve injury, which can be used to assess the therapeutic

benefit o f existing drugs as well as the analgesic potential of novel drugs. Likewise,

at a neuronal level, in vivo electrophysiological techniques can be used to investigate

pathological changes involved in neuropathic pain, also useful for pharmacological

assessment and identification of possible new drug targets.

It is widely acknowledged that nerve injury leads to alterations to the

peripheral nervous system that culminate in aberrant sensory processing and

neuropathic pain symptoms. Observations of many pre-clinical studies employing the

various models o f nerve injury have demonstrated that primary afferent fibres

properties are pivotal and at the origins o f characteristic changes in response to nerve

injury, including the generation of ectopic discharges and anatomical reorganization

(see chapter 1.3.1). Evidence is now also accumulating regarding nerve injury-

induced alterations at the level o f the spinal cord. In vivo electrophysiological studies

in rats with SNL (Pertovaara et ah, 1997; Chapman et a l , 1998b; Suzuki et a l ,

2000b), PSL (Behbehani & Dollberg-Stolik, 1994; Takaishi et a l , 1996), CCI

(Palecek et a l , 1992b; Sotgiu et a l , 1992; Laird & Bennett, 1993; Sotgiu et a l ,

1994a; Sotgiu et a l , 1994b), L4 - 6 dorsal root constriction (Tabo et a l , 1999) or

monkeys with L7 SNL (Palecek et a l, 1992a) have revealed a complex pattern of

changes in the properties o f dorsal horn neurones ipsilateral to nerve injury. In

73

general these include the development o f spontaneous activity and expansion of

receptive fields o f some dorsal horn neurones alongside absent receptive fields of

others, yet relatively small increases in response magnitudes or reductions in

thresholds to mechanical and thermal cutaneous stimulation.

In this chapter I describe the behavioural and dorsal horn neuronal

consequences induced by peripheral nerve injury, using the selective spinal nerve

ligation (SNL) rodent model. This involves tight ligation o f two (L5/6) of the three

(L4/5/6) spinal nerves that comprise the sciatic nerve (Kim & Chung, 1992). In

comparison to CCI and PSL, the magnitude of mechanical allodynia, assessed by

withdrawal frequency to innocuous mechanical stimuli (von Frey filaments), is

greatest and most reliable in the SNL model and experimental variability is

minimized to some extent due to the tight, versus partial, ligation. The SNL model

also preserves segregation of injured and uninjured primary afferent input proximal

to the formation o f the sciatic nerve such that spinal input from damaged afferents

can be selectively handled (see chapter 1.2). Subsequent to the establishment of

neuropathy in vivo electrophysiological recordings o f WDR dorsal horn neurones

were made and the evoked responses to a wide range of natural and electrical stimuli

assessed such that possible alterations in the spinal processing of cutaneous inputs

following SNL might be highlighted.

3 .2 M e t h o d s

3 .2 .1 B e h a v io u r a l A sse ssm e n t

Male Sprague-Dawley rats, initially weighing 130 - 150 g were subject to

either unilateral selective spinal nerve ligation (L5/6) or a sham operation (control

group) as first described by Kim and Chung (1992). For a postoperative (PO) period

of two weeks behavioural testing o f both experimental groups was carried out on

days 2, 3, 5, 7, 9, 12 and 14, in order to determine the sensitivity o f both hindpaws to

punctate mechanical (von Frey filaments) and cooling (acetone) stimuli.

74

3.2.2 D e t e r m in a t io n o f D o r s a l H o r n N e u r o n e R e s p o n s e t o E l e c t r ic a l

S t im u l i

Subsequent to SNL or sham operation and behavioural assessment,

electrophysiological recordings of ipsilateral convergent dorsal horn neurones were

made at PO days 14 - 17. Additionally, neuronal responses were recorded from un­

operated naïve rats. On isolation of a single cell and insertion of stimulating needles

into the receptive field, a rest period o f 10 minutes was allowed before

characterization of the neurones response to electrical stimulation. The threshold

current required to elicit a C-fibre mediated response was determined. An electrical

test o f 16 stimuli at 3 x C-fibre threshold was then performed and a post stimulus

histogram (PSTH) constructed. Quantification o f the A|3- AÔ- and C-fibre mediated

action potentials was performed on a latency basis, as was the post-discharge. Input

and excess spikes measures were also calculated.

3.2.3 D e t e r m i n a t i o n o f D o r s a l H o r n N e u r o n e R e s p o n s e s t o N a t u r a l

S t im u l i

Subsequent to the electrical stimulus a rest period of several minutes was

allowed. Any spontaneous activity was then recorded over a period of 10 minutes.

Following this, the neuronal response to innocuous and noxious punctate mechanical

stimuli (von Frey 9 and 75 g, respectively) and noxious heat (45 °C) was determined

upon application to the most responsive part of the receptive field for 10 seconds. A

recovery interval was allowed between each test, and responses were normalized by

subtraction of any spontaneous activity apparent before application of each stimulus.

3.2.4 D e t e r m in a t io n o f D o r s a l H o r n N e u r o n e R e c e p t iv e F i e l d S iz e

The area o f the neuronal receptive field over the ipsilateral hindpaw was

determined by probing the plantar surface with three von Frey filaments (9, 15 and

75 g). An evoked response greater than 1 Hz was the criteria for inclusion in the

neurone’s receptive field, which was marked on a standard diagram of the hindpaw.

For quantification of the mean receptive field size o f neurones recorded from each

experimental group, the diagrams were photocopied onto plain paper (80 g/m^) and

the shaded areas cut out and weighed and expressed as a percentage of the weight of

75

a control diagram (determined as an average of 20 as 77.8 ± 0.2 mg). Localization of

the receptive field was also classified into lateral (comprising toes 1 and 2 and the

lateral side of the plantar surface of the hindpaw), central (comprising toes 3 and 4

and the central plantar surface of the hindpaw) and medial (comprising toe 5 and the

medial side of the plantar surface of the hindpaw), as determined by von Frey 9 g.

3.3 R e s u l t s

3 .2 .1 B e h a v io u r a l A s se s s m e n t

During the postoperative period the animals showed normal weight gain and

maintained good general health. Rats subjected to SNL exhibited abnormal foot

posture ipsilateral to nerve injury whereby toes were held together in a 'guarding'

behaviour. This did not occur in either the contralateral hindpaw, or in the sham-

operated rats. Furthermore, in the absence of externally applied stimuli, SNL rats

often licked or pulled the ipsilateral hindpaw and body weight was shifted more to

the contralateral hindpaw. Autotomy, signs o f distress and aggression were not

observed.

Successful replication o f the nerve injury model was confirmed by

behavioural testing o f 138 SNL rats and 81 sham-operated rats, which demonstrated

the development of mechanical and cooling allodynia o f the injured hindpaw of SNL

rats (Figure 5). Evoked allodynia, in response to innocuous mechanical (von Frey

filaments bending force 1 - 9 g) and cooling (acetone) stimuli, was displayed as a

brisk withdrawal, accompanied in some cases by shaking and licking o f the foot

ipsilateral to SNL. This was evident at PO day 2, reached maximum PO days 7 - 1 2

and still maintained at PO day 14. Consistent withdrawal responses were never

exhibited by the control group or by the contralateral hindpaw of the experimental

group, and when present were never accompanied by the pain-like behaviours

displayed to stimuli applied to the lesioned hindpaw of SNL rats.

76

von Frev 5gvon Frev Ig

7 -

Shams

l:;SNL 4 -

Q

0 -

I I I I I I I I0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 16

von Frev 9g8

7

6

§ 5

j:5 2

1

0

10 2 4 6 8 10 12 14 16

Acetone3

2

1

0

10 2 4 6 8 10 12 14 16

Postoperative Day Postoperative Day

Figure 5. Development o f mechanical and cooling allodynia in the ipsilateral hindpaw

over the 2 week postoperative periodfollowing SNL (9, n = 138) in comparison to sham

control (O, n = 81). Data is expressed as the mean difference score ± S.E.M. for the

hindpaw withdrawal response to punctate mechanical stimuli (von Frey filaments

bending forces (a) 1 g, (b) 5 g and (c) 9 g and (d) cooling stimulus (drop o f acetone)

applied to the plantar surface o f the hindpaw s (trials o f 10 for the mechanical and 5 for

the cooling). Difference score = (ipsilateral hindpaw response frequency) -

(contralateral hindpaw response frequency).

11

3.2.2 R e s p o n s e C h a r a c t e r i s t i c s o f D o r s a l H o r n N e u r o n e s t o E l e c t r i c a l

S t im u l i

A total o f 233 convergent dorsal horn neurones were characterized, of

which 75 were from naïve animals, 54 were from sham operated rats and 104 were

from rats subject to SNL (Table 4). All neurones recorded from operated groups had

a receptive field over the left ipsilateral hindpaw. No difference between

experimental groups was observed in the mean cell depth, which was for individual

neurones between 500 and 1000 pm corresponding to laminae V o f the spinal cord.

Table 4. Comparison o f the mean electrical threshold required to evoke a C-fibre

mediated response and the mean evoked dorsal horn neuronal action potentials

(APs) in response to a train o f 16 electrical stimuli at 3 x C-fibre threshold in naïve,

sham-operated and SNL rats at post-operative days 14-17. Electrically-evoked APs

are divided into groups based upon their mediating fibre type. Data is expressed as

means ± S.E.M. and ^denotes statistically significant in comparison to naïve.

NAIVE SHAM SNL

Neurone depth (pm) 815 ± 25 879 ± 27 ; 801 ± B^

C-fibre threshold (mA) 1.48 ± 0 .27 1.18±0.1 1.19 ± 0 .07

C-fibre mediated response (AP) 355 ± 17 351 ±21 :y?^345 ±:T5i^

A(3-fibre mediated response (AP) 107 ± 4 109 ± 5 112 ± 3' f

A5-fibre mediated response (AP) 6 2 ± 4 69 ± 5

Post-discharge (AP) 194 ± 15 2 0 4 ± 16.A 4% f 2 f l ± 1 5 -Î,,'d- •

Input (AP) 307 ± 19 282 ± 19 s “¥ 2 9 0 # 1 6 ’*4

Excess spikes (AP) 297 ± 27 % 309 ± 27

The current required to evoke a C-fibre mediated response was similar

across all three groups. Examples of a representative response o f a neurone to a

single electrical pulse at 3 x C-fibre threshold (as seen on the oscilloscope) and a

PSTH following a train o f 16 electrical stimuli is shown in Figure 6.

78

a

4*

800 m s

Postdischarge

400

Time (ms)

Figure 6. (a) Representation o f the evoked dorsal horn neuronal response to a single

electrical stimulus and 3 times C-fibre threshold, as seen on the oscilloscope, (b) An

example o f a typical PSTH produced after a train o f 16 electrical stimuli at a frequency

o f 0.5 Hz at 3 times C-fibre threshold. The action potentials evoked in a single dorsal

horn neurone by the different primary afferent fibre types are displayed, and the

postdischarge.

79

After a train o f 16 electrical stimuli, generally the total evoked neuronal

response mediated by Ap-, Aô- and C-fibres were comparable in SNL, sham and

naïve rats (Table 4), however the A§-fibre response in SNL rats was found to be

significantly larger than that o f naïve rats {P < 0.05), but not in comparison to sham.

The post-discharge, input and excess spikes measures were not significantly different

between groups.

3.2.3 R e s p o n se C h a r .4c t e r is t ic s o f D o r sa l H orn N e u r o n e s to N a t u r a l

S tim u li

The level o f ongoing spontaneous activity was found to be significantly

higher {P < 0.0001, Mann-Whitney) in neurones in SNL rats compared to sham and

naïve (Table 5). Furthermore 64% of neurones characterized in SNL rats exhibited

spontaneous activity at a rate greater than 0.1 Hz in comparison to only 31% of

characterized neurones in sham-operated rats and 27% o f neurones in naïve animals

( f < 0.001, Fishers exact test).

Table 5. Comparison o f the mean dorsal horn neurone depth recorded from, the

mean spontaneous activity (i.e., neuronal activity in the absence o f external stimulus)

and the mean evoked dorsal horn neuronal action potentials (APs) in response to

innocuous (9 g) and noxious (75 g) von Frey filaments and noxious heat in naïve,

sham-operated and SNL rats at post-operative days 14 - 17. Data is expressed as

means ± S.E.M. and indicates statistical significance (P < 0.0001) in comparison

to naïve and sham groups.

NAIVE SHAM^

Spontaneous activity (Hz) 0.36 ± 0.13 0,33 ±0.11: t

*3.51 ± 0 .7 2 ,

von Frey 9 g evoked response (Hz) 14.18 ± 1 .89 A 13.2 ±1.86 . 14.58 ± 1.453W'' ' '

von Frey 75 g evoked response (Hz) 67.4 ±4.11 66.23 ±4.92 65.73 ±3.41,

45 °C heat evoked response (Hz) 38.52 ±3.88 37.55 ±4.04 v3L14 ±2.57

80

A representative response o f a neurone to punctate mechanical stimuli (von

Frey 9 and 75 g) and noxious heat (constant water je t at 45 °C) is shown in Figure 7.

Neurones displayed an increased response to noxious von Frey 75 g in comparison to

non-noxious von Frey 9 g. The heat response was normalized by subtraction of

evoked activity due to mechanical stimulation o f the water je t at 32 °C, when present.

Naturally-evoked neuronal responses were comparable across SNL, sham and naïve

groups (Table 5).

45 °C

3 0 0 vF 75 g

c(Uo 200Oncoo< v F 9 g

1 0 0 -

32 °C

0 50 100 150

Time (s)

Figure 7. A typical rate recording showing the evoked dorsal horn neuronal action

potentials to natural stimuli applied for a period o f 10 seconds each. The mechanically

evoked response to innocuous von Frey 9 g and noxious von Frey 75 g are displayed.

The response to noxious heat over 10 seconds (water je t at 45 °C) was normalized by

subtraction o f the response to water je t at 32 °C over 10 seconds so as to remove any

mechanically evoked activity due to the water je t itself.

81

3.2 .4 R e c e pt iv e F ield C h a r a c t e r ist ic s of D o r sa l H o r n N e u r o n e s

A total of 72 neurones where used for determination o f receptive field

distributions, o f which 16 were from naïve rats, 26 were from sham-operated rats and

30 were from SNL rats. O f the neurones recorded from, the relative proportions that

had medially, centrally or laterally distributed receptive fields tended to be mainly

centrally located within each o f the three groups (Table 6). Between experimental

groups there was a tendency for the receptive fields o f neurones recorded from SNL

rats to be centrally/medially distributed, in comparison to naïve and sham-operated

rats which tended to be centrally/ laterally distributed.

Table 6. The distribution o f dorsal horn neurone receptive fields over the hindpaw

ipsilateral to SNL or sham-operation at PO day 1 4 - 1 7 and also in un-operated

naive rats. Medial, central and lateral assignment was determined by probing with

von Frey 9 g. Data is expressed as percentage o f neurones characterized within each

group.

NAIVE SHAM

Medial 19% 19%

Central 56% 50%

Lateral 25% 31% Î. , 13%

The receptive field size of dorsal horn neurones, mapped with von Frey 9,

15 and 75 g, were generally comparable between SNL, sham-operated and naïve

animals, two weeks after surgery where performed, and increased in size with

increasing intensity (Figure 8). However, the receptive field size mapped to

innocuous von Frey 9 g was significantly greater (by 40%) for neurones recorded

from SNL rats in comparison to either sham or naïve rats (P < 0.05). Receptive field

sizes mapped to greater von Frey intensities o f 15 g and 75 g were comparable across

the three groups. An example of the receptive fields o f a neurone from each

experimental group is shown in Figure 9.

82

5 0 -

40 H5

^ 306

Æ 20o

0 J

TJ

n Naive

Hi Sham

■ SNL

Ig 5g 9g

von Frey Filament

Figure 8. The mean cutaneous receptive fie ld sizes o f WDR dorsal horn neurones

recordedfrom naïve (n = 16), sham-operated (n = 26) and SNL (n = 30) rats, ipsilateral

to surgery where performed Data are expressed as mean % o f total area o f the plantar

surface o f the hindpaw’ ± S.E.M. ^ denotes statistically significant difference in

comparison to sham and naive (P < 0.05).

83

vF 9 g vF 15 g vF 75 g

Naive

Sham 12%

SNL 2 1 %

17%

27%

35%

42%

41%

84%

Figure 9. Examples o f typical receptive fields to o f dorsal horn neurones recorded from

naive, sham-operated and SNL rats, ipsilateral to surgery where performed Receptive

fields were mapped to von Frey filaments 9, 15 and 75 g and the area as a percentage o f

the whole hindpaw> area is given. The numbers assigned to the toes œ^e shown in the top

left-hand diagiam and the receptive fields displayed are classed as lateral in naive,

central in sham and medial in SNL.

84

3.4 D is c u s s io n

3.4.1 B e h a v io u r a l C o n s e q u e n c e s o f SNL

Here, I have demonstrated successful replication of the nerve injury model

of SNL as first described by Kim and Chung (1992). Over the postoperative period

of two weeks the rats subject to L5 and L6 spinal nerve tight ligation displayed

behaviours indicative o f neuropathic pain ipsilateral to nerve injury. These

nociceptive behaviours were not manifest contralateral to surgery nor in the sham-

operated control animals. Brisk withdrawal of the injured hindpaw to stimulation

with normally innocuous punctate mechanical stimuli (von Frey filaments) and

cooling stimuli (acetone) indicated the development o f mechanical and cooling

allodynia, consistent with the original findings (Kim & Chung, 1992). Nociceptive

behaviours were often apparent on the first testing day at PO day 2, and the level of

allodynias increased and were maintained over the two week behavioural assessment

period. Mechanical allodynia has been reported as early as 12 hours after nerve

injury and maintained for up to 15 weeks (Kim & Chung, 1992). Delayed

development o f mechanical sensitivity was additionally been reported in the hindpaw

contralateral to nerve injury, prominent at 2 weeks and maintained for 14 weeks. In

the original paper a sham operation was performed on the contralateral side, however

the lack o f sham-induced behavioural changes seen in a separate sham control group

(Kim & Chung, 1992) and here most likely eliminates this as the cause. I observed

no consistent withdrawal responses contralateral to nerve injury, and on the few

occasions this did occur it was never associated with nociceptive behaviours such as

licking or shaking o f the foot, often observed ipsilateral to nerve injury. However, I

only performed behavioural assessment for 2 weeks after surgery, whereas these

contralateral changes observed previously were delayed in onset. Additionally, these

contralateral changes were observed in response to von Frey 19 g, which is higher

than the greatest von Frey employed in my study. Other models o f peripheral nerve

injury have also reported delayed onset contralateral changes (Takaishi et a l, 1996;

Tabo et ah, 1999), and this may be representative of ‘mirror image’ pain observed

contralateral to nerve injury clinically (Bonica, 1979).

85

Thermal hyperalgesia was also recorded in the original study o f SNL, being

apparent 3 days after surgery and maintained for at least 5 weeks (Kim & Chung,

1992). This was not assessed in my study, however others have failed to observe the

development o f increased sensitivity to suprathreshold thermal stimulus after SNL

(Kontinen et aL, 1998; Roytta et aL, 1999) and other neuropathic pain models

(Luukko et aL, 1994). It appears that there may be some differences in the

development o f nociceptive behaviours dependent upon the evoking stimulus

modality, which is likely mediated by differing transduction mechanisms displaying

differential alterations in response to nerve injury.

In addition to the evoked allodynia observed in response to external stimuli,

rats subject to SNL also displayed certain ongoing nociceptive behaviours. Most

notable was the adopted ‘guarding’ posture of the injured hindpaw and associated

shift o f weight-bearing to the contralateral side observed in SNL and other nerve

injury models. Guarding posture has been shown to be decreased by morphine and

tricyclic antidepressants as expected for a pain-related behaviour (Jazat & Guilbaud,

1991; Ardid & Guilbaud, 1992). This postural characteristic has been suggested to be

an attempt to avoid potentially painful mechanical stimulation of the paw (Bennett &

Xie, 1988) and mechanical stimulation of the injury site itself (Laird & Bennett,

1993). Following CCI it has been shown that the rapid development of scar tissue

around the ligatures tethers the injury site to adjacent muscle (Bennett & Xie, 1988)

resulting in its visible movement upon of leg extension (Laird & Bennett, 1993).

This then excited dorsal horn neurones responding to direct mechanical stimulation

of the nerve injury site and was not due to joint movement. Therefore leg extension

encountered during walking or standing was postulated a likely adequate stimulus for

the mechanosensitive afferents in the neuroma (Laird & Bennett, 1993). They

additionally demonstrated that the majority of these responsive dorsal horn neurones

had a C-fibre input, concluding that some of the afferent input evoked by neuroma

stimulation is mediated by nociceptive afferents.

Despite evidence for ongoing pain after SNL, it has been reported that these

animals show no differences in levels of anxiety or depression in comparison to

sham-operated animals (Kontinen et aL, 1999), therefore this would have no impact

86

upon evoked behaviours. As regards the withdrawal reflex as a valid indicator of

mechanical allodynia, it has been suggested that the shift of weight to the

contralateral side might influence this assessment. However it has been demonstrated

that decreased mechanical thresholds ipsilateral to nerve injury are independent of

this factor, and thus validating quantification o f mechanical allodynia by evoked

hindpaw withdrawal (Kauppila et aL, 1998a).

In comparison to other models of partial denervation of the hindpaw, the

selective tight ligation o f SNL results in the most reproducible and extensive

mechanical allodynia, and minimizes, to some extent, the inter-experimental

variability that is encountered in CCI and PSL models. CCI involves constriction of

the common sciatic nerve with 4 loose ligatures (Bennett & Xie, 1988), whereas PSL

involves tight ligation of 1/3 - 1/2 of the thickness of the sciatic nerve (Seltzer et ah,

1990). Depending on the extent of constriction exerted by the ‘loose’ ligatures or

proportion of the sciatic nerve ligated, respectively, varying proportions of fibre

types within the sciatic nerve innervating the hindpaw could be injured (Basbaum et

aL, 1991; Carlton et aL, 1991). In this respect, for SNL rats the only source of

variability in the contribution that each spinal segment makes to the sciatic nerve can

arise from anatomical differences between individual rats. Furthermore the SNL

model also preserves segregation o f injured and uninjured primary afferent input

proximal to the formation of the sciatic nerve such that spinal input from damaged

afferents can be selectively handled (see chapter 1.2). In CCI and PSL models partial

denervation damage to the common sciatic nerve results in all three spinal segments

and their DRG containing a combination o f damage and undamaged fibres. Self-

mutilating behaviour of the denervated regions o f the hindpaw, thought to be an

attempt to rid itself o f the painful experience (Coderre et aL, 1986), has been

observed in CCI rats (Bennett & Xie, 1988; Attal et aL, 1990) but not after SNL

(Kim & Chung, 1992) or PSL (Seltzer et aL, 1990). I did not observe such autotomy

in my study.

Peripherally reported consequences o f SNL and other models of nerve

injury include the development of ongoing ectopic activity, generated at the site of

injury or within the DRG, may be responsible for alterations to behaviour and

87

somatosensory processing associated with neuropathic pain (Wall & Devor, 1983;

Burchiel, 1984; Kajander et aL, 1992; Steel et aL, 1994; Xie et aL, 1995; Study &

Krai, 1996). In the clinic, positive correlations have been made between the

occurrence of spontaneous firing o f nociceptors innervating the painful region and

neuropathic pain (Nystrom & Hagbarth, 1981; Gracely et aL, 1992). Primary afferent

fibres have also been reported to undergo structural reorganization and exhibit

neurochemical/receptor plasticity, such that large myelinated A^-fibres might

acquire the ability to mediate nociceptive transmission, in particular allodynia (see

chapter 1.3 for discussion). Paradoxically, SNL has been shown to decrease the

number of myelinated primary afferent fibres (Roytta et aL, 1999), which may imply

facilitated spinal nociceptive processing of afferent input. This is substantiated by the

fact that spinal application of different drugs can reduce mechanical allodynia that

develops following SNL (Backonja et aL, 1994; Bowersox et aL, 1996; Chaplan et

aL, 1997; Hwang & Yaksh, 1997) and thermal hyperalgesia (Wegert et aL, 1997).

3.4.2 D o r s a l H o r n N e u r o n a l C o n s e q u e n c e s o f S N L

Subsequent to establishment of neuropathy the operated rats were then used

for electrophysiological studies, alongside naïve, un-operated animals. Making in

vivo recordings o f ipsilateral WDR neurones located in lamina V of the dorsal horn I

was able to characterize the responses o f these neurones driven by electrical and

natural stimulation of their cutaneous receptive fields in order to assess any changes

occurring parallel to the observed behavioural consequences. Before explaining my

observations I will briefly explain the conclusions that can be drawn from the

different electrical measurements made throughout this thesis.

The majority of evoked WDR neurone responses recorded are ultimately

under the peripheral drive o f primary afferent fibre activity. Action potentials

arriving in the spinal cord carried by the different fibre types evokes WDR neurone

activity either by synapsing directly (A|3-fibres) or via a relay o f intemeurones (C-

fibres), such that non-nociceptive and nociceptive information converges (see

chapter 1.3.2.2). The AP-, A5- and C-fibre measurements, reveal the evoked activity

of the WDR neurone mediated by activity in each o f the different fibre types. A^-

88

fibres carry touch evoked stimuli, C-fibres convey nociceptive information and Aô-

fibres carry possibly both. The 'input' is reflective o f presynaptic activity and

neurotransmitter release, and o f the baseline resting postsynaptic C-fibre-evoked

dorsal horn neurone response, quantifying activity in response to the first stimulus of

the electrical train. It represents the excitatory effect of a single afferent barrage on

spinal neuronal circuitry, prior to which the system is at rest and intemeuronal

amplification/hyperexcitability mechanisms, involving NMDA receptor recruitment,

are not in play. 'Postdischarge' quantifies action potentials that cannot have been

evoked by even the slowest-conducting afferent fibre inputs, and is therefore

reflective of spinal cord hyperexcitability generally only apparent after repeated

stimulation. This activity-dependent hyperexcitability can also be quantified as

‘excess spikes’, a measure o f the additional action potentials recorded above the

predicted constant baseline response. Excess spikes and postdischarge are therefore

more reflective o f postsynaptic activity and wind-up, where intemeuronal

amplification/hyperexcitability mechanisms, involving NMDA receptor recruitment,

are now in play.

I found the overall magnitude o f responses to innocuous mechanical,

noxious mechanical and thermal, and supra-C-fibre threshold electrical stimuli were

comparable between SNL, sham-operated and naïve animals, which is surprising

considering the behavioural difference observed between SNL and sham-operated

groups. It could be argued that the lack o f obvious correlation between the

behavioural changes and response properties of dorsal horn neurones is that neuronal

types (nociceptive specific, non-nociceptive, STT neurones), other than WDR, might

have exhibited changes more consistent with behaviour. However, I believe that

WDR neurones are a reasonable population to study as they are a rational target for

some o f the changes underlying hyperalgesia and allodynia for a number o f reasons.

Activity in WDR neurones has been shown to be sufficient for pain perception

(Mayer et aL, 1975; Price & Mayer, 1975) and they comprise more than half o f the

neurones o f the STT (Besson & Chaouch, 1987), a major ascending pain signalling

pathway. Due to the convergence of low- and high-threshold inputs onto WDR

neurones, they can signal sensory-discriminative aspects o f pain (Maixner et aL,

1986) and have been shown to be more accurate predictors of thermal perception

(Dubner et aL, 1989) in comparison to nociceptive-specific neurones.

89

The most obvious difference in neurones recorded from SNL rats here, was

the level of ongoing, spontaneous activity exhibited by neurones in the absence of

application o f an external stimulus to their receptive fields. This was significantly

greater in WDR neurones ipsilateral to nerve injury in comparison to neurones

recorded from sham-operated and naïve animals, in both terms of frequency and the

proportion o f neurones displaying such activity. This is in agreement with

observations made in other electrophysiological dorsal horn neuronal studies after

SNL (Pertovaara et aL, 1997; Chapman et aL, 1998b). The incidence of spontaneous

activity of ipsilateral dorsal horn neurones has also been reported in other rat models

of nerve injury such as PSL (Behbehani & Dollberg-Stolik, 1994), CCI (Palecek et

aL, 1992b; Sotgiu et aL, 1992; Laird & Bennett, 1993; Sotgiu et aL, 1994b) and L4 -

6 dorsal root constriction (Tabo et aL, 1999) and in monkeys with L7 SNL (Palecek

et aL, 1992a). Furthermore, its development has been shown to increase temporally

(Chapman et aL, 1998b; Tabo et aL, 1999), in parallel with the time course of evoked

behavioural allodynia (Chapman et aL, 1998b). The spontaneous activity described

here at the level o f the spinal cord may be secondary to the previously described

nerve injury-induced generation o f ectopic activity in the periphery (Wall & Devor,

1983; Burchiel, 1984; Kajander et aL, 1992; Steel et aL, 1994; Xie et aL, 1995;

Study & Krai, 1996). In SNL animals it has been demonstrated that signals from

injured primary afferent fibres into the spinal cord are critical for the development

and maintenance of neuropathic behaviour (Sheen & Chung, 1993; Yoon et aL,

1996). Although I did not asses the level o f ongoing spontaneous pain behaviour

after SNL, this has been previously described (Kim et aL, 1997) and may relate to

my observation of increased dorsal horn neuronal spontaneous activity following

SNL.

The other significant difference I observed was an increase in the cutaneous

receptive field to innocuous mechanical stimulation (von Frey 9 g) o f WDR dorsal

horn neurones ipsilateral to SNL, in comparison to those of sham-operated and naïve

animals. Other electrophysiology studies have also described enlarged receptive

fields after nerve injury. Following PSL the receptive field areas o f neurones ipsi-

and contralateral to nerve injury 3 - 5 weeks later were significantly larger compared

to those o f neurones recorded in un-operated rats (Behbehani & Dollberg-Stolik,

1994), yet in another study this was not observed until 16 weeks. After L4 - 6 dorsal

90

root constriction WDR dorsal horn neurones were shown to exhibit enlarged

mechanical receptive fields at a later time point of 22 weeks (Tabo et aL, 1999).

Despite the presence o f enlarged receptive fields for a number o f dorsal horn

neurones several of these electrophysiological studies after nerve injury also report

that a fraction o f neurones on the nerve injured side display no mechanical receptive

fields (Palecek et aL, 1992b; Laird & Bennett, 1993) which is indicative of partial

deafferentation.

Reorganization o f receptive fields is indicative of neuronal plasticity after

peripheral nerve damage and is described both clinically and pre-clinically. These

observations described electrophysiologically after experimental neuropathy fit well

with the clinical observations. In patients a cardinal feature o f neuropathic pain is a

partial or complete loss of afferent sensory function and the paradoxical presence of

certain hyperphenomena in the painful area (Lindblom, 1994). Neuropathic pain

patients also have large areas of abnormal sensations and pain, not corresponding to

any known nerve or root innervation territory or dermatomes (Jensen & Rasmussen,

1994). Studies in humans and animals (see Dubner & Basbaum, 1994, for review)

have shown radiation aftersensations and expansion of receptive fields may be

related to pathological changes in dorsal horn neurones. Psychophysical and

neuronal recordings have shown that pain intensity is related to impulse discharges

and numbers o f neurones activated (Price et aL, 1994). WDR neurones are in part

characterized by small receptive zones that can be excited by non-noxious stimuli

(touch, gentle pressure) surrounded by a much larger zone from which noxious

stimuli (pinch, firm pressure, temperature > 45 °C) can evoke neuronal discharges.

These large receptive field zones are overlapping and extend over several

dermatomes and their receptive fields are a reflection o f synaptic propriospinal

interconnections in the spinal dorsal horn that extend over several segments.

Therefore a noxious stimulus can activate several WDR neurones and increasing the

stimulus intensity will result in activation of further WDR neurones rostrocaudally

dispersed. A global increase in spinal cord excitability resulting from either an

increase in synaptic efficacy of excitatory inputs or a decrease o f inhibitory controls

(or both) would have the potential to alter neuronal cutaneous receptive fields.

Indeed, blockade o f inhibitory neurotransmission by picrotoxin or strychnine can

mediate expansion of receptive fields of spinal or meduallry dorsal horn neurones

91

(Yokota & Nishikawa, 1979; Markus & Pomeranz, 1987). Furthermore PAG

stimulation was shown not to attenuate heat-evoked responses o f spinal dorsal horn

WDR neurones in the ligated side in comparison to normal which indicated that SNL

induced a change in the descending inhibitory regulation originating in the PAG

(Pertovaara et aL, 1997). If receptive field size is increased then for any given

stimulus a greater number of spinal neurones will be recruited, actually leading to

greater afferent input and possibly enhancing pain transmission. Thus, alongside

nerve injury-induced dorsal horn neuronal spontaneous activity these observations

are indicative o f such spinal cord hyperexcitability.

Despite the behaviourally observed mechanical and cooling allodynia

consequences o f SNL just prior to the electrophysiological recordings, I saw no

overall increases in the electrically- and naturally-evoked dorsal horn neuronal

response magnitudes in comparison to sham and naïve animals, as might be expected

in the presence o f increased pain responsiveness. Such lack of increases o f dorsal

horn neuronal responses has also been described in other studies. Recordings of

WDR neurones made ipsilateral to SNL have demonstrated subtle intensity-

dependent changes in the response o f dorsal horn neurones to mechanical stimulus,

compared to sham (Chapman et aL, 1998b). The absolute magnitude of response to

brush, prod and acetone were reduced two weeks after neuropathy, the evoked

responses to innocuous punctate mechanical stimuli were comparable between

experimental groups, yet those to a noxious were reduced. Electrically- and noxious

heat-evoked dorsal horn neuronal responses were also reduced in SNL compared to

sham-operated rats. However, the proportion of neurones responding to different

innocuous modalities of stimuli were increased after SNL and not in sham. No

injury-induced modifications to stimulus thresholds were noted even though

behavioural studies have shown that the threshold of the ipsilateral paw is reduced in

response to mechanical stimuli and as is the withdrawal latency to thermal stimuli

following SNL (Kim & Chung, 1992; Wegert et aL, 1997). After L4 - 6 dorsal root

constriction no difference in response properties o f dorsal horn neurones were

observed after 5 weeks, even though behavioural mechanical hypersensitivity was

maximal (Tabo et aL, 1999). Although the mean threshold for eliciting a

mechanically-evoked dorsal horn WDR neurone response was not altered by CCI

(Laird & Bennett, 1993), the number o f neurones sensitive to low intensity

92

mechanical stimulation (Laird & Bennett, 1993) and the magnitude of mechanically-

evoked responses of the neurones were also reduced (Palecek et ah, 1992b). Dorsal

horn neurones ipsilateral to nerve injury have been shown to respond normally to

heat exhibiting no significant changes in response threshold or magnitude of

responses after SNL (Pertovaara et aL, 1997), CCI (Palecek et aL, 1992b; Laird &

Bennett, 1993) and PSL (Takaishi et aL, 1996), even in the presence o f behaviourally

observed thermal hyperalgesia. Furthermore, the heat-evoked responses to

suprathreshold stimuli were actually lower in neuropathic animals (Palecek et aL,

1992b).

In contrast, nerve injury-induced increases in evoked neuronal responses

have been reported. Following L5/6 SNL in rats, WDR responses to mechanical

stimuli were enhanced independent of stimulus intensity (Pertovaara et aL, 1997). In

monkeys previously receiving L7 SNL, ipsilateral spinothalamic tract neurones

rostral to the ligated segment exhibited reduced mechanical and thermal thresholds

and increased magnitude o f response in comparison to the contralateral sham-

operated side. However, neurones within the ligated segment had low responsiveness

(Palecek et aL, 1992a).

In the present study, the magnitude o f evoked dorsal horn neuronal

responses show remarkable comparability between the groups, in agreement with the

majority o f other electrophysiological studies that report no increase in responses

after nerve injury. It has been demonstrated that damage to the peripheral nervous

system actually results in substantial primary afferent fibre loss (Castro-Lopes et aL,

1990), and therefore stimulus-evoked input into the spinal cord is conveyed along

fewer fibres. Consistent with this, a large proportion o f neurones ipsilateral to CCI

possess no cutaneous receptive fields (Laird & Bennett, 1993). In the SNL model,

ligation o f L5 and L6 spinal nerves almost completely abolishes input into the

corresponding spinal segments. However, both of these areas send input to the

undamaged L4 segment, and based upon the estimated contribution of adjacent

unmyelinated nerves to a segment (Besse et aL, 1991), in L4 there could be a deficit

of about 35% in terms o f afferent input. If there are fewer dorsal horn neurones on

the injured side with intact sensory input, and those still receiving input respond

93

normally to noxious stimuli, then one would predict a reduction, rather than the

observed increase in sensitivity of the hindlimb observed behaviourally and the

relatively minor alterations in the evoked responses o f dorsal horn neurones.

Clinically it is clear that allodynia is mediated by low-threshold myelinated AP-

fibres and in order for these normally innocuous inputs to be perceived as painful,

abnormal central processing o f these inputs would have to occur (see Torebjork et

aL, 1995), yet no difference between experimental groups was observed in the

magnitude of AP-fibres nor C-fibre and hyperexcitability measures of excess spikes

and postdischarge. In a similar fashion no nerve injury-induced change in the

magnitude of neuronal response to innocuous and noxious natural stimuli was

observed.

In consolidation with the other studies already discussed, overall, nerve

injury induces behaviourally observed enhanced pain responses alongside complex

changes in dorsal horn neuronal response pattern. Whilst there appears to be a

general increase in the level o f neuronal spontaneous activity after nerve injury,

changes in response to externally applied stimuli vary slightly between studies,

however the predominant tendency appears to be towards little change, or perhaps

even a reduction. The concomitant presence of enhancements and reductions of

neuronal activity and behaviours are reminiscent o f the clinical descriptions of

neuropathic pain where sensory deficits and enhanced sensations can coexist. On

further consideration, the lack of increased neuronal excitability may not be as

unexpected as first thought. These experimental observations, together with the

consideration of symptoms of neuropathic pain, suggest that evoked pains may arise

from central compensations for the loss of normal sensory inputs, whereby spinal

processing of activity from the remaining intact primary afferents must actually be

increased in order to maintain the same magnitude o f response evoked in control

animals. Ongoing peripheral activity may well produce an ongoing level of

neurotransmitter release in the spinal cord, as indicated by high levels o f spontaneous

activity, which may in turn favour hyperexcitability o f responses to subsequent

evoked stimuli. Thus, in neuropathic pain it is likely that aberrant peripheral activity

is amplified and enhanced by spinal mechanisms. This possibly results in increased

VDCC activity leading to more transmitter release and thus increased neuronal

activity mediated by NMDA- and other receptors. The altered stimulus-response

94

relationship may be a basis for the allodynia and thermal hyperalgesia reported in

this model.

95

CHAPTER 4

VOLTAGE-DEPENDENT CALCIUM CHANNELS:

THE CAv2 FAMILY

96

4.1 In tro du c tio n

Neuropathic pain can arise from injury- or disease-evoked damage to the

peripheral or central nervous system and patients often experience sensory deficits,

ongoing pain and stimulus-evoked pain (allodynia and hyperalgesia). Animal models

of neuropathy have been critical in elucidating the complex causal mechanisms,

involving plasticity in nociceptive transmission and modulating systems. The SNL

model (Kim & Chung, 1992) involves tight ligation of two (L5 and L6) of the three

spinal nerves that form the sciatic nerve. Behavioural consequences include thermal

hyperalgesia, and mechanical and cooling allodynia (Kim & Chung, 1992; Chaplan

et aL, 1997).

Sensory neurones express a number of classes of VDCCs (L, N, P/Q, R and

T), distinguished by their electrophysiological and pharmacological profiles. Ten

genes encoding the main pore-forming a i subunit have been identified, termed ociA

to an, and a recently devised nomenclature groups these into three families, Cayl, 2

and 3, based upon structural and functional characteristics (Ertel et aL, 2000). Ca l

and Cay2 comprise the high voltage-activated (HVA) VDCCs, which allow Ca

influx upon substantial membrane depolarization, whereas Cay3 channels are the low

voltage-activated (LVA) VDCCs, such that they permit Ca^ flux at resting

membrane potentials.

N-type VDCCs (am or C a y 2 .2 current) are sensitive to block by co-

conotoxin G VIA. (o-conotoxin GVIA is specific for the N-subtype and binds in an

irreversible manner (see Olivera et aL, 1994; Lynch III, 1997, for reviews). P/Q- type

VDCCs (aiA or Cay2.1 current) are sensitive to block by co-agatoxin IVA. co-

agatoxin IVA is specific for both P- and Q-subtypes, yet block o f P-type Ca " current

is 10-fold more potent than that of Q-type. In contrast to the actions of co-conotoxin

GVIA, co-agatoxin IVA mediated Ca " current block is reversible (see Olivera et aL,

1994; Lynch III, 1997, for reviews). Both N- and P/Q-type VDCCs are widely

expressed throughout the brain and spinal cord (Kerr et aL, 1988; Mintz et aL, 1992;

Gohil et aL, 1994). More specifically, the N-type channel is concentrated in laminae

I and II o f the superficial dorsal horn, where nociceptive primary afferents synapse

97

(Kerr et al. 1988; Gohil et aL, 1994). Staining of the dorsal laminae of the rat spinal

cord has revealed a complementary distribution of aiA and am subunits in nerve

terminals in the superficial laminae (Westenbroek et aL, 1998). Within these

neurones the aiA and am subunits are located predominantly along apical dendrites

at presynaptic locations, Avith substantially lower levels o f somal expression

(Westenbroek et aL, 1998). Further to this, in vitro studies have demonstrated the

requirement of Ca " influx through N- and P/Q-type VDCCs for depolarization-

coupled neurotransmitter release (Miljanich & Ramachandran, 1995). Spinal release

of primary afferent peptides, CGRP and substance P, is (o-conotoxin sensitive (Holz

et aL, 1988; Maggi et aL, 1990; Santicioli et aL, 1992), and glutamate release is both

co-conotoxin and co-agatoxin sensitive (Dickie & Davies, 1992; Turner et aL, 1992;

Turner et aL, 1993). Additionally, Ca " influx into neurones can enhance excitability

and produce intracellular changes including gene induction.

VDCC antagonists are antinociceptive in models of inflammation, based on

behaviour and in vivo electrophysiology, confirming a role for Ca influx into

neurones in the spinal processing o f nociceptive information (see Vanegas &

Schaible, 2000, for review). In the formalin test, behavioural antinociception is

observed to varying extents with N-, P/Q- and L-type VDCC blockers (Malmberg &

Yaksh, 1994; Malmberg & Yaksh, 1995; Bowersox et aL, 1996), and Diaz and

Dickenson (1996) reported phase-dependent effects o f N- and P/Q-type blockers on

neuronal responses. Blockade of N- and P/Q-type VDCCs has been shown to reduce

enhanced neuronal responses (Neugebauer et aL, 1996; Nebe et aL, 1997; Nebe et

aL, 1998) and nociceptive behaviours (Sluka, 1997) after knee joint inflammation

and intradermal capsaicin (Sluka, 1998). Studies in models of neuropathy have been

limited, however Ca chelation (White & Cousins, 1998) and N-type VDCC block

have been shown to be effective (Chaplan et aL, 1994; Xiao & Bennett, 1995;

Bowersox et aL, 1996; White & Cousins, 1998). In contrast L- and P/Q-type

antagonists show little effect (Chaplan et aL, 1994; White & Cousins, 1998).

Currently there are no in vivo electrophysiological studies investigating the

role of spinal VDCCs in the processing of neuropathic pain. This chapter investigates

the role of N- and P/Q-type VDCCs in the transmission o f sensory information at the

98

level of the spinal cord and plasticity of these channels that may occur following

nerve injury. After behavioural establishment of a neuropathic state, using the SNL

model (Kim & Chung, 1992), electrophysiological studies of dorsal horn neurones

were made to investigate the effects of spinally delivered w-conotoxin GVIA (N-type

VDCC blocker) and co-agatoxin IVA (P/Q-type VDCC blocker) on a wide range of

electrically- and naturally-evoked neuronal activity.

4 .2 M e t h o d s

A total o f 72 male adult Sprague-Dawley rats were used in this study, of

which 27 were naïve, 19 were sham-operated and 26 were SNL. The animals were

prepared in the normal way and in separate experiments the effects o f co-conotoxin

GVIA and co-agatoxin IVA were investigated on the evoked dorsal horn neuronal

responses. The toxins were applied directly onto the exposed surface o f the spinal

cord (0.1, 0.4, 0.8, 1.2, 2.2 and 3.2 pig). The effects of each dose were followed until

the responses plateaued (a minimum of 60 minutes), when the next dose would be

applied cumulatively.

4 .3 R e s u l t s

4.3 .1 B e h a v io u r a l A s se s s m e n t

Rats subjected to SNL exhibited 'guarding' foot posture ipsilateral to nerve

injury and successful replication o f the nerve injury model was confirmed by

behavioural testing which demonstrated the development of mechanical and cooling

allodynia o f the injured hindpaw (see chapter 3 for combined behavioural data

obtained from all studies used in this thesis). This was not displayed by either the

contralateral hindpaw, or in the sham-operated rats. Upon establishment o f a

neuropathic state, animals were then used for in vivo electrophysiology and the

pharmacological study at PO days 14-17 .

99

4.3.2 S p in a l C o r d E l e c t r o p h y s i o l o g y : N e u r o n e C h a r a c t e r i z a t i o n

The number of dorsal horn neurones recorded from in each group were 23

in SNL rats, 21 in sham-operated rats and 37 in naïve rats. All neurones had a

receptive field over the hindpaw ipsilateral to surgery (when performed). The level

of ongoing spontaneous activity was found to be significantly higher (P < 0.05) in

neurones in SNL rats (6.36 ± 2.24 Hz) compared to sham (0.34 ± 0.22 Hz) and naïve

(0.83 ± 0.69 Hz). No significant differences were found between experimental

groups in the mean cell depth o f recorded neurones and the mean neuronal responses

evoked by electrical and natural stimulation.

4.3.3 S p in a l C o r d E l e c t r o p h y s i o l o g y : E f f e c t s o f c o - c o n o to x in G VIA

The effect of co-conotoxin GVIA (0.1 - 3.2 p.g), applied directly onto the

spinal cord, on the electrically- and naturally-evoked dorsal horn neuronal responses

was tested in SNL, sham-operated and naïve animals. Since the sham-operation is the

appropriate control for SNL, and for clarity, the results obtained from the naïve, non­

operated, group shall not be displayed on the graphs.

At the higher doses, co-conotoxin GVIA produced inhibitions o f all the

electrically- and naturally-evoked responses in neurones in all experimental groups

(Figure 10). Interestingly, in the sham-operated group, low doses tended to facilitate

the neuronal responses, especially those induced by electrical stimulation. For each

dose (0.1, 0.4, 0.8, 1.2, 2.2, and 3.2 jLig) the effects of the toxin were slow in onset

with clear effects seen around 40 minutes, maximal inhibitions established at 70 - 90

minutes (Figure 11), and these were irreversible. In SNL rats, all doses o f co-

conotoxin GVIA (0.1 - 3.2 fxg) elicited significant effects, compared to pre-drug

control, o f the C-fibre, excess spikes, input, post-discharge and Aô-fibre, responses

(Figures 10a and 12a - e, respectively; P < 0.05; n = 8), with similar mean maximal

inhibitions at top dose in the range 86 ± 6% to 96 ± 3%. The A(3-fibre response was

significantly inhibited with 0.4 - 3.2 jLig co-conotoxin GVIA, but only reached a

maximum inhibition of 41 ± 10% (Figures 10a and 12f).

100

SNL Shama 100-1100-1

7 5 -7 5 -

.2 5 0 - 5 0 -

►5 2 5 - 2 5 -

-25-<0.10.1

co-conotoxin GVIA (fxg) CD-conotoxin GVIA (jig)

NaiveC 100 1

7 5 -

c:o5 0 -

2 5 -

0.1 10(o-conotoxin GVIA (|ig)

A^-fibre

Aô-fibre

C-fibre

Post-discharge

Input

Excess spikes

Figure 10. The effect o f spinally applied co-conotoxin GVIA on the electrically-evoked

dorsal horn neuronal responses recordedfrom (a) SNL (n = 8), (b) sham-operated (n =

6) and (c) naive rats (n = 11) at PO days 14 - 17. Data are expressed as the mean

maximal % inhibition o f the pre-drug conti'ol values S.E.M.

101

Pre-drug co-conotoxin GVIA (cumulative dose)

0.1 lag 0.4]Lig O.SjLig 1.2|Lig 2 .2 |ag 3.2|Lig150 -I

100 -

ocoU

-A— AP-fibrc

O— C-fibre

50 -

von Frey 75g

0 100 200 300 4 0 0

time (min)

Figure 11. Time course o f the effect o f spinally applied co-conotoxin GVIA on the

evoked response o f a typical dorsal horn neurone recorded from a sham-operated rat.

Examples o f the effect on the Afffibre, C-fibre and von Frey 75 g measurements are

shown with the cumulative dose indicated. Data are expressed as % o f the ctveraged

pre-drug control values.

102

Excess spikesC-fibre ba100-1 lOOl

75-75-co

100.1 10 0.1 11Q-conotoxin GVIA (pig) co-conotoxin GVIA (pig)

Post-dischargeInput dc 1001 100

co

25-

-25-'10 0.1 1 10

co-conotoxin GVIA (pig)

Aô-fibree100-1

-250.1 10

co-conotoxin GVIA (pig)

Ap-fibref lOOlSham

75-SNL

50-

25-

-25-'

co-conotoxin GVIA (pig)

Figure 12. Comparison o f the effect o f spinally applied ohconotoxin GVIA on the

electrically-evoked dorsal horn neuronal responses recorded from sham-operated (n =

6) and SNL (n = 8) rats at PO days 14-17. Data are expressed as the mean maximal %

inhibition o f the pre-à'ug control values ± S.E.M. denotes a significantly greater

inhibitory effect (P < 0.05) o f the toxin at a specific dose in SNL rats compared to

sham.

103

In contrast to SNL animals, sham-operated rats required higher

concentrations of the toxin to achieve significant inhibition o f the electrically-evoked

responses, but still reached similar maximal levels of inhibition at the top dose of 3.2

pig. co-conotoxin GVIA (0.8 - 3.2 pig) significantly inhibited the post-discharge and

Aô-fibre responses (Figures 10b and 12d - e; P < 0.05; n = 6), whereas the C-fibre,

excess spikes and input responses (Figures 10b and 12a - c) were significantly

inhibited over the dose range 1.2 - 3.2 pig (P < 0.05; n = 6). In a similar fashion to the

sham-operated group, naïve animals also required higher concentrations of the toxin

to achieve significant inhibition of the electrically-evoked responses (Figure 10c). co-

conotoxin GVIA (0.8 - 3.2 pig) significantly inhibited the C-fibre, input and Aô-fibre

responses (P < 0.05; n = 11) and 2.2 - 3.2 pig produced a significant inhibition (P <

0.05; n = 11) of the post-discharge and excess spikes responses. Maximum

inhibitions ranged from 69 ± 12% to 89 ± 6%. As in SNL rats, co-conotoxin GVIA

had limited effects on the Ap-fibre response in sham (Figure 12f) and naïve rats

(Figure 10c) where the mean maximum inhibitions were 25 ± 7% and 44 ± 6%,

respectively.

After neuropathy the inhibitory effects o f co-conotoxin GVIA on the

electrically-evoked responses appeared to be more marked at the lower end of the

dose range in comparison to sham-operated rats. This was statistically significant (P

< 0.05; n = 8) at 0.1 and 0.4 pig for the C-fibre, excess spikes, input, and post­

discharge responses (Figures 12a - d) and at 0.4 pig for the Aô-fibre response (Figure

12e). 0.1 and 0.4 pig co-conotoxin GVIA had a significantly greater effect (P < 0.05)

on the: C-fibre response (Figure 12a) in SNL rats (29 ±5% and 47 ± 9% inhibitions,

respectively, n = 8) compared to sham rats (7 ± 10% and 5 ± 15% facilitations, n =

7); excess spikes response (Figure 12b) in SNL rats (63 ± 12% and 77 ± 8%

inhibitions, respectively, n = 8) compared to sham (14 ± 10% and 15 ± 17%

inhibitions, n = 7); input response (Figure 12c) in SNL rats (30 ± 5% and 54 ± 9%

inhibitions, respectively, n = 8) compared to sham (19 ± 15% and 14 ± 21%

facilitations, n = 7); post-discharge (Figure 12d) in SNL (50 ±13% and 67 ± 10%

inhibitions, respectively, n = 8) compared to sham (6 ±11% and 20 ± 16%

inhibitions, n = 6). 0.4 pig co-conotoxin GVIA had a significantly greater effect (P <

0.05) on the Aô-fibre (Figure 12e) in SNL rats (50 ± 10% inhibition, n = 8)

104

compared to sham (21 ± 19% inhibition, n = 5).

co-conotoxin GVIA also produced an inhibitory effect on the naturally-

evoked neuronal responses (Figure 13). At 3.2 )Xg co-conotoxin GVIA, the response

to non-noxious mechanical stimulation (von Frey 9 g) was inhibited by 95 ± 3% in

both SNL (n = 5) and sham rats (n = 5), and 75 ± 6% in naïve rats (n = 10) (Figure

13a, naïve not shown). The response to noxious mechanical stimulation (von Frey 75

g) was inhibited by 83 ± 5% in SNL rats (n = 7), 78 ± 12% in the sham control group

(n = 6), and 69 ± 5% in naïve (n = 10) (Figure 13b, naïve not shown). The response

to noxious thermal stimulation (water jet at 45 °C) was inhibited by 89 ± 5% in SNL

rats (n = 6), 93 ± 5% in sham rats (n = 5) and 79 ± 9% in naïve (n = 8) (Figure 13c,

naïve not shown). After neuropathy, co-conotoxin GVIA had a greater inhibitory

effect on the naturally-evoked neuronal responses at the lower end of the dose range

compared to sham-operated animals. This was found to be statistically significant at

0.1 and 0.4 jug co-conotoxin GVIA, for the von Frey 9 g evoked responses (Figure

13a) (55 ± 6% and 69 ± 7% inhibitions, respectively, in SNL rats, compared to 6 ±

12% facilitation and 17 ± 11% inhibition, respectively, in sham rats; P < 0.001; n =

8), and at 0.4 p,g for the von Frey 75 g evoked responses (Figure 13b) (51 ± 6%

inhibition in SNL rats compared to 21 ± 12% inhibition in sham; P < 0.05; n = 8).

105

vF 75gblOOn 1 00-1

75-

50-

25-c 2 5 -

5

-25-'-25-"0.1 1 100.1 1 10

co-conotoxin GVIA (^g) co-conotoxin GVIA (ng)

HeaC1 0 0 -1

7 5 -

cO:-s 5 0 -_o

-250.1 10

Sham

co-conotoxin GVIA (jiig)

Figure 13. Comparison o f the effect o f spinally applied (o-conotoxin GVIA on the

naturally-evoked dorsal horn neuronal responses recorded from sham (n = 6) and SNL

(n = 8) rats. Data are expressed as mean maximal % inhibition of pre-drug control

values ± S.E.M. denotes a statistically significant gt'eater inhibitory effect (P < 0.05)

o f the toxin at a specific dose in SNL rats compared to sham.

106

4.3.3 S p i n a l C o r d E l e c t r o p h y s i o l o g y : E f f e c t s o f cd- a g a t o x i n IVA

co-Agatoxin IVA inhibited the electrically-evoked neuronal responses

measured in all experimental groups, but to a lesser extent than that achieved with co-

conotoxin GVIA (Figure 14). For each dose (0.1, 0.4, 0.8, 1.2, 2.2, and 3.2 p,g) clear

effects o f the toxin were seen around 40 minutes, maximal inhibitions established at

70 - 90 minutes (Figure 15), and these were partially reversible. In SNL animals, at

the highest dose, mean maximal inhibitions from pre-drug controls were in the range

37 ± 13% to 60 ± 10% for excess spikes, input, post-discharge and AÔ-fibre (Figures

14a and 16b - e), and only 18% for the C- and Ap-fibre (Figures 14a and 16a & f,

respectively), at which significance was achieved (P < 0.05; n = 7). Similar maximal

inhibitions were observed in sham-operated rats, and there was no difference in drug

effect at the lower doses in comparison to SNL animals (Figures 14b and 16). co-

agatoxin IVA had no significant inhibitory effects on the responses in naïve animals,

however with the low doses (0.1 - 0.8 jxg) there was a tendency for the electrically-

evoked neuronal responses to be facilitated and this was found to be statistically

significant (P < 0.05; n = 5) for the C-fibre response (Figure 14c).

The naturally-evoked neuronal responses were also inhibited by co-agatoxin

I VA (Figure 17). In SNL rats, at the highest dose, mean maximal inhibitions from

the pre-drug control were in the range 57 ± 11% to 71 ± 12%, and this was

statistically significant for von Frey 9 and 75 g (P < 0.05; n = 7). Again, no

difference in the extent of inhibition was observed between experimental groups.

The spontaneous rate o f activity, when present, was clearly inhibited by

both toxins. Due to the low pre-drug baseline values and variability between

neurones in this ongoing activity was not quantified.

107

SNL Sham

.2 50

co-agatoxin IVA (pig)0.1 1 10

co-agatoxin I VA (pig)

Naive100-1

75 -

S 5 0 -

25 -

- 2 5 -

-50 J

co-agatoxin IVA (pig)

- A^-fibre

—e —- Aô-fibre

— h r — - C-fibre

- Post-discharge

- Input

---* - Excess spikes

Figure 14. The effect of spinally applied (D-agatoxin IVA on the electrically-evoked

dorsal horn neuronal responses recordedfrom (a) SNL (n = 7), (b) sham-operated (n =

7) and (c) naive rats (n = 5) at PO days 14- 17. Data are expressed as the mean

maximal % inhibition o f the pre-drug control values + S.E.M.

108

Pre-drug (o-agatoxin IVA (cumulative dose)

1 IO.SjLig 1.2|Lig 2.2jLig 3.2|Lig150 1

100 -

ocoU^ 50 - A (3-fibre

Post-discharge

von Frey 75g0-*T T TT I

100 200

time (min)300 400

Figure 15. Time course o f the effect o f spinally applied co-agatoxin IVA on the evoked

response o f a typical dorsal horn neurone recorded from a sham-operated rat.

Examples o f the effect on the Afi-fibre, post-discharge and von Frey 75 g measurements

are shown with the cumulative dose indicated. Data are expressed as % o f the averaged

pre-drug control values.

109

C-fibrea100-1

75 -

-25-'0.1 10

Excess spikes

n - T T |

co-agatoxin IVA (jig)

100-1

7 5 -

5 0 -

2 5 -

0.1 1co-agatoxin IVA (p,g)

10

Post-dischargeInput d1 0 0 - 1lOOl

7 5 -75 -

§ 5 0 -_DC 25 -

0.1 10 0. 1 10co-agatoxin IVA (pg) co-agatoxin IVA (pig)

Aô-fibre^ lOOi

75 -co

a 25-

-25-0.1 10

co-agatoxin IVA (pig)

Afi-fibref* lOOi

Sham

SNL

0.1 1 10co-agatoxin IVA (pig)

Figure 16. Comparison of the effect o f spinally applied (o-agatoxin IVA on the

electrically-evoked dorsal horn neuronal responses recorded from sham-operated (n =

7) and SNL (n 7) rats at PO days 14-17. Data are expressed as the mean maximal %

inhibition of the pre-drug control values + S.E.M.

110

vF 9g vF 75gb100-1 100-1

75 -75 -

•2 5 0 - 50 -

Lcc 25 -

-25 -J-25 ^10 0.1 101

co-agatoxin IVA (jiig)

Heat100 1

o

c

-25 -

-50 J0.1 1 10

co-agatoxin IVA ( jL ig )

Figure 17. Comparison of the effect of spinally applied (D-agatoxin IVA on the

naturally-evoked dorsal horn neuronal responses recordedfrom sham (n= 7) and SNL

(n = 7) rats. Data are expressed as mean maximal % inhibition o f pre-dbmg control

values S.E.M.

I l l

4 .4 D is c u s s io n

In this study I have addressed the role of members o f the Cay2 family of

VDCCs in the spinal processing of sensory information after nerve injury. Spinal co-

conotoxin GVIA, an N-type (C a y 2 .2 , a is) VDCC blocker, inhibited the electrically-

and naturally- (innocuous and noxious) evoked dorsal horn neuronal responses. The

effects o f low doses were significantly enhanced after the establishment of

neuropathy. In comparison, the P/Q-type (C a y 2 .1 , Qia) blocker co-agatoxin IVA

inhibited the evoked neuronal responses to a lesser extent and its profile was

unaltered after nerve injury.

As regards the concentration and specificity o f the toxins used it has been

demonstrated in vitro that N-type VDCCs are specifically and completely blocked by

O.I - 0.5 p,M co-conotoxin GVIA. Block of P-type Ca " current is blocked by 2 nM

co-agatoxin IVA, which is 100-fold less than that required to block Q-type Ca "

current (Lynch III, 1997). Although the concentrations of the toxins at their sites of

action in vivo obviously cannot be accurately determined, the concentration applied

to the surface o f the spinal cord at a mid range dose (0.8 p,g) was 5 p-M for co-

conotoxin GVIA and 3 pM for co-agatoxin IVA. These concentrations are higher

than the selective dose-ranges for these toxins in vitro, however in vitro the toxins

are applied directly to the membrane. Within the spinal cord these concentrations

will be several orders o f magnitude less due to the diluting effect o f diffusion into the

surrounding tissue and vasculature. Additionally, since the toxins are large peptides

access into the tissue is likely to be limited as demonstrated by the slow

establishment of maximal inhibitions (Figures 11 & 15). In addition, the different

profile o f action o f co-conotoxin GVIA and co-agatoxin IVA further suggests that

they are acting at the appropriate VDCC subtype.

Blockade of Ca current through both N- and P/Q-type VDCCs inhibited,

to differing extents, the electrically- and naturally-evoked neuronal responses in both

naïve un-operated and sham operated animals as well as after the establishment of

neuropathy in SNL rats. This is indicative of a role o f Cay2 VDCCs in sensory

transmission at the level of the spinal cord. This is in agreement with their known

112

importance in neurotransmitter release (Miljanich & Ramachandran, 1995) and their

reported distribution within the spinal cord (Kerr et al., 1988; Mintz et al., 1992;

Gohil et al., 1994). cû-conotoxin GVIA and co-agatoxin IVA are likely to be acting at

VDCCs located at a number of possible sites within the spinal cord that may be both

pre- and postsynaptic, as will be discussed. Furthermore SNL appears to selectively

impart plasticity on only N-type VDCCs, enhancing their role in sensory

transmission.

4.4 .1 R o l e o f S p in a l N - ty p e VDCCs (CAv2.2, a ie )

I have demonstrated an increased inhibitory effect of co-conotoxin GVIA at

the low doses in SNL animals, compared to sham and naïve, on evoked dorsal horn

neuronal responses. A possible mechanistic explanation for this is that if there are

increases in the probability and/or frequency of opening o f pre-existing VDCCs, the

likelihood o f a given concentration of co-conotoxin GVIA blocking a functionally

important N-type channel would be increased. Indeed, it is likely that Ca " influx via

VDCCs on central terminals of peripheral nerves and neurones, and consequently

neurotransmitter release and membrane depolarization (Smith & Augustine, 1988),

are increased due to enhanced neuronal activity observed peripherally after

neuropathy (Wall & Devor, 1983; Sheen & Chung, 1993). Furthermore, despite

reduced afferent input via L5 and L6 spinal nerves, the magnitude o f neuronal

responses recorded was not diminished after SNL. Conversely, increased frequency

and occurrence of spontaneous activity was observed alongside enlarged receptive

fields of these neurones (as discussed in chapter 3). This suggests that there are

compensatory increases in peripheral and/or spinal neuronal activity after neuropathy

leading to a greater functional role of VDCCs in the observed neuronal responses.

N-type VDCCs are subject to modulation by G-proteins and protein kinases

and this serves to fine tune neurotransmitter release. VDCCs can be tonically

inhibited by several G-protein-activating neurotransmitter systems. Binding of the

activated G-protein Py subunit to the domain I-II linker region o f N-type VDCCs,

and to a lesser extent P/Q- and possibly R-types, results in a depression o f Ca

current (see Zamponi & Snutch, 1998, for review). This tonic inhibition can be

113

overridden specifically in N-type VDCCs by activation of protein kinase C (PKC)

(Swartz, 1993; Swartz et ah, 1993; Zamponi et ah, 1997; Barrett & Rittenhouse,

2000) via so-called PKC/G-protein cross-talk that is mediated by PKC

phosphorylation o f a singe threonine residue (Thr-422) in the (Xm domain I-II linker

(Hamid et ah, 1999). This has been shown to dramatically reduce somatostatin (and

to a lesser extent opioid) receptor-mediated G-protein inhibition o f N-type VDCC

activity (Hamid et ah, 1999; Cooper et ah, 2000). Additionally, PKC activation can

also enhance N-type Ca current in the absence o f tonic inhibition (see Yang &

Tsien, 1993).

Of interest here, is the fact that PKC has been implicated in persistent pain

states. For example, inhibitors o f spinal PKC exert anti-allodynic effects in

neuropathic models o f SNL (Hua et ah, 1999) and diabetic neuropathy (Ahlgren &

Levine, 1994) and prevent the development of thermal hyperalgesia induced by PSL

(Ohsawa et ah, 2000). PKC inhibitors simultaneously decrease enhanced levels of

membrane bound PKC and neuropathic behaviours induced by nerve injury (Mao et

ah, 1992). Furthermore, mice lacking the PKCy isoform gene show normal acute

pain responses in the absence o f tissue or nerve injury, but exhibit a significant

decrease in behavioural aliodynia that occurs days after injury (Malmberg et ah,

1997). Interestingly, PKCy is restricted to a population of local circuit interneurones

in lamina II o f the dorsal horn that are not inhibitory (Martin et ah, 1999a) and its

presence here is critical for sustained hyperexcitability o f lamina V WDR neurones

after mustard oil induced injury (Martin et ah, 2001). PKC has also been directly

linked to neurotransmitter release in a study that demonstrated PKC activation

increased the capsaicin-evoked release of substance P and CGRP from spinal cord

slices (Frayer et ah, 1999). This indicates that PKC’s role in pain behaviours likely

results from its ability to increase nociceptive sensory transmission at the level of the

spinal cord. Given: (1) the current-enhancing effect o f PKC-mediated

phosphorylation on N-type VDCCs; (2) the role o f both N-type VDCCs and PKC in

neuropathy; (3) the localization o f both the N-type aiB subunit and PKCy to the

superficial dorsal horn, (4) that the tonic inhibition o f N-type VDCCs may be

relieved due to nerve injury induced alterations in neurotransmitter systems

responsible; it seems possible that the increased effectiveness o f ca-conotoxin GVIA

after nerve injury on the evoked neuronal responses I demonstrate here could be due

114

to an increased activity o f existing N-type VDCCs mediated by PKC

phosphorylation of the channel

An alternative, or additional possibility is that rather than an increase in

activity of existing VDCCs, via whatever mechanism, there may be an increase in

VDCC expression. In support of this idea, the expression of the N-type VDCC

subunit is increased in small DRG cells and in lamina II of the spinal cord after

neuropathy (Cizkova et a l , 1999). Furthermore, the mRNA and protein for the

auxiliary « 2 subunit o f VDCCs is also upregulated in the ipsilateral dorsal root

ganglia and spinal cord of SNL rats (Luo et ah, 2001). In addition to its ability to

modulate VDCC kinetics (Hobom et aL, 2000), « 2 can modulate the binding

affinity of co-conotoxins to the N-type VDCC (Brust et a l , 1993). Although it is

generally believed that the co-toxins primarily affect the function of the ttj subunit

(Olivera et aL, 1994), evidence indicates that they can also bind the ci component of

the CX2Ô subunit (Barhanin et aL, 1988). It remains to be determined whether

upregulation o f a 20 is responsible for the selective enhancement of co-conotoxin

GVIA’s inhibitory actions I report here.

As previously discussed in chapter 3, peripheral nerve damage leads to the

generation o f ectopic activity at the injury site (Wall & Devor, 1983) or within the

DRG (Wall & Devor, 1983; Kajander et aL, 1992). This provides an ongoing,

continual barrage of neuronal activity via primary afferents into the spinal cord that

could lead to central sensitization and hyperexcitability (Gracely et aL, 1992; Sheen

& Chung, 1993; Yoon et aL, 1996). Evidence indicates that enhanced nociceptive

transmission in neuropathic pain states is mediated, at least in part, by the action of

excitatory amino acids, particularly at NMDA receptor sites (Dubner & Bennett,

1983; Dickenson & Sullivan, 1987; Coderre et aL, 1993), and NMDA antagonists

have been shown to produce antinociception in patients (Bide et aL, 1994; Price et

aL, 1994). Thus, any increased neuronal depolarization, such as that produced by

NMDA receptor events, would increase the likelihood o f VDCC activity in spinal

neurones since they are voltage-operated ion channels. Indeed, events such as these

regulate the expression of new channels and receptor subunits.

115

It is evident here that blockade o f N-type VDCCs results in a greater

inhibition of electrically- and naturally-evoked dorsal bom neuronal responses after

SNL, indicative o f a role for N-type VDCCs in mechanisms of central sensitization

of nociceptive dorsal bom neurones after neuropathy. Increased primary afferent

activity after peripheral nerve damage could result in enhanced exocytosis o f

excitatory transmitters, which would in turn increase activation of receptor systems

such as those for glutamate (NMDA), substance P and CGRP, which produce

excitation o f spinal cord neurones. Nerve terminals throughout the dorsal horn have

been demonstrated immunoreactive for N-type VDCCs, often correlating with the

presence o f substance P (Westenbroek et aL, 1998). Since N-type VDCCs are

abundant on nerve terminals and crucial for neurotransmitter release (Smith &

Augustine, 1988; Miljanich & Ramachandran, 1995) the simplest explanation for the

effects of co-conotoxin GVIA is that it blocks presynaptic N-type VDCCs and

therefore Ca^ influx and evoked neurotransmitter release (see Figure 18). This is

supported by the observation that the non-potentiated ’input’ response recorded in this

study, that likely relates to the level o f synaptic transmission between the central

terminals o f primary afferents and the neurones of the spinal cord dorsal horn, and

non-NMDA receptor mediated postsynaptic events (discussed in chapter 3), is

significantly inhibited by the spinal application o f the toxin. Small DRG cells

express N-type Ca " currents (Scroggs & Fox, 1992a; Scroggs & Fox, 1992b) and

the release o f CGRP (Santicioli et aL, 1992); sensory neuropeptides (Maggi et aL,

1990) and substance P in culture (Holz et aL, 1988) are all co-conopeptide sensitive.

As for the control of glutamate release and subsequent NMDA receptor/channel

activation, implicated in central hyperexcitability, the role of N-type VDCCs is

debatable. Given the evidence for a role o f N-type VDCCs in peptide release it

would be expected that a similar situation would exist for glutamate, although some

studies have reported that N-type VDCCs have a minimal role in mediating

glutamate release (Pocock & Nicholls, 1992; Turner et aL, 1992). On the other hand,

glutamatergic synaptic transmission between DRG cells and spinal cord neurones has

been demonstrated to be inhibited by blockade o f pre-synaptic N-type VDCCs

(Gruner & Silva, 1994), and co-conopeptides can cause significant reduction in

depolarization-evoked glutamate release (Terrian et aL, 1990; Dickie & Davies,

1992).

116

Primary a ffer en ts

R eccptiN on-m jured HyperalgesiaAllodynit

/ Expanded receptive

\ field

tstimulus

Recordingelectrodenerve

C-fibres Superficial dorsal

Ap-fibre

11 N -type VDCC p; opioid-R

I Ï P-type VDCC AAorphine

O T - W V D C C M

I * L-type VDCC GlutamateA peptides

Deep dorsal horn convergent neurone

Figure 18. Hypothetical sensory pathway through the dorsal horn o f the spinal cord. Possible locations o f VDCCs, p opioid receptors (R)

and effects o f peripheral nerve injury are indicated. Neurotransmitter release is mediated by Ca ' influx via presynaptic N- and P/Q-type

VDCCs, which can be inhibited by activation of pre.synaptic opioid receptors. Nerve damage causes ongoing afferent input and thus

increased neurotransmitter release. In particular the role o f N-type VDCCs is enhanced and implicated in altered sensory perception.

It is also likely that (O-conotoxin GVIA inhibits Ca " influx through N-type

VDCCs at postsynaptic sites (Takahashi & Momiyama, 1993) and presynaptic

locations on spinal cord intemeurones such that enhanced N-type VDCC-mediated

exocytosis of transmitters and modulators is inhibited, thereby reducing excitation of

spinal cord neurones. Nociceptive signals arriving at the spinal cord from damaged

peripheral nerves may be subject to amplification via relays o f intemeurones that

project onto an output neurone (such as the WDR neurones recorded from here)

resulting in the phenomena o f wind-up and central sensitization. In support o f a

possible interneuronal site o f action, is the fact that NMDA receptors are not

considered to be directly postsynaptic to primary afferents (Davies & Watkins,

1983). Also, the most substantial inhibitions observed in this study, produced by the

low doses of (O-conotoxin GVIA after SNL, were on the ‘excess spikes’ and ‘post­

discharge’ neuronal responses, indicators of neuronal hyperexcitability and mediated

by postsynaptic events (discussed in chapter 3). However, in the latter case, blockade

of presynaptic transmitter release by the VDCC blockers could indirectly also

prevent postsynaptic hyperexcitability due to decreased activation o f postsynaptic

NMDA receptors (see Figure 18).

In both control and SNL animals the AP-fibre mediated response was

inhibited the least by (O-conotoxin GVIA compared to all the evoked responses. This

is in agreement with the facts that substance P containing terminals within the

superficial dorsal horn do not normally arise from A^-fibre primary afferents, yet

they often do co-localize with N-type VDCC containing terminals (Westenbroek et

aL, 1998). Furthermore, nerve injury induces upregulation of the N-type VDCC am

subunit in only small DRG cells and in lamina 11 of the spinal cord after neuropathy

(Cizkova et aL, 1999) that are not associated with large diameter A^-fibre cell bodies

or normal termination patterns. An additional explanation is that the electrical

stimulus employed in this study is applied at three times the threshold required to

mediate a C-fibre evoked response. C-fibres are high threshold activated primary

afferents that convey nociceptive information, in contrast to low-threshold Ap-fibres.

Thus the experimental methodology used here is suprathreshold for Ap-fibres,

rendering inhibition of their evoked neuronal response more difficult.

118

Other in vivo electrophysiology studies have only investigated the role of

VDCCs in models o f inflammation. Inflammation can result from non-neuronal

tissue damage, and can induce a set of excitatory changes in the periphery and CNS,

which can establish a more persistent, but often reversible hypersensitivity in the

inflamed and surrounding tissue. In some cases, for example arthritis, inflammation

becomes chronic and leads to a pathological pain state. The dysfunctional

mechanisms at play are often different, commonly showing distinct and often

contrary pharmacological sensitivities to those displayed in neuropathic pain.

However, inflammation is often associated with hyperalgesia and expansion of

receptive fields, indicative of central hyperexcitability (see chapter 3), similar to that

seen in neuropathy. There are several animal models of inflammatory pain. Injection

of formalin into the hindpaw results in a biphasic response characterized by acute

(phase 1) and persistent (phase 2) nociceptive behaviours (see Porro & Cavazzuti,

1993). This is also observed neuronally, and NMDA receptor activity has been

shown to be critical in the 2" phase (Haley et aL, 1990). Of interest here, co-

conotoxin GVIA has been demonstrated to reduce the responses o f dorsal horn

neurones in both phases of the formalin test (Diaz & Dickenson, 1997). (O-conotoxin

GVIA has also been shown to inhibit the evoked responses of dorsal horn neurones

receiving both noxious and innocuous mechanical input from the knee joint in the

presence and absence o f inflammation induced by carrageenan or kaolin

(Neugebauer et aL, 1996) and mustard oil (Nebe et aL, 1998). This evidence is in

accordance within the present investigation in that blockade of N-type VDCCs has

the ability to reduce evoked dorsal horn neuronal responses to both innocuous and

noxious stimuli, indicating that the toxin is not selective with regard to stimulus

under such circumstances. However in contrast to SNL, establishment of

inflammation does not appear to alter the sensory processing profile o f the N-type

VDCC in comparison to control responses.

4.4 .2 R o l e o f S p in a l P - ty p e VDCCs (C av2.1 , Œia )

In contrast to the actions of co-conotoxin GVIA, the P/Q-type VDCC

blocker co-agatoxin IVA had limited inhibitory effects on the evoked neuronal

responses and no difference was seen between experimental groups. Interestingly, a

significant inhibition of ‘excess spikes’ and ‘post-discharge’ was seen with higher

119

doses in sham and SNL rats, although the maximum effect was lower than after co-

conotoxin GVIA. In naïve and sham-operated animals, the electrically-evoked

responses tended to be facilitated at low doses and similarly in an in vivo

electrophysiological study co-agatoxin IVA increased neuronal responses in the

absence of inflammation but mediated inhibition in its presence (Nebe et aL, 1997).

In vitro studies have demonstrated that co-agatoxin IVA blocks excitatory (Luebke et

aL, 1993; Castillo et aL, 1994; Yamamoto et aL, 1994) as well as inhibitory

(Takahashi & Momiyama, 1993) synaptic transmission, which may explain the

facilitatory and inhibitory effects observed here. P/Q-type VDCCs may be involved

in the release of glutamate, aspartate, dopamine, serotonin, noradrenaline, G ABA

and glycine (Turner et aL, 1992; Takahashi & Momiyama, 1993; Turner et aL, 1993;

Kimura et aL, 1995; Miljanich & Ramachandran, 1995).

I have demonstrated a limited role for P/Q-type VDCCs in the maintenance

o f neuropathy. The P/Q-type channel appears to be linked selectively to NMDA

receptor-mediated events as here post-SNL administered co-agatoxin IVA had most

marked effects on NMDA receptor-mediated ‘excess spikes’ and ‘postdischarge’

measurements. This is in accordance with results from inflammatory studies.

Spinally applied co-agatoxin IVA has been shown to inhibit dorsal horn WDR

neuronal responses only in the 2" phase, not the phase (Diaz & Dickenson, 1997).

In other inflammation models, P/Q-type channels may be most important in the

initiation of the facilitated pain state, since intrathecal administration of co-agatoxin

IVA to rats prior to capsaicin injection into the hindpaw (Sluka, 1997) and

carrageenan injection into the knee joint (Sluka, 1998) has been behaviourally shown

to prevent the development o f hyperalgesia.

Evidence to support a postsynaptic role for P/Q-type VDCCs is based on the

fact that postsynaptic currents in numerous CNS neurones are markedly suppressed

by co-agatoxin IVA and reduced to a lesser extent by co-conotoxin GVIA (Takahashi

& Momiyama, 1993). Presynaptic P/Q-type channels are also present, but

interestingly within the superficial dorsal horn the distribution of P/Q-type channels

complements rather than co-localizes with N-type VDCC containing neurones

(Westenbroek et aL, 1998). Immunolocalization has demonstrated the presence of

120

P/Q-type VDCCs primarily on the nerve terminals o f dorsal horn neurones in

laminae II-VI, but unlike N-type channels, rarely in substance P containing neurones

(Westenbroek et ah, 1998). Intriguingly, this may indicate that N-type channels

regulate the release o f neurotransmitters, including glutamate, from peptide-

containing C-fibres, whereas P/Q-type channels could possibly be localized to non­

peptide, IB4-positive neurones. Thus based on their physical location P/Q-type

VDCCs are not implicated in the nociceptive transmission that occurs at the first

synapse between primary afferent neurones and neurones of the superficial dorsal

horn. This is supported by the observation here that the non-potentiated ‘input’

measurement, reflective of the level of activity encountered at the primary afferent

terminal/superficial dorsal horn synapse was not as inhibited by co-agatoxin IVA in

comparison to the potentiated ‘post-discharge’ and ‘excess spikes’ measurements.

Furthermore, both N- and P/Q-type VDCCs are found in the deeper laminae of the

dorsal horn, suggesting localization on terminals of large fibres and spinal neurones

(Westenbroek et aL, 1998).

4 .4 .3 Su p p o r t in g E v id e n c e a n d I m p l ic a t io n s

There are numerous behavioural studies that support a role for Ca influx

into neurones in the spinal processing of nociceptive information. Non-specific

VDCC blockers, lanthanum (LaClg) or neodymium (NdCb), administered

intrathecally have been shown to be antinociceptive in acute tail flick and hot plate

tests (Reddy & Yaksh, 1980), and have produced dose-dependent inhibition of both

the acute phase 1 and tonic phase 2 of the formalin test (Malmberg & Yaksh, 1994;

Malmberg & Yaksh, 1995). Further to this, specific block of N-type VDCCs has

been shown to be antinociceptive against acute, high-threshold, thermal stimulus,

using co-conotoxin GVIA in the rat tail-flick test (Wei et aL, 1996) and SNX-111

(also known as ziconotide), a synthetic co-conopeptide that selectively and reversibly

blocks N-type VDCCs, in the hot plate test (Malmberg & Yaksh, 1994; Malmberg &

Yaksh, 1995; Wang et aL, 2000b). Conversely, co-agatoxin IVA was without effect

(Malmberg & Yaksh, 1994).

121

In the formalin model o f inflammation, intrathecal administration of SNX-

111 has been shown to mediate potent inhibition o f flinching behaviour in both

phases in the majority of studies and appears to be dependent on chronic infusion

(Malmberg & Yaksh, 1994; Malmberg & Yaksh, 1995; Bowersox et aL, 1996).

Acute intrathecal administration of SNX-111 appears to only inhibit the 2° tonic

phase (Bowersox et aL, 1996; Wang et aL, 2000b). In contrast P/Q-type VDCC

block produced only modest inhibition of the acute phase, but potent inhibition o f the

prolonged phase (Malmberg & Yaksh, 1994; Malmberg & Yaksh, 1995). Likewise,

in other models o f inflammation, intrathecally delivered co-conotoxin GVIA and co-

agatoxin IVA given prior to injection of capsaicin into the hindpaw plantar (Sluka,

1997) and carrageenan into the knee joint (Sluka, 1998) prevented the development

o f secondary mechanical hyperalgesia and allodynia and heat hyperalgesia,

respectively. In all these models of inflammation blockade o f P/Q-type VDCCs is

only effective when the agent is administered before inflammation induction,

whereas it appears that N-type block is effective both before and after its

establishment. Intrathecal administration o f SNX-111 has also been demonstrated to

alleviate established heat hyperalgesia and mechanical allodynia in a rat model of

post-operative pain, selective to the injured paw (Wang et aL, 2000b).

This electrophysiological study, investigating the role o f N- and P/Q-type

VDCCs in sensory transmission at the level of the spinal cord after nerve injury,

extends results from behavioural studies confirming that spinal N-type VDCCs are

the predominant isoform involved in pre- and postsynaptic processing of nociceptive

information. Tactile allodynia induced by the SNL model has been shown to be

alleviated by intrathecal N-type blockers, SNX-111, -239 and -159, without effect

when administered intravenously or regionally to the nerve. In contrast, (D-agatoxin

IVA and SNX-230 (synthetic P/Q-type selective conopeptide homologue) were

inactive (Chaplan et aL, 1994; Bowersox et aL, 1996). In the CCI model o f nerve

injury, SNX-111 and -124 reduced heat hyperalgesia and mechano-allodynia when

delivered directly to the site o f injury and these were without effect in the uninjured

nerve (Xiao & Bennett, 1995). Mechanical hyperalgesia in rats with PSL was

reduced by subcutaneous injection o f SNX-111 into the receptive field whereas

SNX-230 had no effect. Neither compound had any effect in control animals (White

& Cousins, 1998). I found a broad spectrum o f activity o f the N-type channel

122

blocker, co-conotoxin GVIA, on both electrically- and naturally-evoked neuronal

responses encompassing the single modalities used in these behavioural studies. It

has been suggested that in the brain N-type channels are located predominantly on

presynaptic terminals with a smaller proportion found postsynaptically, whereas P/Q-

type channels appear to have functional roles both pre- and postsynaptically (see

Lynch III, 1997). If this holds for sensory pathways in the dorsal horn, then the

present results would suggest that the shift in the dose-response curve for the N-type

VDCCs but not the P/Q-type is indicative o f considerable plasticity in the N-type

channels located presynaptically with less pronounced changes at postsynaptic sites.

It may then be that the peripheral neuropathy selectively impacts upon the terminals

of the injured and remaining non-injured peripheral neurones. The increased activity

o f spinal N-type VDCCs would be expected to lead to increased neurotransmitter

release promoting spinal mechanisms o f hyperexcitability that contribute to the

ensuing neuropathic syndromes (see Figure 18).

Recent studies have ventured into the role o f N-type VDCCs in nociception

in mice. The distribution of aiB subunits in murine DRG has been observed to be

similar to that observed in rat (Murakami et aL, 2001). Intrathecal administration of

(O-conotoxin SVIB, a novel analogue of co-conotoxin GVIA which blocks N-type

VDCCs via an alternative mechanism, has been shown to be antinociceptive in the

2" phase o f the formalin test, whereas co-conotoxin GVIA was antinociceptive in

both phases (Murakami et aL, 2001). Exploitation of mice permits the use of knock­

out animals and recently the nociceptive responses of mice lacking the N-type

channel cxis subunit have been characterized. Cay2.2-/- mutant mice, not showing

any motor deficits, have been shown to exhibit normal responses to acute nociceptive

stimuli in one study (Saegusa et aL, 2001), yet reduced responses to mechanical and

thermal stimuli in another (Kim et aL, 2001). Both studies demonstrated reduced

nociceptive responses in the 2" phase o f the formalin test, but not in the acute 1®

phase (Kim et aL, 2001; Saegusa et aL, 2001). Of most interest here, is the greatly

reduced development of neuropathic pain symptoms, including thermal hyperalgesia

and mechanical allodynia induced by SNL, in comparison to that observed after SNL

in normal mice (Kim et aL, 2001; Saegusa et aL, 2001). This further substantiates the

predominant role of N-type VDCCs in pathophysiological pain states.

123

Although not investigated in this study, due to the lack of any known

specific blocking pharmacological tools, the third member of the Cay2 family, Cay2.3

or R-type, (%iE, has been investigated by use of mutant mice. Mice lacking the aiE

subunit were shown to exhibit normal pain behaviours to acute mechanical, thermal

and chemical stimuli, yet in comparison to mice expressing the aiE subunit, mutant

mice displayed reduced response to somatic inflammatory pain (Saegusa et aL,

2000). Neuropathic pain behaviours were not investigated, however this provides

further evidence of a role of the neuronal Cay2 VDCC family in facilitated pain

states.

Since in this study the rats were under halothane anaesthesia any

disturbance to the general behaviour caused by the toxins, in particular motor effects,

would go undetected. The doses utilized here are comparable to those used in

behavioural studies where no obvious side-effects were noted. Limiting effects of the

VDCC blockers observed in behavioural studies include intermittent spontaneous

shaking behaviour and serpentine-like movement o f the tail (Malmberg & Yaksh,

1994; Malmberg & Yaksh, 1995). In a phase I/II open-label clinical study SNX-111

delivered intrathecally to 31 patients with chronic pain conditions with diverse

backgrounds including cancer, AIDS, spinal cord injury, thalamic pain and

postherpetic neuralgia reported partial to complete pain relief in instances were

opioids were found to be ineffective (Brose et aL, 1996). The most common side-

effects were nystagmus, mental confusion, difficulty in word finding, nausea,

dizziness, headache and disturbance of gait and balance. Although intrathecal

administration of SNX-111 may selectively target spinal cord N-type VDCCs and

limit hypotensive effects that would be caused by a reduced sympathetic efferent

action, from this clinical study it appears that there is still some spread of the drug

from the spinal cord to the brain. However, it should be noted these adverse effects

reversed upon discontinuation or decrease in dose o f SNX-111 since this analogue of

the natural toxin has reversible effects whereas co-conotoxin GVIA effects tend to be

irreversible in nature. It appears that pharmacological targeting o f these channels

with toxin blockers as a potential therapy for the treatment o f chronic pain is

hindered by the adverse systemic side-effects (Penn & Paice, 2000) and the

inconvenient spinal route of administration.

124

The present results, together with a number of previous behavioural reports,

emphasize the key roles of VDCCs in sensory events within the spinal cord. Here, I

show that N-type, in comparison to P/Q-type, VDCCs are the predominant functional

type in terms of a wide range of sensory modalities and exhibit plasticity following

nerve injury. The increased ability of co-conotoxin GVIA to inhibit evoked neuronal

responses to a wide range o f stimuli suggests that N-type VDCCs are a key

component of increased central excitability that follows nerve injury.

125

CHAPTER 5

VOLTAGE-DEPENDENT CALCIUM CHANNELS:

THE CAv3 FAMILY

126

5.1 In tro du c tio n

Neuropathic pain, arising from injury- or disease-evoked damage to the

peripheral or central nervous system, often responds poorly to traditional analgesics.

Patients can experience sensory deficits, persistent and stimulus-evoked pain

(allodynia and hyperalgesia). Peripheral nerve damage provides abnormal input into

the central nervous system that then leads to dorsal horn hyperexcitability. Under the

influence o f both excitatory and inhibitory neurotransmitter systems, the dorsal horn

is a site of peripheral input modulation before projection to higher brain centres, thus

it controls the stimulus-response relationship. Reduction o f this excitability is a

possible key to neuropathic pain management. In this context, as discussed in the

introduction, calcium influx into neurones is one event that increases excitability.

High voltage-activated (HVA) Ca " channels, of the Cayl and 2 families (L-

, N-, P/Q- and R-types), consisting of a pore-forming subunit and modulatory

accessory subunits, (3, a2-0 and y (Walker & De Waard, 1998), are widely expressed

throughout the brain and spinal cord (Kerr et al., 1988; Mintz et al., 1992; Gohil et

al., 1994; Westenbroek et al., 1998). They are activated by relatively strong

membrane depolarization and permit Ca " influx in response to action potentials.

Consequential secondary actions include neurotransmitter release; thus these

channels establish a major link between neuronal excitability and synaptic

transmission. For these reasons HVA Ca channels have been the focus o f both

acute and persistent pain transmission studies. Animal models have demonstrated the

antinociceptive abilities o f antagonists specific for L-, N- and P/Q-type Ca^

channels (see also chapters 4 and 6) highlighting the differential role each subtype

plays in nociception, often dependent on the nature o f the pain state (see Vanegas &

Schaible, 2000, for review).

In addition to HVA Ca " channels, kinetically distinct low voltage-activated

(LVA) Ca " channels o f the Cay3 VDCC family, or T-type channels, also exist both

in neuronal and non-neuronal cells. T-type VDCCs are present on cardiac muscle,

smooth muscle, secretory cells, fibroblasts and neurones (see Tsien et al., 1988;

Bean, 1989a, for reviews), including mammalian sensory neurones (Bossu et al.,

1985; Fedulova et al., 1985; Bossu & Feltz, 1986; Carbone & Lux, 1987; Kostyuk &

127

Shirokov, 1989). They are designated LVA since they activate at voltages near the

resting membrane potential, and T-type for their transient kinetics whereby they

inactivate rapidly, deactivate slowly and have a small single channel conductance

(Huguenard, 1996). In contrast to the HVA VDCCs, these unique gating properties

help determine firing frequency of cardiac pacemaking cells (Hagiwara et a l, 1988).

In neurones their unusual kinetics prohibit T-type channels alone in supporting

neurotransmission, yet permit their involvement in low amplitude oscillations,

neuronal bursting, synaptic signal boosting, Ca " entry promotion and lowering

threshold for high-threshold spike generation (Jahnsen & Llinas, 1984; White et al.,

1989). Ca^ current through these channels appears to play an important

physiological role in near-threshold phenomena and regulation o f neuronal

excitability.

Until recently the LVA current was considered a single entity. However the

genes encoding subunits responsible for Cav3,l (ajo), Cav3.2 (a^n) and Cay3.3 (a^i)

Ca " currents have now been cloned, showing 30% homology to HVA channel-

forming subunits (Cribbs et aL, 1998; Perez-Reyes et a i , 1998; Lee et al., 1999)

and displaying hallmark native T-type Ca " channel properties when expressed

heterologously. In situ hybridization studies on the rat brain have shown that T-type

Ca channels have unique distributions, including the dorsal horn of the spinal cord

and sensory ganglia (Talley et a l , 1999). This is complemented by the presence o f

T-type currents in primary sensory neurones (Carbone & Lux, 1984; Kostyuk et al.,

1992; Scroggs & Fox, 1992a; Todorovic & Lingle, 1998) and some superficial rat

dorsal horn neurones (Ryu & Randic, 1990), a site important for the processing and

integration o f sensory information including pain. T-type Ca channels were first

described in peripheral sensory neurones o f the DRG (Carbone & Lux, 1984), yet

their function in sensory processing is still unclear. Although their unique

biophysical properties make LVA currents relatively easy to study in vitro, the

involvement of T-type channels in pain-related central sensitization has been

hindered by a scarcity o f specific pharmacological agents. Unlike the HVA Ca "

channels, no natural toxins or venom components have been identified that target T-

type VDCCs selectively.

128

Neuropathic pain and epilepsy both share neuronal hyperexcitability as a

common underlying mechanism. There are established antiepileptic drugs that target

the generation of neuronal hyperexcitability in the brain and some of these have been

proven effective in the treatment of various forms of neuropathic pain (Swerdlow &

Cundill, 1981; McQuay et aL, 1995). The succinimide derivative ethosuximide (2-

ethyl-2-methylsuccinimide) is an anticonvulsant (Macdonald & McLean, 1986)

effective in the treatment o f absence epilepsy (Coulter et al., 1989c); a condition

characterized by spike-wave rhythm likely generated by T-type Ca current.

Ethosuximide has been demonstrated to be a relatively specific T-type VDCC

antagonist in thalamic (Coulter et ah, 1989b) and DRG neurones (Kostyuk et ah,

1992). The present study uses the SNL model, confirmed by behavioural testing, to

induce a neuropathic state, and subsequently electrophysiological studies of dorsal

horn spinal neurones were made to investigate the effects o f spinally delivered

ethosuximide on responses to a wide range o f electrically- and naturally-evoked

neuronal activity.

5 .2 M e t h o d s

A total of 24 male adult Sprague-Dawley rats were used in this study, of

which 7 were naïve, 6 were sham-operated and 11 were SNL. The animals were

prepared in the normal way and the effect of ethosuximide (5, 55, 555 and 1055 pg),

applied directly onto the exposed surface o f the spinal cord, was investigated on the

evoked dorsal horn neuronal responses. The effects of each dose were followed until

the responses plateaued (a minimum of 50 minutes), when the next dose was applied

cumulatively.

5.3 R e s u l t s

5 .3 .1 B e h a v io u r a l A s se s s m e n t

Rats subjected to SNL exhibited 'guarding' foot posture ipsilateral to nerve

injury and successful replication o f the nerve injury model was confirmed by

behavioural testing which demonstrated the development of mechanical and cooling

allodynia o f the injured hindpaw (see chapter 3 for combined behavioural data

129

obtained from all studies used in this thesis). This was not displayed by either the

contralateral hindpaw, or in the sham-operated rats. Upon establishment of a

neuropathic state, animals were then used for in vivo electrophysiology and the

pharmacological study at PO days 14 - 17.

5 .3 .2 S p in a l C o r d E l e c t r o p h y s i o l o g y : N e u r o n e C h a r a c t e r i z a t i o n

The number of dorsal horn neurones recorded from in each group were 11

in SNL rats, 6 in sham-operated rats and 7 in naïve rats. All neurones had a receptive

field over the hindpaw ipsilateral to surgery (when performed). No significant

differences were found between experimental groups in the mean cell depth of

recorded neurones and the mean neuronal responses evoked by electrical and natural

stimulation. However, 45% of neurones characterised in SNL rats exhibited

spontaneous activity at a rate greater than 0.1 Hz in comparison to only 17% of

characterised neurones in sham-operated rats and 25% of naive.

5 .3 .3 S p in a l C o r d E l e c t r o p h y s i o l o g y : E f f e c t s o f e t h o s u x im id e

The effect o f ethosuximide (5 - 1055 p.g), applied directly onto the spinal

cord, on the electrically- and naturally-evoked dorsal horn neuronal responses was

tested in SNL, sham-operated and naïve animals. Since the sham-operation is the

appropriate control for SNL, and for clarity, the results obtained from the naïve, non­

operated, group shall not be displayed on the graphs. However, no difference in the

effects o f ethosuximide was observed between sham and naïve groups. Ethosuximide

produced a dose-related inhibition of the electrically- and naturally-evoked responses

in neurones in all experimental groups (Figures 1 9 -2 1 ) and there was no difference

in the extent of its effects between groups. Effects were seen around 40 minutes, with

maximal inhibitions established at around 60 minutes. For sham and SNL groups, all

doses of ethosuximide (5 - 1055 pg) elicited statistically significant inhibitions of

the electrically-evoked responses, compared to pre-drug control values (Figure 19; P

< 0.05; n = 6 - 11). For naïve animals statistically significant inhibitions were elicited

by ethosuximide: at all doses (5 - 1055 pg) for the A|3-fibre, Aô-fibre, postdischarge

and excess spikes measurements; at 55 - 1055 pg ethosuximide for the C-fibre

response; and at 555 and 1055 pg for the input response {P < 0.05; n = 7). In all 3

130

«>00

80 -

o 60 -

1 4 0 -c

^ 2 0 -

0

-20 ^

Post-discharge Excess spikes1 0 0 -

80 -

J 40 H c

I10

—1------- 1 1100 1000 10000

20 i

0

-2 0 -J

Ethosuximide (jig)

1 1 1 110 100 1000 10000

Ethosuximide (p.g)

InputlOOilOOi

8 0 - 8 0 -

6 0 -6 0 -

4 0 -^ 4 0 -

f 2 0 -

0 ----

-20 J1000 00001 10 100 1000 10000

Ethosuximide (pg) Ethosuximide (pg)

Ad-fibre AB-fibrelOOi lOOl

8 0 - 80 - Shamo 60 -

4 0 -

60 - SNL

;P20-

0 ---

10 100 1000 Ethosuximide (pg)

10 100 1000 10000 Ethosuximide (pg)

Figure 19. Comparison o f the effect o f spinally applied ethosuximide on the electrically-

evoked dorsal horn neuronal responses recorded from sham-operated (n = 6) and SNL

(n = 6 - 11) rats at PO day’s 14 - 17. Data are expressed as the mean maximal %

inhibition o f the pre-drug control values ±S.E.M.

131

Sham8 0 n

6 0 -

o 4 0 -

< 2 0 -

00 2 4 6 8 10 12 14 16 18

b1 0 0 -

80 -CO2c<D 6 0 -o

c_o 4 0 -

o<

2 0 -

0 -

Stimulus Number

SNL

~ i 1 r2 4 6

n— I— I— I— I— I 8 10 12 14 16 18

-O— Pre-drug control

-A— Ethosuximide 5 jLtg

-O— Ethosuximide 55 jig

-V— Ethosuximide 555 jiig

Ethosuximide 1055f4g

Stimulus Number

Figure 20. Examples of the inhibitory effect of spinally applied ethosuximide on

individual dorsal horn neurones exhibiting wind-up recorded from (a) a sham-

operated and (b) a SNL rat. Data are expressed as the number of evoked action

potentials, in the 90 - 800ms latency band, to each electrical stimulus.

132

10(h

80 -

60-

c 40-

20-

01 10 100 1000 10000

lOOl

80-1

o;-5 60H

- 40-

204

0

Ethosuximide (^g)

Heat

vF 75g100-1

8 0 -

60-

2 0 -

010 100 1000 100001

Ethosuximide (p,g)

—r~10

—I------- 1-------- 1100 1000 10000

Ethosuximide (pig)

- Sham

—• —- SNL

Figure 21. Comparison of the effect of spinally applied ethosuximide on the naturally-

evoked dorsal horn neuronal responses recorded from sham (n = 6) and SNL (n = 6 -

11) rats. Data are expressed as mean maximal % inhibition o f pre-drug control values ±

133

groups the AP-fîbre response was least affected with mean maximal inhibitions at

top dose in the range 17 ± 6% to 26 ± 8% (Figure 19f). In all 3 groups the C-fibre

and Aô-fibre responses reached similar mean maximal inhibitions at top dose in the

range 33 ± 13% to 47 ± 17% (Figures 19d & 19e, respectively). Greater inhibitory

effects over the entire ethosuximide concentration range were observed on the input,

postdischarge and excess spikes measurements (Figures 19a - c, respectively). At

1055 |xg, ethosuximide maximally inhibited the input and excess spikes to within the

range 48 ± 8% to 61 ± 14%. The greatest effect was observed on postdischarge,

which was maximally inhibited in SNL, sham and naïve rats over the range 58 ±

10% to 75 ± 13%.

Figures 20a and 20b are examples of the effects o f ethosuximide on the

wind-up o f an individual neurone recorded from a sham and SNL rat, respectively.

These particular neurones were selected as those displaying a clear wind-up response

i.e. an increasing number of evoked action potentials (300 ms - 800 ms latency band)

observed with increasing electrical stimulus number, and the drug effects are

consistent with the rest o f the neurones tested. It is apparent that ethosuximide

reduces the wind-up, in a dose-related manner (evident by a flattening of the curve),

with a much less pronounced effect on the initial C-fibre input (response to stimulus

number 1) and no difference was observed between experimental animal groups.

Ethosuximide also produced an inhibitory effect on the naturally-evoked

neuronal responses (Figure 21). In SNL animals this was significant (P < 0.05; n = 6

- 11) for all concentrations of ethosuximide employed in this study on the response

to non-noxious mechanical stimulation (von Frey 9 g. Figure 21a), noxious

mechanical stimulation (von Frey 75 g. Figure 21b) and noxious thermal stimulation

(water jet at 45 °C, Figure 21c). In sham rats the von Frey 9 g and heat responses

were significantly inhibited at all concentrations o f ethosuximide, and the von Frey

75 g response was inhibited by 555 and 1055 jig (P < 0.05; n = 5 - 6). In naïve

animals responses to both von Frey filaments were significantly inhibited by all

concentrations o f ethosuximide, and inhibition o f the heat response reached

significance at 555 and 1055 ]Lig (P < 0.05; n = 5 - 7). No differences in the effects of

the drug were apparent between experimental groups and the mean maximal

134

inhibitions established at top dose were in the range 61 ± 11% to 77 ± 10%.

5 .4 D isc u ss io n

In this study I have addressed the role of the Cay3 family of VDCCs (LVA

T-type Ca " channels) in the spinal processing of sensory information after nerve

injury. Spinal ethosuximide, a relatively specific T-type channel antagonist, mediated

significant inhibition o f the electrically- and naturally- (innocuous and noxious)

evoked rat dorsal horn neuronal responses, suggesting some role for T-type Ca

channels in sensory transmission.

Extensive behavioural and electrophysiological studies into nociception

have demonstrated an important role for HVA Ca channels in the processing of

pain (see Vanegas & Schaible, 2000). The contribution of N-, P/Q- or L-type

channels appears to alter depending on the nature o f the pain (acute or chronic,

inflammatory or neuropathic in origin), and this has been established by the use of

specific channel antagonists (see also chapters 4 and 6). A predominant nociceptive

role for N-type channels has been shown, an action that is enhanced after neuropathy

(Chaplan et al., 1994; Xiao & Bennett, 1995; Bowersox et al., 1996; Brose et al.,

1997; White & Cousins, 1998). Upon substantial membrane depolarisation N- and

P/Q-type VDCCs mediate the release o f excitatory neurotransmitters, such as

glutamate, substance P and CGRP, critical for wind-up and central sensitization, in

the presence o f constant afferent input (Dickenson, 1994).

This study highlights the role of Ca "*" influx via T-type channels in these

nociceptive pathways, on which ethosuximide may be exerting its effects at a

number o f possible sites. The biophysical characteristics of the Cay3 VDCCs have

led to their main implication being in the regulation o f cell excitability. The broad

action of ethosuximide on all the dorsal horn neuronal responses observed here, with

no marked selectivity for a particular modality or evoked response, is in keeping with

actions on postsynaptically located Ca " channels. A postsynaptic effect would

equally effect all responses due to convergence o f differential afferent inputs that

occurs at this level. Since LVA Ca^ channel activation occurs close to resting

potential, Ca influx is permitted when cells are at rest (Magee & Johnston, 1995) or

135

in response to subthreshold synaptic inputs (Markram & Sakmann, 1994; Magee &

Johnston, 1995). Thus these channels enhance neuronal excitability and contribute to

the generation of subthreshold membrane potential oscillations that lead to bursts of

Na^-dependent action potentials (Huguenard, 1996). Whilst unable to mediate

synaptic transmission alone, T-type channels do serve to boost synaptic inputs and

potentiate the generation o f high-threshold action potentials. Their block would

result in an overall reduction in the underlying level o f neuronal excitability. This

would render the achievement o f levels of membrane depolarization at ion channel-

activating thresholds less likely, such that action potential production would be

reduced. Postsynaptically, NMDA receptor activation would be decreased, as would

the consequential development of central sensitization.

In this study, ethosuximide exerted its greatest inhibitory effects upon the

input, postdischarge and excess spikes. Each of these electrical measures can be

related to a specific part o f the nociceptive pathway. The non-potentiated input

response can be related to the baseline level o f C-fibre mediated synaptic

transmission between the central terminals of primary afferents and the neurones of

the spinal cord dorsal horn, in the absence of NMDA receptor mediated wind-up

events. Thus blockade of T-type channels at this location is likely causing a

reduction in the exocytosis o f excitatory neurotransmitters by prohibiting the

depolarization required for activation o f HVA Ca " charmels. Alternatively,

ethosuximide could be exerting its effects directly on neurones located early in

polysynaptic spinal pathways. Even more susceptible to the actions o f ethosuximide

were the postsynaptic NMDA receptor-mediated postdischarge and excess spike

measurements, indicative o f central sensitization and neuronal hyperexcitability.

Since T-type Ca " channels are heavily linked to the level of neuronal excitability it

follows that they would have a greater functional role here (see Figure 18).

The results presented here show that after the establishment o f neuropathy

no difference was detected in the effects o f ethosuximide. This appears somewhat

surprising especially since neuronal hyperexcitability, in which LVA Ca " channels

are implicated, is a key underlying factor. Interestingly it has been shown that

although unilateral cortical ablation causes a 68% increase in T-current measured

136

from isolated rat thalamic relay neurones, methy 1-phenyl-succimide (another related

Cav3 VDCC antagonist) was more effective in reducing T-current in normal rats than

in axotomized animals (Chung et a l, 1993). The authors suggest an injury-induced

alteration in the pharmacological properties the Cay3 VDCCs either by de novo

synthesis and/or modification. In the present study, it may be that there is indeed no

increase in the functional role o f LVA Ca^ channels after nerve injury.

Alternatively, the specificity and/or potency of ethosuximide may be such that subtle

differences in Cay3 VDCC function were not highlighted.

Peripheral nerve injury results in reduced afferent input via L5 and L6

spinal nerves, yet as observed here, the magnitude of neuronal responses recorded

was not diminished in comparison to sham and normal rats. Conversely, increased

frequency and occurrence of spontaneous activity was observed. This suggests that

perhaps compensatory increases in peripheral and/or spinal neuronal activity come

into play after neuropathy (see chapter 3). Ectopic C-fibre activity originating within

DRG, the nerve injury site, or residual intact afferents provides an ongoing, continual

barrage o f neuronal activity via primary afferents into the spinal cord (Wall & Devor,

1983; Kajander et a l , 1992). Evidence suggests that this is one possible source for

the initiation and maintenance o f central sensitization and therefore the positive

sensory symptoms observed after nerve damage (Devor & Seltzer, 1999).

Interestingly, SNL has been shown to increase the prevalence o f subthreshold

membrane potential oscillations in DRG neurones which augments ectopic discharge

(Liu et a l , 2000). It does not seem unreasonable to postulate a role for Cay3 VDCCs

in this underlying oscillatory behaviour, which may imply increased activity via

these channels in the remaining afferents tantamount to a restoration of excitability to

that in the non-injured situation. However, the spinal application o f ethosuximide

used in this study would not inhibit any T-type mediated oscillatory behaviour at the

level of the DRG.

Another recent study also indicates a role for peripheral T-type VDCCs in

nociceptive pathways (Todorovic et a l, 2001). In isolated rat sensory neurones

shown to participate in thermal nociception, reducing agents have been demonstrated

to augment T-currents, but not HVA Ca " currents, and Na" currents, whereas

137

oxidizing agents decreased LVA Ca current. A T-type VDCC antagonist mibefiradil

(Todorovic & Lingle, 1998), the first member of a relatively new class of VDCC

antagonists to be clinically approved for the treatment o f hypertension, abolished the

enhanced DRG T-currents, leaving baseline currents intact. In the same study, this

was mirrored in vivo, where peripheral injection of the endogenous reducing agent L-

cysteine produced thermal hyperalgesia that was blocked by peripherally applied

mibefiradil, but not NMDA receptor blockers. Redox sites in sensory neurones could

involve any o f the many L-cysteine residues present in putative extracellular regions

of T-type VDCCs (Cribbs et ah, 1998; Perez-Reyes et al., 1998) and could be targets

for agents that modify pain perception.

The existence o f a neuronal Ca " current elicited just above the resting

potential was first established in primary sensory neurones (Carbone & Lux, 1984;

Bossu et al., 1985; Fedulova et al., 1985; Nowycky et al., 1985), and LVA currents

have since been observed in a wide variety o f cell types (Huguenard, 1996). The

presence o f a relatively large T-type current in some superficially located rat spinal

dorsal horn neurones is o f much interest because this region is involved in processing

and integration of sensory information, including pain (Ryu & Randic, 1990). T-type

channels are pharmacologically and physiologically heterogeneous (Akaike, 1991;

Huguenard, 1996; Tarasenko et al., 1997), which may reflect differential expression

of the three known subtypes Cay3.1, Cay3.2 and Cay3.3 (ttio, (%iH and «n ,

respectively). The regional and cellular distribution o f gene expression for the

different Cay3 VDCC family members in the rat central and peripheral nervous

systems has recently been determined using in situ hybridization and these channels

appear to have unique and relatively non-overlapping distributions (Talley et al.,

1999). All three transcripts were detected in sensory areas and in the dorsal horn of

the spinal cord where in particular am was mainly restricted to the outermost

laminae I and II. In the DRG high levels o f am and moderate levels o f an mRNA

were found, and am has been shown to display the most similar pharmacological

profile to native DRG T-type currents (Todorovic et al., 2000). The presence o f T-

type VDCCs in DRG determined by in situ hybridization appeared to be restricted to

small and medium sized neurones, whereas the extremely large cells were not

labelled. This correlates with substantial T-type current observed in medium

diameter DRG neurones isolated fi-om adult rats that is absent in larger DRG cells

138

(Scroggs & Fox, 1992a). Since DRG cell body diameter is correlated to axon

conduction velocity and sensory modality (Yaksh & Hammond, 1982), this evidence

is indicative o f T-type current specifically localized to smaller A5- and C-type

sensory neurones that convey thermal and nociceptive information. In contrast larger

AP-type neurones that subserve tactile and proprioceptive pathways do not appear to

possess T-type current. In the present study the extent o f inhibition observed with the

highest dose of ethosuximide was Aô-fibre > C-fibre > Ap-fibre, which fits well with

these studies. However, the Aô- and C-fibre responses were not markedly inhibited

over the AP-fibre response, as one might expect if smaller diameter sensory neurones

exhibited substantial LVA Ca current. Additionally, no difference in the extent of

inhibition was observed for the evoked responses to innocuous and noxious

mechanical and thermal stimuli, which are conveyed to the spinal cord by the

different primary afferent fibre types. This may again be explained by the existence

of the three different ai subunits that comprise the Cay3 VDCCs , each showing a

unique neuronal distribution.

Collation of these observations suggests different neuronal populations that

have alternative patterns o f T-type channel localization: cells in which functional

channels are localized in the somatic membrane and others in which functional

channels are only expressed dendritically. Electrophysiological studies o f dissociated

neurones allow robust recordings to be made only from cell soma and proximal

processes. Immunohistochemistry (Craig et al., 1999), Ca imaging (Christie et al.,

1995) and electrophysiological studies (Kavalali et al., 1997; Mouginot et al., 1997)

suggest that in certain neurones, LVA channels have a dendritic location. A

predominantly dendritic expression of T-type channels has been described for

hippocampal neurones (Karst et al., 1993).

Progress in revealing the differential physiological roles of the Cay3 family

of VDCCs have been impeded by the lack o f ligands that distinguish not only

between LVA and HVA VDCCs but also between C a y 3 .1 , C a y 3 .2 and C a y 3 .3 . The

Cay3 VDCCs are not inhibited by many types of antagonists, and those that do, lack

specificity. For example in DRG neurones mibeffadil has been shown to be one of

the best ligands for T-type channels with an IC 50 o f 3 p.M (Todorovic & Lingle,

139

1998). However mibefradil also blocks VDCCs of the Cayl (Classer, 1998) and Cay2

(Bezprozvanny & Tsien, 1995) families. Variable sensitivity o f T-type Ca channels

to dihydropiridines (DHPs) has also been described. This ranges from quite high

LVA Ca " current DHP sensitivity in cerebellar Purkinje (Kaneda et a l, 1990),

hippocampal (Takahashi & Akaike, 1991) neurones with K<i’s o f around 10 piM, to

virtually DHP insensitive in primary sensory neurones (see Kostyuk, 1999).

T-type Ca " current is also reported to be sensitive to block by volatile

anaesthetics (Todorovic & Lingle, 1998; McDowell et a l , 1999). It has been shown

that isofluorane is 3.5 times more potent than halothane in blocking rat DRG T-

current which parallels the observation that isofluorane is more analgesic (Levine et

a l, 1986) and more effective at inhibiting somatosensory evoked potentials (Peterson

et a l , 1986) than halothane. Whether anaesthetic effects are in part mediated via T-

type VDCCs remains to be determined, however this may be indicative of a potential

role for T-type Ca channels in sensory transmission.

As regards ethosuximide, its current clinical use for the treatment of

epilepsy is via block of T-current in thalamic neurones (Coulter et a l , 1989c; Coulter

et a l , 1989b). In contrast, the T-current in GH3 cells is relatively resistant to block

by ethosuximide (Herrington & Lingle, 1992) and in DRG neurones ethosuximide

has been shown to be over an order o f magnitude less effective (Coulter et a l ,

1989c; Coulter et a l , 1989b). However, the blockade o f T-current mediated by

ethosuximide observed in DRG is complete, but only partial in thalamic neurones. A

preference o f ethosuximide for T-type over L-type Ca " channels in terms of both

efficacy and affinity has also been reported for isolated DRG neurones (Kostyuk et

a l, 1992), specificity (Kj 7 p.M and 15 )liM, respectively) being preserved up to

concentrations o f ImM. Furthermore, in thalamic neurones ethosuximide has been

shown to inhibit T-type currents by about 40% with an E C 50 of 200 p,M, with no

effects on HVA currents (Coulter et a l, 1989b). Thus the data suggests that diversity

exists between T-currents of different cell types both in terms o f kinetics and

pharmacological sensitivity (see also Huguenard, 1996). It is not yet known whether

ethosuximide blocks all members o f the Cay3 VDCC family equally or whether it

displays subtype selectivity. It must be taken into consideration that a io is the

140

predominant subtype found in thalamic relay neurones, the target site o f

ethosuximide for the treatment of absence epilepsy. Therefore the possibility exists

that Cav3.1 (a io )is more sensitive to the effects o f ethosuximide in comparison to

the Cav3.2 (am ) VDCC which is more abundant than a io in the outer lamina o f the

spinal cord and dorsal root ganglia (Talley et al., 1999). The synthesis of ligands able

to distinguish between the LVA Ca " channels, in particular the am subtype which

anatomically appears to be implicated in nociceptive pathways, is eagerly awaited.

Until then the role o f T-type Ca "*" channels in nociception cannot be truly

established. The results from the present study may have underestimated the

importance o f LVA VDCCs in sensory transmission as ethosuximide may not have

been a sensitive enough pharmacological tool. Nevertheless, I deemed ethosuximide

the best pharmacological option available, over other known T-type VDCC

antagonists, to study the role of LVA Ca channels within the pain pathway, due to

its reported selectivity for LVA over HVA VDCCs (Coulter et al., 1989b; Kostyuk et

al., 1992) and its ability to completely block T-type current in DRG neurones

(Todorovic & Lingle, 1998).

An alternative explanation for the limited effect o f ethosuximide on evoked

dorsal horn neurone responses and the lack of increased effects after nerve injury

could be due to the fact that T-type channels are rapidly inactivating and require

hyperpolarized membrane potentials for removal o f inactivation (Coulter et al.,

1989a; Toselli & Taglietti, 1992). Since during repetitive stimulation, postsynaptic

dorsal horn neurones would be rapidly depolarized, an effect likely enhanced after

nerve injury due to the ensuing hyperexcitability, T-type VDCCs at this location are

maybe largely inactivated and thus the inhibitory effects mediated by ethosuximide

would be relatively redundant. On a similar note, in 5 - 10 day old rat DRG sensory

neurones, PKC activation selectively inhibited low-threshold T-type Ca current,

present only on a subpopulation of neurones, with no effect on high threshold current

(Schroeder et al., 1990). Thus, activation o f PKC could diminish excitability of some

sensory cells while leaving others unaffected, providing the potential for selective

modulation of particular sensations. This is contrast to the Ca current-enhancing

effects of PKC activation observed through N-type VDCCs (see chapter 4.4). Given

the increased role of PKC activity in mediating neuropathic pain behaviours and the

antinociceptive effect exerted through its inhibition, it follows that in the presence of

141

neuropathy T-type VDCCs on some sensory neurones would actually be inhibited.

Again this supports my finding of an unaltered role o f T-type VDCCs after nerve

injury, as observed through the actions of ethosuximide. However, on the basis of the

PKC evidence a decreased inhibitory effect o f ethosuximide on evoked neuronal

responses might have been expected.

Two other licensed anticonvulsants, carbamazepine and gabapentin, have

been investigated using the same experimental protocol as the present study

(Chapman et al., 1998a). Carbamazepine, a Na" channel blocker, and gabapentin,

thought to act via Ca channels (see chapter 7), were found to have similar efficacy

and range o f effectiveness as ethosuximide observed here. Although both gabapentin

and ethosuximide were equally effective in sham and neuropathic animals,

carbamazepine was only effective in the latter group which could possibly be a

consequence of differential regulation of particular Ca " and Na^ channels following

nerve injury. To my knowledge the present study is the first to demonstrate a

possible role of T-type Ca channels in the spinal processing of sensory information

related to pain. Given the parallels between epilepsy and pain, the likelihood of

common causal mechanisms and the ability o f antiepileptic drugs to be effective in

neuropathic pain states, the results indicate that ethosuximide may merit both

behavioural testing in animals and human studies.

142

CHAPTER 6

VOLTAGE-DEPENDENT CALCIUM CHANNELS:

THE CAvl FAMILY

143

6 .1 I n t r o d u c t io n

High voltage-activated (HVA) Ca " channels, o f the Cayl (L-type) and Cay2

families (N-, P/Q- and R-types) are activated by relatively strong membrane

depolarization and permit Ca influx in response to action potentials. Consequential

secondary actions include neurotransmitter release, thus these channels establish a

major link between neuronal excitability and synaptic transmission. On the other

hand, the low voltage-activated (LVA) Ca channels of the Cay3 (T-type) family are

activated at voltages close to resting potential. These VDCCs permit Ca " influx

when cells are at rest, thus regulating cell excitability and most likely the

depolarization required to activate HVA VDCCs. Extensive behavioural and

electrophysiological studies into nociception have demonstrated an important role in

general for VDCCs in the processing of pain see chapters 4 and 5 and (Vanegas &

Schaible, 2000) for review. The employment of specific Ca channel blockers and

antagonists has revealed that the contribution of L-, N- and P/Q- type channels

differs dependent on the nature o f pain (acute or chronic, inflammatory or

neuropathic in origin). Results from the previous chapters, in conjunction with other

studies, highlight a predominant pronociceptive role for N-type channels, which is

enhanced after neuropathy (Chaplan et al., 1994; Bowersox et a l , 1996; Brose et al.,

1997; White & Cousins, 1998).

Pre-clinical studies provide varied conclusions regarding the antinociceptive

effects of L-type VDCC blockade. The general consensus is that inhibition o f Ca "

influx through these channels has little or no antinociceptive effect. For completeness

of this thesis I thought it necessary to establish the role of the L-type VDCCs in the

spinal transmission of sensory information and assess alterations that might occur in

their functioning following peripheral nerve injury, given the plasticity I discovered

in N-type VDCCs at this level and the roles of the P- and T-type channels in sensory

events. Of the four members that comprise the family o f L-type or Cayl VDCCs

(Ellis et ah, 1988; Williams et ah, 1992; Tomlinson et ah, 1993; Bech-Hansen et ah,

1998) see also chapter 1.4, it is the Cay 1.2 (aic) and Cayl.3 (a ^ ) channels that are

located in the CNS. In rat brain, the a ic subunit (first defined in cardiac muscle) and

the aiD subunit (first defined in endocrine cells) are found throughout the brain (Hell

et ah, 1993), where they are important for regulation of gene expression and the

144

integration o f synaptic inputs (Bean, 1989b; Bean, 1989a; Murphy et ah, 1991;

Bading et al., 1993; Bito et aL, 1996; Deisseroth et a l, 1996). Immunohistochemical

studies o f the rat spinal cord have revealed staining o f these channels scattered

throughout the entire dorsal horn, localized predominantly to the cell bodies and

proximal dendrites o f neurones (Ahlijanian et a l , 1990; Hell et a l , 1993;

Westenbroek ef ût/., 1998).

L-type VDCCs are selectively blocked by 1,4-dihydropyridines (DHPs),

phenylalkylamines and benzothiazepines (see Tsien et al., 1988; Spedding &

Paoletti, 1992). DHPs, such as nifedipine, are a class of synthetic Ca channel

blockers used for treatment o f hypertension, due to their depressive action on the

cardiovascular system. Specifically they prevent Ca "*" entry through VDCCs on

vascular smooth muscle, reducing contraction which then leads to vasodilation and

thus they lower arterial pressure. The present study uses both naïve and SNL animals

to investigate the effects o f spinally delivered nifedipine on a wide range of

electrically- and naturally-evoked dorsal horn neuronal activity.

6 .2 M e t h o d s

A total o f 12 male adult Sprague-Dawley rats were used in this study, of

which 6 were naïve, and 6 were SNL. The animals were prepared in the normal way

and the effect o f nifedipine (10, 110 and 260 |Lig), applied directly onto the exposed

surface o f the spinal cord, was investigated on the evoked dorsal horn neuronal

responses. The effects o f each dose were followed until the responses plateaued (a

minimum of 60 minutes), when the next dose would be applied cumulatively.

6.3 R e s u l t s

6.3.1 B e h a v i o u r a l A s s e s s m e n t

Rats subjected to SNL exhibited 'guarding' foot posture ipsilateral to nerve

injury and successful replication o f the nerve injury model was confirmed by

behavioural testing which demonstrated the development of mechanical and cooling

allodynia o f the injured hindpaw (see chapter 3 for combined behavioural data

145

obtained from all studies used in this thesis). This was not displayed by the

contralateral hindpaw. Upon establishment of a neuropathic state, animals were then

used for in vivo electrophysiology and the pharmacological study at PO days 14 - 17.

6 .3 .2 S p in a l C o r d E l e c t r o p h y sio l o g y : N e u r o n e C h a r a c t e r iz a t io n

The number of dorsal horn neurones recorded from in each group were 6 in

SNL rats and 6 in naïve rats. All neurones had a receptive field over the hindpaw

ipsilateral to surgery (where performed). No significant differences were found

between experimental groups in the mean cell depth o f recorded neurones and the

mean neuronal responses evoked by electrical and natural stimulation. However, 2

out o f the 5 neurones characterised in SNL rats exhibited spontaneous activity at a

rate greater than 0.9 Hz whereas none of the neurones characterized from naïve

animals showed any spontaneous activity.

6 .3 .3 S p in a l C o r d E l e c t r o p h y sio l o g y : E f f e c t s o f n if e d ip in e

The effect of nifedipine (10 - 260 p.g), applied directly onto the spinal cord,

on the electrically- and naturally-evoked dorsal horn neuronal responses was tested

in SNL and naïve animals. Although the sham-operation is the appropriate control

for SNL, results obtained from the naïve, non-operated group and the SNL group

show that nifedipine exerted only limited inhibitory effects on the evoked neuronal

responses. Furthermore, no difference in the effects o f nifedipine was observed

between these experimental groups, thus it was deemed unnecessary and wasteful to

perform these experiments upon sham-operated rats. This is supported by the results

from chapter 5 where no difference in the effects o f ethosuximide was observed

between sham and naïve groups, nor between SNL and either sham or naïve animals.

Nifedipine produced limited inhibitions of the electrically- and naturally-evoked

responses in neurones in naïve and SNL rats (Figures 22 - 24), and there was no

difference in the magnitude of its effects between groups. For measurements where

nifedipine-mediated inhibition was more marked, its effects were dose-related and

seen around 60 - 90 minutes. Nifedipine was dissolved in mixture o f cremophor,

saline and ethanol, due its insolubility in saline alone (see chapter 2). No vehicle

control was performed in this series o f experiments since it has already been

146

a Post-dischargelOOnlOOn

80-80-

o 60-2 60-

40-

-20-J100 100010 100 1000

Nifedipine (pg)

Input

Nifedipine (pg)

C-fibre100-1

80-80-

o 60-.2 60-

4 0 -

0 —0 —

- 2 0 - 'TTH]1000 10 100 1000100

Nifedipine (pg) Nifedipine (pg)

Ad-fibre AP-fibreOOiOOi

80-80- Naive

o 60-60- SNL

4 0 -

0 - —

- 20-1100010 100

Nifedipine (pg)10 100

Nifedipine (pg)1000

Figure 22 Comparison of the effect of spinally applied nifedipine on the electrically-

evoked dorsal horn neuronal responses recorded from un-operated, naive (n = 6) and

SNL (n = 5) rats at PO days 14 - 17. Data are expressed as the mean maximal %

inhibition of the pre-drug control values ± S.E.M.

147

demonstrated to be without effect (Stanfa et al, 1996). For naive animals the highest

dose o f nifedipine used, 260 \xg (dictated by its poor solubility), produced

statistically significant inhibitions, compared to pre-drug control values, on the

electrically-evoked responses, postdischarge, excess spikes, input and C-fibre

(Figures 22a - e, respectively; P < 0.05; n = 6). Additionally, the input response in

naïve animals was also significantly inhibited at the middle dose, 110 |Lig nifedipine.

In contrast, for the SNL group significant inhibition from pre-drug control value was

only seen after the top dose of nifedipine for the input measurement (P < 0.05; n =

6). In both groups the different fibre type-evoked responses (Ap, Aô and C) were

least affected with mean maximal inhibitions at top dose in the range 6 ± 21% to 38

± 12% (Figures 22d - f, respectively). Greater inhibitory effects of nifedipine were

observed on the postdischarge, excess spikes and input measurements (Figures 22a -

c, respectively). At 260 pg, nifedipine maximally inhibited these measures within the

range 32 ± 12% to 56 ± 12%. It should also be noted that low doses of nifedipine

mediated facilitation of the electrically-evoked responses in the naïve group (2 out of

6 cells) and in the SNL group (2 out o f 5 cells), for the Aô-fibre, C-fibre,

postdischarge, input and excess spike measurements. This explains the large error

bars displayed on some of the graphs. These excitatory effects were later inhibited at

the top dose, 260 pg nifedipine.

Figure 23 is an example of the effects of nifedipine on the wind-up o f an

individual neurone recorded from a SNL rat. This neurone was selected as one

displaying a clear wind-up response i.e. an increasing number of evoked action

potentials (300 ms - 800 ms latency band) observed with increasing electrical

stimulus number, and the drug effects are consistent with the rest o f the neurones

tested. It is apparent that nifedipine reduces the wind-up, in a dose-related manner

(evident by a flattening of the curve), with a much less pronounced effect on the

initial C-fibre input (response to stimulus number 1) and no difference was observed

in comparison to neurones recorded from naïve rats.

148

C /D

[gC0 c-

1o<

120-,

100 -

80 -

60 -

40 -

20 -

00 2 4 6 8 10 12 14 16 18

-O— Pre-drug control

-A— Nifedipine 10 pg

-O— Nifedipine 110 pg

-V— Nifedipine 260 pg

Stimulus Number

Figure 23. Example o f the inhibitory effect o f spinally applied nifedipine on an

individual dorsal horn neurone exhibiting wind-up recorded from SNL rat. Data is

expressed as the number o f evoked action potentials, in the 90 - 800 ms latency band, to

each electrical stimulus.

Nifedipine also produced an inhibitory effect on the naturally-evoked

neuronal responses (Figure 24). This was only found to be significant in SNL

animals at all doses for the noxious heat-evoked response (P < 0.05; n = 5). No

differences in the effects of the drug were apparent between experimental groups and

the mean maximal inhibitions established at top dose were in the range 32 ± 14% to

74 ± 8%. Unfortunately, for the limited amount of cells studied in this investigation,

difficulties were encountered in locating dorsal horn neurones in SNL rats that

responded to the innocuous von Frey 9 g mechanical stimulation (1 out of 5 cells).

The results from this one cell have been displayed on the graph (Figure 24a), since

the effects of nifedipine upon the von Frey 9 g evoked response o f this neurone were

similar to that obtained in the naïve group, which is in keeping with the findings

from the rest o f the study.

149

a vF 9g vF 75g100-1

80-

6 0 -

2 0 -

0101 100 1000

_cc

lOOi

8 0 -

6 0 -

4 0 -

20 -

010 100 10001

N ifed ip ine (pg)

H e a t

N ifed ip ine (pg)

100-

80-co

T#-1

>5 60 -rC I

4 0 -o-

i

2 0 -

0 - r T—I r-n m-r-

N aive

10 100 1000 N ifed ip ine (pg)

Figure 24 Comparison o f the effect o f spinally applied nifedipine on the naturally-

evoked dorsal horn neuronal responses recordedfi^om un-operated, naive (n = 5) and

SNL (n = 5) rats. (Note for SNL rats n = 1 for the von Frey 9 g response). Data are

expressed as mean maximal % inhibition o f pre-drug control values ± S. E.M.

150

6 .4 D is c u s s io n

In this study I have addressed the role of the Cayl family of VDCCs (HVA

L-type Ca channels) in the spinal processing of sensory information after nerve

injury. Spinal nifedipine, a relatively specific L-type VDCC antagonist, mediated a

small inhibition of the electrically- and naturally- (innocuous and noxious) evoked

rat dorsal horn neuronal responses, significant only at the highest dose used for

measures predominantly related to dorsal horn neuronal excitability. These results

are suggestive of a limited role o f L-type Ca " channels in sensory transmission at

the spinal cord level, in a manner unaltered by nerve injury.

Extensive behavioural and electrophysiological studies into nociception

have demonstrated an important role in general for HVA Ca " channels in the

processing of pain see chapters 4 and 5 and (Vanegas & Schaible, 2000) for review.

The employment of specific Ca " channel blockers and antagonists has revealed that

the contribution of L-, N- or P/Q- type channels differs dependent on the nature of

pain (acute or chronic, inflammatory or neuropathic in origin). A predominant

nociceptive role for N-type channels has been established, which is enhanced after

neuropathy (Chaplan et ah, 1994; Xiao & Bennett, 1995; Bowersox et ah, 1996;

Brose et ah, 1997; White & Cousins, 1998). Upon substantial membrane

depolarisation N- and P/Q-type VDCCs mediate the release o f excitatory

neurotransmitters, such as glutamate, substance P and CGRP, critical for wind-up

and central sensitization, in the presence o f constant afferent input (Dickenson,

1994). On the other hand, the LVA T-type Ca channels are activated close to

resting potential. These VDCCs permit Ca^ influx when cells are at rest thus

regulating cell excitability and most likely the depolarization required to activate

HVA L-, N- and P/Q-type channels, the latter two Cay2 VDCCs strongly implicated

in neurotransmission.

L-type VDCCs are found within the sensory transmission pathways from

periphery to the spinal cord to the brain. Immunohistochemical studies have

demonstrated the neuronal presence o f a ic and aiD subunits upon the cell bodies and

dendrites in a variety of cell populations, including the spinal cord, where they are

distributed throughout the deep dorsal and ventral horns (Ahlijanian et ah, 1990; Hell

151

et ah, 1993; Westenbroek et ah, 1998). Further evidence cornes from the

observations that nifedipine reduced the sustained component o f HVA Ca " current

in rat dorsal horn neurones (Ryu & Randic, 1990) and that L-type VDCCs comprise

50% of the whole Ca current of small diameter DRG neurones, much greater than

that observed in large diameter cell bodies (Scroggs & Fox, 1992a),

It is, however, generally accepted that the Cay2 family o f VDCCs are

responsible for depolarization-coupled neurotransmitter release (Miljanich &

Ramachandran, 1995), with no contribution to this physiological function from the

Cayl family. However, some reports have demonstrated that Ca current through L-

type channels is linked to the endogenous release of glutamate in rat hippocampus

(Terrian et ah, 1990) and substance P release from sensory neurones (Pemey et ah,

1986; Rane et ah, 1987). These relatively few studies are somewhat discredited by

the overwhelming evidence to the contrary and L-type VDCCs are generally thought

not to be directly related to synaptic transmission in the dorsal horn. In most

investigations DHPs have no effect on neurotransmitter release (Holz et ah, 1988;

Kamiya et ah, 1988; Takahashi & Momiyama, 1993; Wheeler et ah, 1994; Wu &

Saggau, 1994). Release of CGRP from rat spinal afferents has been demonstrated not

to be mediated by nifedipine-sensitive L-type VDCCs, in contrast to the key roles of

(o-conotoxin GVIA sensitive channels (Santicioli et ah, 1992). Furthermore, in rat

hippocampal slices, nimodipine or nifedipine were found to reduce Ca " entry

evoked by repetitive electrical stimulation only postsynaptically, and not that

occurring presynaptically (Jones & Heinemann, 1987; Igelmund et ah, 1996). Taken

together these studies indicate that L-type VDCCs are not involved in presynaptic

Ca " entry, and subsequent neurotransmitter release, but rather to have a postsynaptic

location (see also Spedding & Paoletti, 1992). L-type VDCCs are therefore

implicated in neurone depolarization and influx of Ca " for intracellular mechanisms

linked to the activation of signal transduction cascades.

Here I have demonstrated that most susceptible to the effects of nifedipine

were the postsynaptic NMDA receptor-mediated postdischarge and excess spikes

measurements, indicators o f central sensitization and neuronal hyperexcitability.

Nifedipine potently inhibits L-type currents in a use-dependent manner. Nifedipine

152

has a low affinity for L-type VDCCs that are at rest and preferentially binds to the

inactivated state of the channel (Bean, 1984). At negative holding potentials DHPs

have little effect upon the whole cell Ca " current suggestive that L-type VDCCs are

minimally active. Upon increasing neuronal depolarization, as might occur during

the central sensitization that ensues after peripheral nerve damage or repetitive

electrical stimulation, L-type VDCCs may mediate a greater proportion of the Ca

current (Bean et ah, 1993). The voltage-dependent interaction exhibited by

nifedipine would mean that the likelihood o f its binding to and blocking L-type

VDCCs increases with increased neuronal excitability. It might be expected then that

the effectiveness o f nifedipine would positively correlate with maintained activation

of the channel due to the reported ongoing, continual barrage o f neuronal activity via

primary afferents into the spinal cord arising from ectopic C-fibre activity that

follows nerve injury (Wall & Devor, 1983; Kajander et ah, 1992). If a C-fibre

afferent barrage results in activation of VDCCs then nifedipine could reduce

postsynaptic Ca entry, which also implies that nifedipine would be less effective in

the naïve animal. However, as demonstrated here this was not the case. Peripheral

nerve injury did not induce plasticity in the events mediated by L-type VDCCs at the

level of the spinal cord. Furthermore, equally susceptible to the effects of nifedipine

was the non-potentiated input response which can be related to the baseline level o f

C-fibre mediated synaptic transmission between the central terminals of primary

afferents and the neurones of the spinal cord dorsal horn, in the absence of NMDA

receptor mediated wind-up events. Even if nerve injury did not induce changes in the

functioning of L-type VDCCs, considering the above discussion it might be expected

that this measurement would be inhibited less by nifedipine in comparison to

postdischarge and excess spikes. It could be that the depolarization mediated by the

first electrical stimulus was sufficient to maximally activate L-type VDCCs within

the sensory pathway, therefore no difference would be observed in the effects of by

L-type VDCC block in the presence or absence of NMD A-mediated wind-up.

There are a few reports that may relate to these observations, where L-type

VDCCs are implicated in regenerative Ca^^-dependent plateau potentials o f deep

dorsal horn neurones that possibly comprises another critical component o f wind-up,

operating downstream of the predominant synaptic NMD A-mediated processes

(Morisset & Nagy, 1996; Morisset & Nagy, 1998). Electrical stimulation (0.4 - 1

153

Hz) o f high-threshold primary afferent fibres in the dorsal root, induced action

potential wind-up and afterdischarge correlating with the production o f Ca^ -

dependent plateau potentials, as recorded intracellularly from deep dorsal horn

neurones in slice preparation of rat spinal cord. These phenomena were suppressed

by nifedipine indicating that Ca '^-dependent plateau potentials were involved in this

form o f wind-up (Morisset & Nagy, 2000). Being associated with Ca '*' influx,

generation o f plateau potentials could be a link between short-term plasticity and the

long-term modification of dorsal horn neuronal excitability associated with central

sensitization.

In pre-clinical acute nociception studies, L-type VDCC antagonists o f the

DHP, phenylalkylamine and benzothiazepine classes can block evoked nociceptive

responses (Del Pozo et al., 1987; Miranda et al., 1992), although these observations

have not been consistently confirmed (Hoffmeister & Tettenbom, 1986; Contreras et

ah, 1988; Omote et al., 1995). Furthermore, given the marked vascular effects o f

these antagonists any behavioural study with heat or inflammatory stimuli has to be

viewed with caution as changes in blood flow will alter heat transfer across the skin.

Intrathecal L-type VDCC antagonists, verapamil and diltiazem, have been

demonstrated to have no effect on the acute phase and produce only modest

inhibition o f the 2" , tonic phase of nocifensive behaviours induced by formalin

(Malmberg & Yaksh, 1994). In the same experimental model nifedipine was found to

mediate some antinociceptive actions, but these were much less effective than the

NMDA receptor antagonist MK-801 (Coderre, 1992). Furthermore, intrathecal

nifedipine failed to prevent or reverse thermal hyperalgesia or spontaneous

behaviours induced by a rodent model o f acute arthritis, where N- and P/Q-type

VDCC blockade was antinociceptive (Sluka, 1998). This is indicative o f no spinal

role for L-type Ca " channels in the secondary heat hyperalgesia induced by joint

inflammation. However, in one model o f inflammation intrathecal nifedipine was

shown to prevent secondary mechanical hyperalgesia and allodynia developed in

response to injection of capsaicin into the hindpaw (Sluka, 1997). In models of

neuropathic pain no role for spinal L-type VDCCs has been uncovered. Using the

SNL model of neuropathy, administration o f L-type VDCC antagonists, diltiazem,

verapamil and nimodipine, intrathecally, intravenously and to the site of injury, also

had no antinociceptive effects (Chaplan et al., 1994), nor did injection of nifedipine

154

into the hindpaw in PSL model of peripheral nerve injury (White & Cousins, 1998).

Although there is behavioural evidence regarding the role of L-type VDCCs

in neuropathy, only studies investigating acute or inflammatory pain have been

executed electrophysiologically. Recordings o f single dorsal horn neurones, in a

manner identical to that used in this study, demonstrated that verapamil was without

effect on responses to formalin injection, in contrast to intrathecal co-agatoxin IVA

and m-conotoxin GVIA which reduced the 2" phase and additionally the 1® phase,

respectively (Diaz & Dickenson, 1997). It was, however, demonstrated that L-type

VDCCs might be involved in the central sensitization o f dorsal horn neurones to

mechanical stimuli induced by joint inflammation in the model o f acute arthritis, yet

the dose administered also reduced the response o f unsensitized neurones

(Neugebauer et a i, 1996).

The data I present here is in keeping with the theme demonstrated by the

vast majority o f studies surrounding L-type VDCCs and sensory transmission.

Unlike predominantly the N-type, and to a lesser extent P/Q-type VDCCs, spinal L-

type channels do not appear to be critical to the transmission o f nociceptive

information observed both behaviourally and electrophysiologically. It could be

argued that the lack of effects of nifedipine on the evoked dorsal horn neuronal

responses in this study were due to insufficient doses o f the drug, which were limited

by its solubility. Similar problems have been encountered in other studies using

nifedipine where the maximum concentration possible was 1 mM and this required

30% DMSO to enable nifedipine to go into solution (Sluka, 1997; Sluka, 1998). One

of these studies reported comparable antinociceptive effects to capsaicin treatment

mediated by block o f spinal L- or N-type VDCCs, using intrathecal nifedipine at a

concentration 100-fold greater than that o f cü-conotoxin GVIA (Sluka, 1997).

However in a subsequent study in the formalin model the same concentration of

nifedipine was without effect (Sluka, 1998). In my study, the maximum

concentration o f spinally applied nifedipine used (15 mM) was similarly 1000-fold

greater than the concentrations of co-conotoxin GVIA required to exert maximum

effects (15 - 20 jiM) upon the evoked neuronal responses (see chapter 4). Thus it is

unlikely that an increase in dose of nifedipine would cause further inhibitory effects.

155

Nonetheless, blockade of L-type channels would thus be expected to

reduce neuronal activity and reduce the influx o f Ca " , not unlike that observed with

block of T-type channels (see chapter 5). Since intracellular Ca can activate second

messenger pathways, in addition to acting as a second messenger itself, intracellular

events downstream of synaptic transmission might be affected. Maximal effects

mediated by nifedipine decreased input, postdischarge and excess spikes

measurements by approximately half and it appears that L-type block was the basis

for some decrease in neurotransmitter release, and possibly attenuation of second

messenger systems activation and sensitization o f dorsal horn neurones. It is unlikely

however that this would be manifest behaviourally as demonstrated by other studies.

The effects of nifedipine upon the different fibre type-evoked deep dorsal horn

neuronal responses were not markedly differential to noxious or innocuous afferent

input from the periphery. Given the distribution of L-type VDCCs in the deep dorsal

horn this it not surprising. At this location Ca influx through these channels would

participate in convergent sensory pathways, not specific to nociception in particular.

Given the similarities in the effects o f nifedipine on evoked dorsal horn

neuronal responses with those mediated by block o f T-type VDCCs with

ethosuximide (chapter 5), it should be noted that variable sensitivity of T-type Ca "

channels to DHPs has also been described. However, this ranges from quite high

LVA Ca current DHP sensitivity in cerebellar Purkinje (Kaneda et ah, 1990) and

hippocampal (Takahashi & Akaike, 1991) neurones, to virtually DHP insensitive in

primary sensory neurones (see Kostyuk, 1999). Since the latter relates more to the

sensory systems investigated here, it may be that LVA VDCCs on dorsal horn

neurones within the sensory pathway may also be DHP insensitive. However the fact

that ethosuximide inhibitions were greater than those observed for nifedipine, but

with a similar profile, means that the possibility that nifedipine-mediated inhibitions

were due to inhibition of Ca " influx through LVA VDCCs, cannot be ruled out.

Although inhibition of L-type VDCC activation per se does not generally

mediate antinociception in experimental conditions (Chaplan et ah, 1994; Omote et

ah, 1995; Sluka, 1998), it has been demonstrated that blockers or antagonists may be

useful in the potentiation or restoration of opioid analgesia. This effect seems to rely

156

upon the establishment of opioid tolerance due to chronic administration, or a pain

state that is opioid resistant. A synergistic effect interaction between L-type Ca "

channels blockers and opioids under either or both o f these situations has been

consistently reported (Welch & Olson, 1991; Antkiewicz-Michaluk et aL, 1993; Diaz

et al., 1995). Since DHPs, phenylalkylamines and benzothiazepines are clinically

established therapeutic drugs there have been some clinical assessments into their

potential use in analgesia. Clinically relevant doses o f L-type Ca " channel

antagonists demonstrate a moderate potentiation of the analgesic effect o f opioids

(von Bormann et aL, 1985; Boldt et aL, 1987; Carta et aL, 1990; Filos et aL, 1993;

Santillan et aL, 1998), whereas others did not confirm this (Lehmann et aL, 1989;

Roca et aL, 1996; Hasegawa & Zacny, 1997b). In keeping with the hypothesis L-

type VDCC block has been shown to correlate with opioid responsiveness

(Hasegawa & Zacny, 1997b; Santillan et aL, 1998). Findings suggest that chronic

blockade o f L-type VDCCs prevents some compensatory mechanisms responsible

for opioid tolerance and/or expression. Alternatively the vasodilatory effects o f

DHPs could increase absorption o f morphine into the vasculature resulting in

increased distribution to its sites of action, and thus increased effectiveness.

Radioligand binding studies to assess alterations in L-type channels in the tolerant

state have varied in methodology and provide contrasting reports. Chronic morphine

treatment has been reported to increase the number o f membrane bound L-type

VDCCs in the brain (Ramkumar & el-Fakahany, 1988; Zharkovsky et aL, 1993) but

not in the spinal cord (Bernstein & Welch, 1995).

Thus the consensus from results presented here, in conjunction with pre-

clinical behavioural studies is that L-type VDCCs have little role in the spinal

processing o f sensory information in both acute and chronic pain states, unlike its N-

type VDCC counterpart strongly linked to neurotransmitter release and pathological

conditions. Results from pre-clinical and clinical investigations do however reveal

that in situations o f opioid tolerance, L-type VDCC antagonism may be o f

therapeutic benefit.

157

CHAPTER 7

GABAPENTIN AND MORPHINE:

A COMBINED APPROACH TO INHIBITION

OF CA " INFLUX

158

7 .1 I n t r o d u c t io n

Damage to the peripheral and central nervous system can lead to the

development o f neuropathic pain where patients often experience a combination of

sensory deficits with spontaneous and stimulus-evoked pain (allodynia and

hyperalgesia). Due to the multiplicity o f causes o f neuropathy, ranging from trauma,

viral infections to diabetes, and the number o f possible resultant symptoms, the

underlying dysfunctional mechanisms are likely to be diverse. This may contribute to

the problematic clinical management o f nerve injury pains. Neuropathic pain and

epilepsy both share neuronal hyperexcitability as a common underlying mechanism.

There are established antiepileptic drugs that target the generation o f neuronal

hyperexcitability and some o f these have been proven effective in the treatment of

various forms of neuropathic pain (Swerdlow & Cundill, 1981; McQuay et aL,

1995). Excitability blockers, antidepressants and opioids can be useful therapies but

the numbers needed to treat are about three even for the most effective agents

(Sindrup & Jensen, 2000) and their use can be limited by unfavourable side-effects

(Foley & Inturrisi, 1987). The diverse mechanisms and symptoms of neuropathic

pain lends credence to the idea that combination therapy, based on multiple

pharmacological targets and low drug doses, could improve both the pain relief and

side-effect profiles.

The anticonvulsant gabapentin (GBP) is widely becoming accepted as an

alternative treatment for various types o f neuropathic pain (Nicholson, 2000) as it

provides reasonable efficacy and is well tolerated. Designed as an analogue of the

inhibitory neurotransmitter y-aminobutyric acid (GABA) (Satzinger, 1994) to cross

the blood brain barrier, it does not however, clearly modulate G ABA receptor

function, and the mechanism(s) o f its anticonvulsant/analgesic actions remain

undetermined. GBP does not bind to any known neurotransmitter receptor but binds

to a unique site in the CNS identified as « 28, a VDCC modulatory accessory subunit

(Gee et aL, 1996). Animal models have demonstrated the antinociceptive abilities of

specific VDCC blockers in line with electrophysiologically observed reductions in

spinal cord hyperexcitability (as demonstrated in chapters 4, 5 and 6) and together

these highlight the differential role each subtype plays in nociception, often

dependent on the nature of the pain state (see Vanegas & Schaible, 2000, for review).

159

Morphine acts via a number of CNS sites, including the spinal cord where

pre- and to a lesser extent, postsynaptic p opioid receptors mediate the majority of

morphine’s effects upon nociceptive transmission (see section 1.3.2.5 for further

discussion). Of the four identified opioid receptors (p, 5, k and ORL-1), the affinity

of morphine for p opioid receptors is 50 times higher than that for Ô opioid receptors,

and morphine shows minimal affinity for the K and ORL-1 receptors (Wick et aL,

1994; Kieffer, 1997). The p opioid receptors (and 5, K and ORLl receptors) are

coupled to G-proteins and upon binding of morphine (or the appropriate ligand for

the other receptors, see Table 2) to the extracellular domain, a conformational change

in the receptor is elicited that subsequently opens channels (p, 5 and ORLl

receptors). The resultant neuronal hyperpolarization leads to a decrease in the

opening o f VDCCs, and in the case o f k opioid receptor activation this reduction of

Ca^ influx is direct. At presynaptic locations a reduction in the release of

neurotransmitter from the afferent nerve ensues. This presynaptic action is common

to all opioid-mediated inhibitory effects at central and peripheral sites, the net result

being a reduction in neuronal excitability. At the less frequent postsynaptic receptors,

the same mechanism results in the direct hyperpolarization o f the neurone and the

attenuation of firing.

Clinically it is generally agreed that morphine and other opioid drugs are

less effective in treating neuropathic pain compared to nociceptive or inflammatory

pain although the effectiveness of opioids in neuropathic pain as reported in the

literature is still controversial (Portenoy et aL, 1990; Foley & Portenoy, 1991;

Rowbotham et aL, 1991; Jadad et aL, 1992; Cherny et aL, 1994; Kingery, 1997;

Dellemijn, 1999). Possible explanations include the many associated underlying

pathophysiological mechanisms (Dubner, 1991; Besson et aL, 1993), and dose-

escalation (Portenoy et aL, 1990; Rowbotham et aL, 1991). Animal models of

neuropathic pain have been used in an attempt to clarify the relative efficacy of

opioids in different types of neuropathic pain. Studies have revealed that several

factors appear to be accountable for the variability observed in opioid efficacy such

as pre-clinical model, route of administration and testing stimuli used (Attal et aL,

1991; Lee et aL, 1994; Bian et aL, 1995; Lee et aL, 1995; Ossipov et aL, 1995b;

Jasmin er u/., 1998; Suzuki etaL, 1999).

160

Due to the differing ionic mechanisms of inhibition mentioned above, it

could be predicted that morphine and GBP would interact positively, through

concomitant decrease of excitation and increase of inhibition. Thus far, there have no

in vivo electrophysiological studies investigating the role of GBP in conjunction with.

morphine in the spinal processing of neuropathic pain. This study uses the SNL

model, confirmed by behavioural testing, to induce a neuropathic state. Subsequent

to this, electrophysiological studies o f dorsal horn spinal neurones were made to

investigate the effects of low does of subcutaneously delivered GBP and morphine

on a wide range of electrically- and naturally-evoked neuronal activity. Due to the

serious difficulties in conducting combination studies in patients with neuropathic

pain it is hoped this approach may provide a guide to potential improvements in the

treatment of this type of pain.

7.2 M e t h o d s

A total o f 62 male adult Sprague-Dawley rats were used in this study, of

which 19 were naïve, 19 were sham-operated and 24 were SNL. The animals were

prepared in the normal way and the drugs were administered subcutaneously into the

scruff o f the neck (250 ^ volume). In one series o f experiments the effects of

morphine alone (1 & 4 mg/kg applied cumulatively) was monitored until exertion of

maximum effects (a minimum of 50 minutes). In a second series o f experiments the

effects o f a combination of GBP and morphine were investigated. The drug

combination protocol consisted of a single dose o f GBP (either 10 or 20 mg/kg),

monitored for 60 minutes, proceeded by two doses o f morphine (1 & 4 mg/kg).

Reversal of opiate-mediated effects was assessed by spinally application o f 5 |uig

naloxone.

7.3 R e su l t s

7.3.1 B e h a v io u r a l A sse ssm e n t

Rats subjected to SNL exhibited 'guarding' foot posture ipsilateral to nerve

injury and successful replication of the nerve injury model was confirmed by

behavioural testing which demonstrated the development o f mechanical and cooling

161

allodynia o f the injured hindpaw (see chapter 3 for combined behavioural data

obtained from all studies used in this thesis). This was not displayed by either the

contralateral hindpaw, or in the sham-operated rats. Upon establishment o f a

neuropathic state, animals were then used for in vivo electrophysiology and the

pharmacological study at PO days 14 - 17.

7 .3 .2 S p in a l C o r d E l e c t r o p h y s io l o g y : N e u r o n e C h a r a c t e r iz a t io n

The number o f dorsal horn neurones recorded from in each group were 24

in SNL rats, 19 in sham-operated rats and 19 in naïve rats. All neurones had a

receptive field over the hindpaw ipsilateral to surgery (when performed). No

significant differences were found between experimental groups in the mean values

of recorded neurone depth or responses evoked by electrical and natural stimulation.

However, the mean level of ongoing spontaneous activity recorded from neurones in

nerve injured rats was 1.73 ± 2.53 Hz which was significantly higher (P < 0.05) than

that observed in sham and naïve (0.39 ± 0.89 and 0.42 ± 0.71 Hz, respectively). It is

also worth noting that 71% of neurones characterized in SNL rats exhibited

spontaneous activity at a rate greater than 0.1 Hz in comparison to only 37% of

characterized neurones in sham-operated rats and 39% of naive.

7.3 .3 S p in a l C o r d E l e c t r o p h y s io l o g y : E f f e c t s o f M o r p h in e A l o n e

The effect of subcutaneously administered morphine (1 and 4 mg/kg applied

cumulatively), on the electrically- and naturally-evoked dorsal horn neuronal

responses was tested in SNL, sham-operated and naïve animals. Morphine produced

a dose-related inhibition o f the electrically- and naturally-evoked responses in

neurones recorded from naive and sham-operated animals (Figures 25 - 28), with

clear effects seen around 40 - 50 minutes (Figures 25a & 25b). The greatest effect

was seen in the sham group (Figures 26 & 27) with maximal inhibitions achieved

using 4mg/kg morphine in the range 45 ± 15% to 80 ± 11% (excluding Ap-fibres).

Statistically significant inhibitions (P < 0.05) compared to pre-drug control values,

of the post-discharge (Figure 26a), input (Figure 26b), excess spikes (Figure 26c) and

the C-fibre (Figure 27a) measurements were achieved at 4mg/kg morphine for both

naïve (n = 6) and sham (n = 6) groups. 1 mg/kg morphine significantly inhibited the

162

excess spikes and heat response in the sham group as did 4mg/kg for von Frey 9 g

(Figure 28a) and von Frey 75 g (Figure 28b) and heat (Figure 28c) evoked responses

(n = 6, P < 0.05). Morphine was noticeably less effective at inhibiting the evoked

responses in nerve injured rats in comparison to naïve and sham-operated rats

(Figures 25c & 26 - 28). In fact 2 out of 7 cells showed and increase in their response

resulting in and average excitation for the input (Figure 26b) and excess spikes

(Figure 26c) measurements at the low dose. The only inhibitions found to be

significant (n = 7, P < 0.05) were von Frey 9 g with 1 mg/kg morphine and von Frey

75 g with 4 mg/kg morphine (Figures 28a & 28b). No statistical significance was

determined for the direct comparison o f the effects o f morphine between

experimental groups. Morphine mediated inhibitions were reversed with spinally

applied naloxone (Figure 25).

7 .3 .4 S p in a l C o r d E l e c t r o p h y s io l o g y : E f f e c t s o f G a b a p e n t in A lo n e

The effect of subcutaneously administered GBP (10 & 20 mg/kg), on the

electrically- and naturally-evoked dorsal horn neuronal responses was tested in SNL,

sham-operated and naïve animals. GBP produced a dose-related inhibition o f the

electrically- and naturally-evoked responses in neurones in SNL and sham-operated

animals, with clear effects seen around 40 - 50 minutes. The greatest effect was

observed in the neuropathic animals for which statistically significant inhibitions (n

= 10, P < 0.05) compared to pre-drug control values, were achieved with 20 mg/kg

for all measurements (excluding A-fibres, which were relatively spared) in the range

of 24 ± 9% to 56 ± 9% (Figures 26 - 28). The response to noxious heat (Figure 28c)

was most susceptible to the effects o f GBP with statistically significant inhibitions (P

< 0.05) achieved at 10 and 20 mg/kg in nerve injury rats (n = 10) and 20 mg/kg in

sham (n = 7). In contrast GBP did not have a constant inhibitory effect on all

neurones and measurements recorded from the naïve group. With the exception of

the heat response, either 10 or 20 mg/kg GBP (or in some cases both) produced an

overall excitation o f the neuronal responses. This was most marked for post­

discharge (Figure 26a) and excess spikes (Figure 26c) for which 18% ± 29 and 42 ±

28% excitation was achieved, respectively. In the naïve group both doses o f GBP

elicited excitation in 4 out of 7 neurones. No statistical significance was determined

for direct comparison of the effects o f GBP between experimental groups.

163

a 140-1 control1 2 0 -

° 1 0 0 -

1 mg/kg morphine

4 mg/kg morphine

5ltgnaloxine

coo80 -

1 60- G,"o 40 -

20 -

- A — A(3-fibre

■o— C-fibre -•— von Frey 75

100 150 200

b 140-1 j

1 2 0 -

1 0 0 -

60 -

40 -

20 -

c 160-

140-

1 120-oo 100-ÛÛ5"9 80 néa. 60 -

40 -

20-

0 -

I 50

50

100 150 200

If

100

time (min)150 200

Figure 25. Time course of the effect o f subcutaneously applied morphine, with spinal

naloxone reversal, on the evoked response of a typical dorsal horn neurone recorded

from (a) naïve, (h) sham and (c) SNL rats. Examples o f the effect on the Af-fibre, C-

fibre and von Frey 75 g measurements are shown with the cumulative dose indicated.

Data is expressed as % of the averaged pre-drug control values.

164

Postdischarge *a 100

r- 6 0

4 0

Inout□ Naive

El Sham

c 6 0 -

rS 4 0 -

Excess spikesC 1 00 -

I 10 ^ 2 0 I I 1 4— G abapentin — — M orp hine -

1 4—M orphine “

+

G abapentin 10 m g/k g

4 ,(m g /k g)

—M orp h in e •+

G abap en tin 2 0 m g/k g

Figure 26. Effect of subcutaneously administered morphine and gabapentin, both

individually and in combination, on electrically-evoked dorsal horn neuronal responses

recorded from naive, sham-operated, and SNL rats (n = 6 — 10 for each experimental

gi oup in each drug testing protocol). Data are expressed as maximal mean % inhibition

of the pre-dimg values + S.E.M. ^ denotes statistically significant inhibitory response

compared to pre-drug control value; lAr denotes a significantly greater inhibitory effect

of morphine in the presence o f gabapentin compared to morphine alone at a specific

165

^ 1 0 0 -

8 0 -

6 0 -co3 4 0 - IS

20

0 -

-20 -

C-fibre

1 0 0 -,Naive

8 0 - Sham

6 0 - SNL

2 0 -

-2 0 -I

c 100 -,

8 0 -

6 0 -

4 0 -

^ 2 0 -

-20 J G abapentin — ^

A5-fibre

Ap-fibre

1 4-M orphine -

1 4

“ M orp hine “4-

G abap en tin 10 m g/k g

I

— M orphi

4 |(m g/kg)

G abapentin 2 0 m g/k g

Figure 27. Effect o f subcutaneously administered morphine and gabapentin, both

individually and in combination, on electrically-evoked dorsal horn neuronal responses

recorded from naive, sham-operated, and SNL rats (n = 6 - 10 for each experimental

gf oup in each drug testing protocol). Data are expressed as maximal mean % inhibition

of the pre-à^ug values + S.E.M. ^ denotes statistically significant inhibitory response

compared to pre-é‘ug control value; lAr denotes a significantly gi'eater inhibitory effect

of morphine in the presence of gabapentin compared to morphine alone at a specific

dose (P < 0.05).

166

a 1 00 -

8 0 -

6 0 -co

4 0 --5c

2 0 -

0 -

-20 -

b 100-j

8 0 -

6 0 -c_o4 0 -

c7 2 0 -

0 -

-20 -

C 00-]

8 0 -

6 0 -co4 0 -

c

s 2 0 -

0 -

. 2 0 -

von Frev 9g

von Frev 75g

Heat

* *

Ulpenti J I 4-M orphine -

1 4“ M orp hine “

4-

G abapentin 10 m g/k g

□ Naive

[3 Sham

1 4 ,(m g/k g)

“ M orphine —4-

G abapentin 20 m g/kg

Figure 28. Ejfect of subcutaneously administered morphine and gabapentin both

individually and in combination, on naturally-evoked dorsal horn neuronal responses

recorded from naive, sham-operated, and SNL rats (n = 6 - 1 0 for each experimental

group in each drug testing protocol). Data are expressed as maximal mean % inhibition

of the pre-drug values + S.E.M. denotes statistically significcü'it inhibitory response

compared to pre-dmg control v a lu e d e n o te s a significantly greater inhibitory effect

of morphine in the presence of gabapentin compared to morphine alone at a specific

dose (P < 0.05).

167

7 .3 .5 S p in a l C o r d E l e c t r o p h y s i o l o g y : E f f e c t o f M o r p h in e in t h e

P r e s e n c e o f G a b a p e n t in

The effect of subcutaneously administered morphine (1 and 4 mg/kg applied

cumulatively), 60 minutes after the subcutaneous administration of either 10 or 20

mg/kg GBP, on the electrically- and naturally-evoked dorsal horn neuronal responses

was tested in SNL, sham-operated and naïve animals. In combination with GBP,

morphine produced a dose-related inhibition of the electrically- and naturally-evoked

responses in neurones in naïve, sham-operated and nerve injured animals, with clear

effects seen around 40 - 50 minutes, and these were reversed by spinally applied

naloxone (Figure 29). Maximal inhibitions within each experimental group were

achieved with a combination o f 10 mg/kg GBP and 4mg/kg morphine (Figure 26 -

28). The greatest effect was observed in the nerve injury group with maximal

inhibitions in the range 57 ± 12% - 92 ± 5%. In the presence of 10 mg/kg GBP the

levels o f inhibition attained after administration o f morphine of all evoked responses

in each experimental group were greater than those observed with morphine alone. In

comparison to the pre-drug control values, the drug combination inhibitions were

found to be statistically significant in: SNL rats for C-fibre, postdischarge and input

using 1 mg/kg morphine in conjunction with 10 mg/kg GBP; in SNL and sham for

C-fibre using 4 mg/kg morphine with 10 mg/kg GBP; and in SNL, sham and naïve

for post-discharge, input and excess spikes using 4 mg/kg morphine with 10 mg/kg

GBP (Figures 26 & 27). In the presence o f 10 mg/kg GBP both doses o f morphine

significantly inhibited the von Frey 9 g, von Frey 75 g and heat (Figure 28) evoked

responses, in all 3 experimental groups (with the exception of naïve with 4 mg/kg

morphine for von Frey 75 g and heat). When morphine was used in combination with

20 mg/kg GBP the levels of inhibition reached were no greater than those achieved

in the presence of 10 mg/kg GBP (Figures 26 - 28). Effects of the GBP/morphine

combination were fairly comparable between sham and SNL groups. However, in the

neuropathic group there was a marked increase in the inhibitions achieved with

morphine in the presence of GBP compared to the effects o f morphine alone. This

was found statistically significant (n = 7 - 10, P < 0.05) for; C-fibre (Figure 27a) and

input (Figure 26b) using 4 mg/kg morphine in conjunction with both 10 and 20

mg/kg GBP; post-discharge (Figure 26a) using 4 mg/kg morphine with 10 mg/kg

GBP; and von Frey 75 g (Figure 28b) using 1 mg/kg with 20 mg/kg GBP. Since in

the sham group the levels of inhibition mediated by morphine alone were greater

168

than compared to that observed after SNL, the increase observed in the presence of

GBP in sham was not as marked. However, statistical significance was found for

input (Figure 26b) using 4 mg/kg morphine in conjunction with 20 mg/kg GBP and

for heat (Figure 28c) using 1 mg/kg morphine with 20 mg/kg GBP. The neuronal

responses of the naïve group were least inhibited by the drug combination (Figures

26 - 28) for which no statistically significant difference was found in the comparison

o f the effects of morphine in the presence of GBP to morphine alone.

Pre-drug Gabapentin Morphine Naloxone

20020mg/kg 1 mg/kg 4mg/kg

150

o

O 100

von Frey 75g

r T ~T100

T T150 200 2500 50

time (min)

Figure 29. Time course of the effect of subcutaneously applied morphine, with spinal

naloxone reversal, on the evoked response of a typical dorsal horn neurone recorded

from (a) naive, (b) sham and (c) SNL rats. Examples of the effect on the Af-fibre, C-

fibre and von Frey 75 g measurements are shown with the cumulative dose indicated.

Data is expressed as % of the averaged pre-drug control values.

169

7 .4 D is c u s s io n

In this study I have examined the effect o f a combination o f subcutaneously

administered GBP and morphine on the spinal processing o f sensory information

after nerve injury. In naïve, sham-operated and SNL groups significant inhibitions of

the evoked dorsal horn neuronal responses were achieved with a combination o f the

two drugs at doses that when given alone lacked any effect. The greatest increase in

the effectiveness of morphine after GBP treatment occurred after neiiropathy, where

systemic morphine was almost ineffective.

7.4 .1 E f f e c t o f M o r p h i n e A l o n e

The present study has shown that subcutaneously administered morphine

has a reduced inhibitory effect on the electrically-evoked dorsal horn neuronal

responses in rats subject to SNL in comparison to sham-operated and naïve animals.

Additionally, morphine dose escalation did not improve the inhibition mediated in

nerve injured rats, whereas its effect was dose-related in sham and naïve groups.

Doses used here were identical to those demonstrated to have effects in behavioural

studies o f neuropathy and inflammation (Field et aL, 1999b). Within the SNL group

the greatest observed morphine mediated inhibitions were on the mechanical von

Frey stimuli evoked responses. Across all three experimental groups AP-fibre

evoked responses were relatively spared. A reduced sensitivity o f evoked neuronal

responses in SNL rats to systemic morphine has previously been demonstrated in a

similar electrophysiological study (Suzuki et aL, 1999) and this is in line with a

behavioural observation (Kontinen et aL, 1998). However, the antiallodynic and

antinociceptive abilities o f morphine in behavioural studies involving neuropathy is

somewhat variable and appears to be dependent on the model o f neuropathy

employed, behavioural assessment and nature o f stimuli utilized, as well as the route

of morphine administration. In keeping with my findings a reduced effectiveness of

intrathecally administered morphine after neuropathy has been observed for thermal

hyperalgesia in the SNL model (Ossipov et aL, 1995a; Ossipov etaL, 1995b; Nichols

et aL, 1997; Wegert et aL, 1997), the CCI model (Mao et aL, 1995) and PSL model

(Yamamoto & Sakashita, 1999).

170

On the other hand, it has been demonstrated that under neuropathic

conditions, adopting the systemic route of administration, morphine can be effective

in alleviating mechanical allodynia in the SNL model (Bian et aL, 1995; Lee et aL,

1995; Wegert et aL, 1997) see however (Kontinen et aL, 1998), for which the spinal

route has been proven to be both with (Hwang et aL, 2000) and without effect (Bian

et aL, 1995; Lee et aL, 1995; Nichols et aL, 1995; Wegert et aL, 1997). Likewise, in

the CCI model, systemic morphine was shown to be effective against mechanical

allodynia (Kayser et aL, 1995) and hyperalgesia (Attal et aL, 1991 ; Desmeules et aL,

1993; Koch et aL, 1996), thermal hyperalgesia (Yamamoto & Yaksh, 1991; Lee et

aL, 1994; Backonja et aL, 1995), cold allodynia (Hedley et aL, 1995; Jasmin et aL,

1998) and spontaneous pain (Jazat & Guilbaud, 1991). Furthermore, in various

experimental conditions an increased sensitivity to morphine after neuropathy has

also been observed (Kayser et aL, 1995; Suzuki et aL, 1999). In an

electrophysiological study, Suzuki et al (1999) demonstrated that intrathecal

morphine had an enhanced potency on the C-fibre evoked and noxious natural

stimuli evoked neuronal response o f SNL rats. In the same study however, it was

also shown that systemic morphine was almost without inhibitory effect, as I have

also demonstrated. Thus, it appears that there is an inconsistency between the

inhibitory effects of systemically administered morphine observed behaviourally,

where its has a proven effect, in contrast to its lack of mediated inhibition observed

electrophysiologically. The majority o f behavioural data supports the effectiveness of

systemically administered morphine for the relief of neuropathic behaviours. This

may possibly reflect a direct action of morphine at supraspinal (Yaksh, 1997; Sohn et

aL, 2000) or spinal sites (Duggan & North, 1984; Dickenson & Sullivan, 1986),

areas abundant in opioid receptors (see Mansour et aL, 1995), or an indirect action

upon spinal transmission via spinal descending inhibitory controls (see Basbaum &

Fields, 1984). With regards to the lack o f inhibitory effect observed

electrophysiologically here and by Suzuki et al (1999), it should be noted that in

behavioural studies only threshold responses are employed, which contrasts to

suprathreshold neuronal responses employed electrophysiologically, most likely

requiring higher morphine concentrations for inhibition. Indeed, the naturally-evoked

neuronal responses to von Frey 9 g in the present study were most susceptible to

morphine after neuropathy, which may relate more to behavioural testing.

Intravenous versus subcutaneous administration routes may be another issue since in

171

many of the animal studies intravenous morphine mediated greater effects.

Reduced efficacy of opioids on nerve injury-induced static mechanical

allodynia and on hyperalgesia (which implies traditional nociceptive pathways)

could be a result of increased excitation o f spinal neurones (Woolf, 1983) due to

activation o f NMDA receptors (Dickenson, 1997), resultant from peripheral nerve

injury and suprathreshold stimuli (as already mentioned). This enhanced activity may

simply require more opioid drug in order to be controlled at the spinal cord level as

demonstrated by the studies already discussed. This is further supported by an

observed increase in the efficacy of morphine in neuropathic pain models in the

presence o f NMDA receptor antagonists (Chapman & Dickenson, 1992; Yamamoto

& Yaksh, 1992; Nichols et a/., 1997; Wegert et ah, 1997; Christensen et aL, 1998;

Kauppila et aL, 1998b). Postsynaptic cell hyperpolarization can produce inhibition of

sensory transmission by activation of opioid receptors on dendrites of projection

neurones, intemeurones and cell bodies of output neurones. It has been demonstrated

electrophysiologically comparing control animals to those lacking presynaptic opioid

receptors that postsynaptic actions o f opioids require higher does o f systemic

morphine in comparison to presynaptic-mediated effects (Lombard & Besson, 1989).

Thus, the reduced effect o f opioids on hyperalgesias could partly be overcome by

dose escalation. Indeed, neuropathic states may not be completely resistant to spinal

opioids, but rather less sensitive, since it has been reported that intrathecal opioids in

a number o f animal models can produce effective antinociception, albeit at higher

doses than those used in normal rats (Bian et aL, 1995; Lee et aL, 1995; Ossipov et

aL, 1995a; deGroot et aL, 1997; Abdulla & Smith, 1998; Zurek et aL, 2001). It

should also be considered, however, that spread of the opioid drug away from the site

of injection to the brain might occur with the required increase in dose, which could

be responsible for the antinociceptive effects (Zurek et aL, 2001).

Another possible mechanism for the occurrence o f opioid resistance is that

spinal cord opioid mechanisms may be disturbed. As previously discussed (section

1.3.2.5) opioids predominantly act a presynaptic site upon nerve terminals.

Autoradiographic and immunohistochemical methods have demonstrated high levels

of spinal cord opioid receptors around C-fibre terminal zones in lamina I and the

172

substantia gelatinosa, and lower levels in deeper laminae (Besse et aL, 1990b;

Rahman et aL, 1998; Zhang et aL, 1998). A lack of opioid receptors on the terminals

of large diameter, low-threshold A|3-fibres that convey non-noxious information,

explains the nociception-specific actions o f morphine (Dickenson & Suzuki, 1999),

also seen in the present study. Opioid receptors are synthesized in the cell bodies of

small afferent fibres in the DRG and transported to the peripheral and central

terminals, permitting modulation o f neurotransmitter release (Zhang et aL, 1998).

This makes spinal opioid receptors susceptible to the effects o f peripheral nerve

damage in such a way that nerve section has been shown to cause a substantial

reduction in spinal opioid receptors at this level, mostly a result of disturbed axonal

transport (Besse et aL, 1990b). Immunohistochemistry and in situ hybridization has

revealed a down-regulation of jLi opioid receptors in rat and monkey DRG neurones

and in the dorsal horn after complete or partial sciatic nerve injury (Zhang et aL,

1998) Autoradiographic binding studies indicate that peripheral nerve injury induces

a reduction in the level o f opioid receptors in the dorsal horn of the spinal cord

(Lombard et aL, 1990; Stevens et aL, 1991; Besse et aL, 1992). Furthermore a

reduction o f opioid receptors localized to the L6 region, has been observed 7 days

after SNL, ipsilateral to nerve injury (Porreca et aL, 1998). In the case of less severe

peripheral nerve injury where axonal transport is somewhat intact, down-regulation

or destruction of opioid receptors could also possibly be mediated by increased

production o f PKC following activation o f NMDA receptors in postsynaptic cells

(Mao et aL, 1995).

Activation o f opioid receptors on output neurones, unlike the other

postsynaptic and presynaptic locations would not be selective for noxious

transmission, and could explain the small reduction o f low-threshold activity with

doses that abolish nociceptive C-fibre activity, observed in this study and by others

(Chapman & Dickenson, 1992; Suzuki et aL, 1999). Thus, minor inhibitions of low

threshold inputs by opioids are likely to account for the weak effects o f spinal

opioids on allodynias.

Another explanation for the reduced efficacy in neuropathy is the

antagonism of the inhibitory actions o f opioids by the CCK system (see Wiesenfeld-

173

Hallin et aï., 1999). There are two identified receptors for the peptide CCK, named

CCKl and CCK2 (Alexander, 1998), the latter being the predominant type in the

spinal cord of rats, whereas it is CCKl in primates (Hill et aL, 1990; Ghilardi et aL,

1992). Under non-pathological conditions primary afferent neurones do not contain

the peptide CCK, whereas intemeurones in the superficial laminae and laminae IV-

VI o f the dorsal horn appear to be its primary source. Peripheral nerve injury has

been shown to induce an upregulation o f CCK mRNA in ipsilateral DRG neurones

(Verge et aL, 1992; Xu et aL, 1993), though not accompanied by an increase in

peptide immunoreactivity. Whereas upregulation o f the CCK2 receptor mRNA in

DRG neurones (Zhang et aL, 1993b) is reflected by increases in the receptor protein,

particularly in the superficial layers o f the lumbar dorsal horn (see Wiesenfeld-Hallin

et aL, 1999). CCK appears to function as an anti-opioid peptide although its

mechanism is not fully understood. It is not due to a direct hyperalgesic effect, since

CCK does not alter non-pathological pain thresholds, nor mediated via binding to

opioid receptors. CCK receptor antagonists have been found to be antinociceptive

against neuropathic pain administered alone and are able to re-establish the effect of

morphine lost after neuropathy (see Wiesenfeld-Hallin et aL, 1999). Thus, it is

suggested that nerve injury leads to increased activity in the CCK system, through

either changes in the peptide or its receptors, and this in turn reduces the efficacy of

opioids in neuropathic pain which can possibly be overcome by dose escalation.

Inconsistencies in the literature surrounding morphine efficacy in

neuropathic situations may also be explained by the method employed to assess

nociceptive behaviour, variable in both stimulus nature and intensity and thus likely

to be processed via different pathways. For example two types o f mechanical

allodynia, static and dynamic, are evident following nerve damage, determined by

the type o f stimuli required for their detection (Koltzenburg et aL, 1992; Ochoa &

Yarnitsky, 1993; Field et aL, 1999a). Static allodynia is evoked by increasing

pressure to the skin (such as von Frey stimulation) and in neuropathic pain patients

has been shown not to depend on A(3-fibres, since it survives compression-ischaemia

that interrupts conduction in myelinated fibres (Ochoa & Yarnitsky, 1993). Static

allodynia is however, dependent upon capsaicin-sensitive A ô-fibres. Dynamic

mechanical allodynia, induced by lightly stroking the surface of the skin appears to

be signalled by the large diameter myelinated A|3-sensory neurones (Kajander et aL,

174

1992; Koltzenburg et al., 1992; Ochoa & Yarnitsky, 1993; Field et al., 1999a).

Morphine has previously been shown to block static allodynia following systemic

administration in nerve ligation models o f neuropathic pain (Bian et ah, 1995; Lee et

al., 1995), where however, spinal administration was ineffective. In contrast in the

model o f diabetes-induced neuropathy both spinal (Calcutt & Chaplan, 1997) and

subcutaneously administered morphine (Field et al., 1999b) were shown to be

effective at blocking static allodynia, yet the dynamic component was left intact

(Field et al., 1999b). The latter study, which also employed identical doses of

subcutaneous morphine to those used here, is in accordance with the reductions in

the evoked neuronal responses to low von Frey stimuli that I have demonstrated.

Furthermore to the effects o f morphine upon static but not dynamic allodynia, the

responses of dorsal horn neurones to Aô and C-fibre, but not AP-fibre stimulation

have also been shown to be blocked by morphine (Le Bars et al., 1976). It has also

been demonstrated that stroking o f the hindpaw ipsilateral to CCI induced Fos

expression (considered an indirect marker o f nociceptive processes) at the spinal

cord level in superficial (laminae I-II) and deep dorsal horn (laminae V-VI). This

was not observed in control animals, and was insensitive to morphine, in contrast to

that evoked by noxious heat (Catheline et al., 2001). The lack o f effect of morphine

on stroking-evoked Fos expression in the dorsal horn supports the hypothesis that

tactile allodynia is related to the activation o f large primary afferent fibres

(Koltzenburg et al., 1994a; Koltzenburg et al., 1994b; Woolf, 1994). This

appearance of Fos immunoreactivity in the superficial dorsal horn evoked by low-

threshold stimuli possibly results from Ap-fibre sprouting into laminae I-II (see

section 1.3.2.3) or activation o f C-fibres with reduced thresholds after nerve injury.

The results o f the present study and pre-clinical investigations discussed

provide a basis for the difficulties encountered clinically surrounding the efficacy of

opioids in the treatment o f neuropathic pain where their effectiveness remains

controversial, as will be discussed later (Amer & Meyerson, 1988; Portenoy et al.,

1990; Foley & Portenoy, 1991; Kupers et al., 1991; Rowbotham et ah, 1991; Jadad

et a l , 1992; Cherny et a l, 1994; Moulin et a l , 1996; Kingery, 1997; Benedetti et a l,

1998; Dellemijn, 1999). Indeed, the route o f administration, the test used for pain

evaluation (possibly relating to the nature and symptoms of pain experienced in the

clinic) and the origin o f lesion employed in pre-clinical studies impacts upon the

175

relative effectiveness of opioids which may translate to it clinical management.

7.4 .2 E f f e c t o f G a b a p e n t i n A l o n e

It has been demonstrated here that subcutaneously administered GBP, at the

highest dose used (20 mg/kg, which was intentionally low for the main focus o f the

study yet is within the dose range used clinically), had a greater inhibitory effect on

the electrically- and naturally-evoked dorsal horn neuronal responses in rats subject

to SNL in comparison to sham-operated and naïve animals. Effects o f GBP were

established around 4 0 - 50 minutes. The systemic doses used here and time course of

effects observed electrophysiologically are in accordance with those seen

behaviourally, not associated with any motor deficits (Field et ah, 1999b; Kayser &

Christensen, 2000).

Behavioural studies show GBP has a negligible effect against physiological

sensory nociception to thermal (Hunter et a l, 1997), mechanical (Field et al., 1997b)

and chemical stimuli (Shimoyama et al., 1997b; Yoon & Yaksh, 1999), yet it is

effective in pathophysiological situations involving the presence o f central

sensitization. Indeed, in various models of nerve injury, inflammation and surgery,

GBP has been shown to reduce any resultant mechanical/thermal allodynias and

hyperalgesias. Systemically, GBP has been shown to be antinociceptive and

antiallodynic in nerve injury models of SNL (Hunter et al., 1997; Abdi et al., 1998;

Field et al., 2000), CCI (Hunter et al., 1997; Kayser & Christensen, 2000) and PSL

(Pan et al., 1999), and also reduced evoked dorsal horn neuronal responses after SNL

(Chapman et al., 1998a). In the nerve injury model o f diabetic neuropathy systemic

and intrathecal GBP blocked both static and dynamic components o f mechanical

allodynia, (Field et al., 1999b; Field et al., 2000). GBP prevented only the second

phase, which reflects central sensitization (Coderre et al., 1990), o f the formalin

response when administered intrathecally (Field et al., 1997b; Shimoyama et al.,

1997b; Yoon & Yaksh, 1999) and systemically (Field et al., 1997b). Both systemic

and intrathecal GBP blocked carrageenan-induced thermal and mechanical

hyperalgesia (Field et al., 1997b), which was supported by its systemic ability to

block the maintenance of carrageenan-induced sensitization o f dorsal horn neurones

(Stanfa et al., 1997). Additionally after SNL, subarachnoid GBP relieved tactile

176

allodynia (Hwang & Yaksh, 1997), systemic GBP prevented the development of

mechanical allodynia and hyperalgesia in a model of post-operative pain (Field et al.,

1997a), and intrathecal GBP was demonstrated to reverse thermal hyperalgesia (Jun

& Yaksh, 1998). GBP also had some inhibitory effect in the experimentally

appropriate sham-operated control group, yet had a tendency to increase the neuronal

responses of the un-operated naïve group . This is in accordance with other similar

studies and may well result from the ability of the drug to act in inflammatory states

(Stanfa et a l , 1997; Chapman et a l, 1998a).

Peripheral nerve damage leads to the generation of ectopic activity at the

injury site (Wall & Devor, 1983) or within the DRG (Wall & Devor, 1983; Kajander

et al., 1992). This provides an ongoing, continual barrage of neuronal activity via

primary afferents into the spinal cord which could lead to central sensitization and

hyperexcitability (Gracely et ah, 1992; Sheen & Chung, 1993; Yoon et al., 1996).

Interestingly it has been demonstrated that GBP inhibits the ectopic discharge

activity from injured nerve sites in the PSL model at doses determined therapeutic

for tactile allodynia (Pan et al., 1999), yet not after SNL (Abdi et al., 1998).

Throughout all these pre-clinical studies, systemic doses of GBP are within a similar

range (10 - 300 mg/kg). However required therapeutic doses o f GBP utilizing the

spinal route of application are much lower (10 - 100 pg per animal) and the positive

effects of direct administration indicates GBP has a spinal site o f action. As yet

mechanism(s) o f its anticonvulsant/analgesic effects remain elusive.

GBP was originally designed as an analogue o f GABA that would cross the

blood brain barrier (Satzinger, 1994), for indication as an anticonvulsant.

Nonetheless, GBP does not bind to either GABAa or GABA b receptors, it is not

converted metabolically into GABA and is neither a substrate nor and inhibitor of

GABA transport (Taylor et al., 1998). Some evidence suggests that GBP increases

GABA release, thereby enhancing inhibitory neurotransmission (Gotz et al., 1993).

More recently it has been shown that the antiallodynic action of GBP is not sensitive

to either GABAa or GABAb receptor antagonists (Hwang & Yaksh, 1997; Patel et

al., 2001), therefore the increase in CNS GABA levels are unlikely to be responsible

for the mode o f action o f GBP in models o f pain. In vitro, GBP stimulated the

177

activity o f catabolic enzyme glutamate dehydrogenase at high concentrations

implicating effects on glutamate metabolism (Goldlust et ah, 1995). However, the

onset o f these metabolic effects on neurotransmitters would be too slow to solely

account for the time course of antinociceptive effects observed as early as 15 minutes

after intrathecal GBP administration (Shimoyama et ah, 1997a). GBP does not affect

ligand binding at a variety of commonly studied drug and neurotransmitter binding

sites and voltage-activated ion channels including GABA, glutamate and glycine

receptors, as demonstrated by lack of inhibition of [^H]-GBP binding (Taylor, 1995).

A range of anticonvulsants including, carbamazepine and ethosuximide, also do not

displace bound [^H]-GBP, suggesting a novel pharmacological site to which binding

is important for GBP’s antiepileptic activity (Taylor et ah, 1998). In vivo behavioural

studies have shown that the antinociceptive effects o f systemic GBP are prevented by

D-serine, a glycine/NMDA receptor agonist (Singh et ah, 1996). However,

radioligand binding assays have not shown GBP to inhibit strychnine-insensitive

[^H]-glycine binding to brain membranes or to influence the binding of [^H]-MK-801

to the NMDA receptor channel, and D-serine does not bind to the novel GBP binding

site (Suman-Chauhan et a l, 1993). Although many studies have attempted to

establish the mechanisms responsible for the actions o f GBP, identification o f

definite cellular or molecular target(s) is still awaited. Given the possibilities, it is

fairly probable that different mechanisms account for the anticonvulsant,

antinociceptive and anxiolytic actions of GBP. This is supported by the fact that GBP

has a delayed anticonvulsant effect after bolus injection in rats (Welty et ah, 1993)

compared to the analgesic action attained rapidly after intrathecal injection (Field et

a l, 1997b).

By far the most convincing, and experimentally supported theory for the

mode action of GBP, over the other possibilities, is a mechanism involving the

inhibition o f Ca " influx via VDCCs, which is of interest to this thesis. The only

GBP binding site to be identified was originally purified from pig brain and found to

correspond to the «28 subunit o f VDCCs (Gee et a l , 1996). The VDCC aid subunit

is expressed in skeletal muscle, brain and heart (Ellis et a l , 1988). In vitro studies

have shown that the carboxyl ô peptide contains the transmembrane domain that

plays a role in anchoring the aid subunit and in stabilizing subunit interactions (Jay

et a l , 1991; Gurnett et a l , 1996). The remaining majority o f the protein is

178

extracellular (Brickley et al., 1995; Gurnett et a l , 1996; Wiser et al., 1996)

suggesting little interaction with intracellular molecules such as protein kinases and

G-proteins. The «26 subunit is likely a rate-limiting factor in VDCC assembly, since

coexpression with other VDCC subunits results in enhancement o f current amplitude

(Mikami et al., 1989; Brust et al., 1993; Gurnett et al., 1996) and increased

expression o f co-conotoxin and DHP binding at the cell surface (Singer et al., 1991;

Brust et al., 1993). Since ociS is able to modulate VDCC function, and coexpression

/assembly with the a i subunit is required for the physiological activation o f Ca

currents (Gurnett et al., 1996), interaction of GBP with this subunit may permit the

drug to influence neuronal excitability. Indeed, there are several in vitro studies

establishing this link. GBP inhibited Ca currents in isolated rat brain neurones

(Stefani et al., 1998) and K^-induced Ca influx in synaptosomes (Ma & Woolf,

1995; Fink et al., 2000; Meder & Dooley, 2000) and subsequent neurotransmitter

release from rat neocortical slices (Fink et al., 2000). GBP presynaptically inhibited

glutamatergic synaptic transmission in rat spinal cord slices (Shimoyama et al.,

2000), predominantly in the superficial lamina where nociceptive primary afferent

fibres terminate. Therefore antinociceptive effects of GBP may involve the inhibition

o f the release of excitatory amino acids from presynaptic terminals. Furthermore

GBP inhibited synaptic transmission in the superficial dorsal horn, using spinal cord

slices from hyperalgesic, but not normal, rats (Patel et al., 2000). GBP also reduced

HVA Ca^ current amplitude in DRG cells in a use-dependent manner (Alden &

Garcia, 2001) over a clinically effective range (10 - 100 p-M) as determined from the

serum of human patients (Taylor et al., 1998). A consistent inhibition o f HVA Ca "

conductance, but not Na" or glutamate currents and GABA responses, was mediated

by GBP in isolated adult rat cortical neurones (Stefani et al., 2001). Similarly GBP

inhibited HVA Ca currents in the same neurones in which Na^ channels were

unaffected (Wamil & McLean, 1994). Furthermore, the inactivation steady-state

properties o f Ca^ channels were not significant modified by GBP, whilst

deactivation gating was slightly modulated, probably through interaction with the

ŒzÔ subunit. When studied in different expression studies (X2Ô has been shown to be

responsible for a substantial increase o f the whole-cell Ca " current (Stefani et al.,

2001). This is probably not only a result o f increased a i subunit in the plasma

membrane, but mainly through alterations in the mean open time o f the channel itself

(Shistik et al., 1995; Gurnett et al., 1996; Saegusa et al., 2000). In contrast, GBP has

179

been shown to be without effect on current effects when tested in cultured

nodose and DRG neurones (Rock et a l , 1993) and acutely dissociated human

hippocampal granule cells (Schumacher et al., 1998), which may be explained by

their use of cultured neuronal cells subject to developmental variables.

The a 20 subunit is common to all VDCCs, yet GBP has been demonstrated

to specifically target chronic pain with minimal adverse effects (Field et al., 1997a;

Field et al., 1997b). The ubiquitous distribution of VDCCs may thus require specific

subunit combinations and particular modulatory conditions to allow GBP to exert its

inhibitory actions under pathophysiological circumstances. Three different azô

subtypes («28-1, 2 and 3) and several splice variants with differing distributions have

recently been described (Klugbauer et al., 1999; Hobom et a l, 2000), and it appears

that GBP may only interact with a 2Ô-l, based on comparative amino acid sequences

and mRNA distributions with the GBP-binding protein (Taylor et al., 1998).

Expression of a rat DRG 0C2 subunit, unique at least at the post-translational

modification level, has also been reported (Luo, 2000). Nerve injury induces marked

tissue-specific up-regulation of the DRG VDCC «26 subunit (Luo et al., 2001). The

latter study demonstrated that upregulation o f DRG «28 mRNA and protein was

evident in rats with SNL, unilateral sciatic nerve crush, but not dorsal rhizotomy,

indicative of a peripheral site o f regulation o f expression. After SNL a temporal

correlation between DRG «28 subunit upregulation and allodynia was also

established. Differential expression of distinct (X28 subunits in spinal cord and DRG

indicated that spinal cord «28 was not synthesized and retrogradely transported from

DRG. The reported spinal antihyperalgesic actions of intrathecal GBP could actually

result from drug interactions with spinal cord and/or DRG (X28 subunits as other

intrathecally delivered agents have been shown to access DRG (Porreca et al., 1999;

Lai et al., 2000). Since tissue-specific expression o f the DRG OC28 is implicated in

models o f neuropathic pain, this subunit may be an important component in

governing tissue-specific functions o f VDCCs and neuroplasticity contributing to

allodynia after nerve injury. It is known that the (X28 subunit is common to all

VDCCs (Hofmann et al., 1994; Isom et al., 1994) therefore GBP actions may

involve more that one type o f VDCC. Given the predominant role of N-type VDCCs

in neuropathic pain after nerve injury (see chapter 4), in conjunction with the

180

heterogeneity o f a 2Ô (Luo, 2000) subunits it is likely that functional and tissue-

specific expression of the N-type VDCCs may depend on its subunit composition.

This would allow N-type VDCCs to make unique contributions to nerve injury-

induced neuropathic pain and also provides GBP with a potential subunit-specific

interaction associated with the pain pathway, which may explain its specificity of

action.

Some other intriguing evidence supporting the 0C2Ô subunit as the mediator

of GBP’s antihyperalgesic actions comes from studies involving stereoisomers of

GBP analogues. To date GBP and (S)-3-isobutylgaba (pregabalin) are the most

active compounds that interact with the isolated GBP-binding site, predicted to be

a 2Ô. Pregabalin stereoselectively inhibits [^H]-GBP binding to brain membranes with

the (S)-isomer showing similar affinity as GBP in contrast to the R-isomer which

was shown to be 10 times weaker (Taylor et al., 1993). Using the model of diabetes-

induced peripheral neuropathy (Field et al., 1999b; Field et al., 2000) and SNL

(Field et al., 2000), GBP and pregabalin blocked allodynia with similar potency

whereas its (R)-isomer was inactive. These reports are consistent with the

involvement of the azS subunit in the mediation of GBP’s antinociceptive effects.

Similar stereoselective effects have also been reported in animal models of epilepsy

(Bryans et al., 1998), reinforcing the hypothesis that modulation o f HVA Ca "

currents is critical (although probably not sufficient) in the putative effect of GBP as

an anticonvulsant.

There is an established role for VDCCs in nociception and the contribution

of the different channel subtypes appears to alter depending on the nature of the pain

(see Vanegas & Schaible, 2000). Chapter 4 clearly demonstrates a predominant

nociceptive role for N-type VDCCs, which is enhanced after neuropathy. In contrast,

chapter 4 also demonstrates that the P/Q -type channels play a modest role in sensory

transmission in a manner that is not altered after nerve injury. Likewise, in chapter 5,

inhibition o f the T-type chaimel vvdth the anti-epileptic ethosuximide was also shown

to have a limited effect on evoked dorsal horn neuronal responses, again remaining

unchanged after neuropathy. Upon substantial membrane depolarization N- and P/Q-

type VDCCs mediate the release o f excitatory neurotransmitters, such as glutamate.

181

substance P and CGRP, critical for wind-up and central sensitization, in the presence

of constant afferent input (Dickenson, 1994). Since T-type channel activation occurs

close to resting potential, they allow Ca " influx when cells are at rest thus regulating

cell excitability and most likely the depolarization required to activate high-voltage

activated N- and P/Q-type channels necessary for neurotransmission. The discussions

raised in chapters 4, 5 and 6 suggest that the N-type VDCCs are selectively subject to

plasticity after nerve injury and thus could potentially be the main candidate target

for GBP, if indeed it does exert its actions through VDCCs.

A few electrophysiological studies have attempted to define the VDCCs

inhibited by GBP, but no general consensus has been determined. It has been

demonstrated that GBP alters the affinity of DHP for the Ca channel (Stefani et al.,

2001). In another study GBP inhibited whole-cell Ca currents recorded from

neurones isolated from adult rat brain in a fast and voltage-independent manner.

DHPs (nimodipine and nifedipine) prevented the GBP-mediated inhibition, whereas

co-conotoxin G VIA and MVIIC (a P/Q-type VDCC blocker) partially reduced its

effect (Stefani et al., 1998). They hypothesize that L-type, in comparison to N- and

P/Q-type, are the predominant VDCC target of GBP in these neurones. However this

study failed to present data comparing the level o f inhibition achieved with the used

doses o f L-, N- and P/Q-type VDCC blockers alone. It appears that blockade of L-

type Ca " current produced substantial inhibition o f the Ca current leaving very

little to be further inhibited by a dose o f GBP that did not produce a similar

inhibition alone. In contrast, the inhibition produced by the co-conotoxins alone was

comparable to that achieved with GBP alone. Additionally, differences in the mode

o f VDCC antagonism exerted by DHPs and the co-conotoxins might possibly affect

the binding o f GBP to the «26 subunit or its subsequent inhibitory mechanism and

thus these results should be viewed with caution. Given that L-type VDCCs are not

involved in neurotransmitter release and play a minimal role in sensory transmission,

not altered by peripheral nerve damage, it seems functionally unlikely that GBP

would exert its antinociceptive effects through these channels in vivo. Moreover, the

critical role o f N- and P/Q-type VDCCs in neurotransmission and the similarities in

the enhanced role of N-type VDCCs in nociception and antinociceptive actions

exerted by GBP after nerve injury highlight these channels as a much more

convincing substrate. GBP has been shown to inhibit glutamate and aspartate release

182

by 20% from synaptosomes (Fink et aL, 2000) and this suggests that GBP inhibited

Ca influx into glutamatergic terminals by acting on P/Q-type VDCCs, which

predominate on synaptosomal terminals (Burke et al., 1993; Luebke et al., 1993). In

contrast, other studies GBP has been shown not to significantly affect L-, N-, or T-

type Ca^ currents (Rock et a l , 1993). However given the diversity, and tissue

specific expression o f the various VDCC subunits, the potential combination

heterogeneity o f VDCCs at the structural level is enormous. It is possible that GBP

exerts functional effects only with a particular combination of subunits. Moreover,

these effects specific to chronic pain states may be observed only under

pathophysiological conditions, perhaps dependent upon situations that mimic the

neuronal excitability that characterizes epilepsy, GBP’s initial clinical target.

Presented with all the pre-clinical and in vitro evidence there is a substantial

body o f evidence linking an aid-mediated GBP mode of action, certainly with

respect to chronic pain. It remains to been seen whether the aid subunit has an as yet,

undiscovered alternative physiological function. Although so far, the interaction of

GBP with the aid most likely impacts upon the functioning of one or more subtypes

of VDCCs.

7 .4 .3 E f f e c t o f G a b a p e n t i n a n d M o r p h i n e C o - a d m i n i s t r a t i o n : C l i n i c a l

I m p l i c a t i o n s

The present investigation has convincingly demonstrated that after nerve

injury the evoked dorsal horn neuronal responses, shown to be refractory to systemic

morphine, become susceptible to its inhibitory actions in a dose-related manner when

in the presence o f systemic GBP (itself at a dose that alone mediated negligible

neuronal inhibitions). These inhibitions were reversed by spinally applied naloxone,

which indicates a spinal site of action and provides a way of reversing the effects of

the combination if adverse effects arose with the combination in patients. Since the

two agents, at ineffective doses alone, interacted in a positive manner, block of the

effect o f one of the agents abolishes the entire effect. The spinal cord site of

interaction of the combination indicates that the substrate acted upon by GBP must

reside on circuitry also possessing )Li opioid receptors. This would support the ideas

183

expounded above that tissue-specific, and in this context, spinal (X2Ô subunit events

could well mediate the observed effect.

In behavioural studies systemic GBP has been shown to have effects that

peak 1 hour after administration, and that are still apparent at 5 hours (Hunter et al.,

1997). This indicates that GBP would still be active throughout the entire time-

course o f an experiment in the present study where morphine was given 1 hour after

GBP and then followed for a further 2 hours. Thus, there is great clinical promise for

the treatment o f poorly opioid-responsive neuropathic pains for which many cases

morphine dose-escalation becomes limited by adverse side-effects. The majority of

the evoked neuronal responses were significantly inhibited by a combination o f GBP

and morphine, the greatest effects being seen on the C-fibre, postdischarge, input and

von Frey evoked responses. These observations are important since high-threshold

C-fibres specifically relay nociceptive information, post-discharge is a measure of

spinal cord hyperexcitability and von Frey evoked responses may relate to

behaviourally observed mechanical allodynia, all important features of neuropathy.

In contrast, the AP-fibre response was relatively spared, which may indicate that

physiological touch sensations would be left intact (see chapter 3).

Other pre-clinical studies have also investigated the use of morphine in

conjunction with other drugs. GBP has been shown to enhance the antinociceptive

effects of morphine in the rat tail flick test (Shimoyama et al., 1997a). Intrathecal

morphine and N-type VDCC blockers have also been proven to mediate additive and

synergistic inhibitions in rat models of nociception (Wang et al., 2000a; Pirec et al.,

2001), which is supportive of an N-type VDCC mediated inhibitory effects of GBP.

After nerve injury synergistic effects of morphine have been observed with clonidine

(Ossipov et al., 1997) and with neostigmine (Hwang et al., 2000). However, GBP’s

central action does not involve an opiate mechanism, tolerance does not develop and

morphine tolerance does not cross-generalize to GBP’s antihyperalgesic effects

(Field etal., 1997b).

L-type VDCC blockers have been clinically investigated for their analgesic

effects when administered alone and in combination with morphine. In the absence

184

of a pathophysiological pain state, these studies revealed no L-type VDCC mediated

analgesic effects (Hasegawa & Zacny, 1997a). However, nimodipine decreased the

required morphine dose in cancer patients with severe pain who had developed

morphine tolerance (Santillan et a i , 1998), an effect that possibly reflects alterations

in L-type VDCCs related to the mechanism of morphine tolerance (Bernstein &

Welch, 1995, see also chapter 6). This lends credence to the notion that GBP is

unlikely to exert its analgesic effects via L-type VDCCs.

There is an increasing amount o f clinical data in support o f the control of

various neuropathic pains with GBP, such as postherpetic neuralgia, painful diabetic

neuropathy, multiple sclerosis and trigeminal neuralgia, to name a few (Nicholson,

2000). Two separate randomized, double-blind studies in postherpetic neuralgia

patients (Rowbotham et al., 1998) and patients with diabetic peripheral neuropathy

(Backonja et al., 1998) reported significant reductions in pain scores compared to

placebo over a period of 8 weeks. Furthermore, improvements on sleep and mood

were noted for the GBP-treated patients and tolerance did not appear to develop.

Reported adverse effects included somnolence and dizziness, however occurrence of

these was minimal and where experienced only mild to moderate. Alongside other

studies and case-reports it has been proposed that GBP is a useful first line treatment

for chronic neuropathic pain, especially where other therapies fail (see Nicholson,

2000, for review). GBP has substantial advantages concerning pharmacokinetics,

safety and tolerability in comparison with established anticonvulsants, as their

clinical use is often limited by a small therapeutic range and several drug interactions

(Patsalos & Duncan, 1993; Dallocchio et ah, 2000). Clinically, GBP is well tolerated

and does not show any acute or chronic toxicity, it is excreted by glomerular

filtration unmetabolized, and does not present any unwarranted drug interactions

(Goa & Sorkin, 1993). However larger long-term controlled trials are needed to

determine the place of GBP in relation to other therapeutic options (Hemstreet &

Lapointe, 2001).

On the other hand, the clinical use o f opioids in neuropathic pain is more

complex and varied responsiveness is reported which may be due to the different

types of pain, mediated by differing causal dysfunctions. There have been many

185

randomized, double blind controlled trials conducted in this area (Dellemijn, 1999)

and morphine effectiveness ranges from opioid resistant (Amer & Meyerson, 1988)

to modest pain relief (Rowbotham et al., 1991) to good pain relief (Likar et al.,

1999). It is widely believed that neuropathic pains are less susceptible, but not

resistant to systemic morphine (Jadad et al., 1992; Cherny et al., 1994), as observed

in the present study. Results from pre-clinical studies indicate that the route of

administration, the test used for pain evaluation (possibly relating to the nature of

pain experienced in the clinic) and the origin o f lesion employed in pre-clinical

studies impacts upon the relative effectiveness o f opioids which may translate to it

clinical management. Although dose-escalation appears to overcome the lack of

antinociceptive actions o f opioids in animal models o f neuropathic pain, clinically

this is hindered by side-effects (such as sedation, nausea, vomiting, constipation and

respiratory depression) (Foley & Inturrisi, 1987), or the development of tolerance

(Portenoy, 1994). As a consequence, opioid monotherapy may result in inadequate

analgesia. Because a multiplicity of mechanisms are involved in pain the limitations

in the use o f opioids can be overcome by combination with one or more non-opioid

analgesics to obtain a more favourable balance of analgesia and side-effects.

There are many benefits to a multi-drug regimen. Since neuropathic pain

has many causes and resultant symptoms, the underlying dysfunctional mechanisms

are likely to be diverse, hence its problematic clinical management. The use o f GBP

to target the excitatory system and morphine to target the inhibitory system, tackles a

sensitized spinal cord from two perspectives. It is also likely to offer a better side-

effect profile since, as shown in this study, only low doses o f both GBP and

morphine, each alone being ineffective, were required to mediate significant effects.

The clinical use of GBP in conjunction with morphine may also confer important

characteristics o f GBP-mediated antinociception, seen in pre-clinical studies, that

morphine lacks. In particular its ability to block both static and dynamic components

of mechanical allodynia (Field et al., 1999b), and the lack o f tolerance development

with chronic use (Field et al., 1997b).

Since the undertaking o f the present investigation two clinical studies have

been published also examining the potential therapeutic potential of a morphine and

186

GBP combination for analgesia (Caraceni et al., 1999; Eckhardt et al., 2000).

Eckhardt et al (2000) investigated in a randomized, placebo-controlled, double­

blinded study, the pharmacodynamic and pharmacokinetic interaction of GBP and

morphine in 12 healthy volunteers using the cold pressor test. GBP \vas shovm to

have no analgesic effect on its own after a single oral dose of 600mg in comparison

to placebo, which was not surprising since the cold pressor test is not accompanied

by tissue or nerve damage. The same dose of GBP significantly enhanced the

analgesic effect o f morphine in comparison to morphine alone. Only the

pharmacokinetics of GBP and not the pharmacokinetics and metabolism of morphine

were significantly altered upon coadministration. Morphine alone led to the expected

opioid-mediated side-effects, however these were no different to those observed with

the morphine and GBP combination. Caraceni et al (1999) studied the combination

of morphine and GBP in patients suffering neuropathic cancer pain where improved

analgesia in comparison with morphine alone was demonstrated. These results

warrant further controlled trials in patients for the relevance of the combination of

morphine and GBP for treating severe pain, where the analgesic effect of GBP

should be more pronounced than in healthy volunteers, considering the data derived

from animal models of chronic pain.

Pre-clinical and clinical results support the clinical investigation of GBP and

morphine in combination for the treatment of chronic pain, which responds poorly to

opioids. Although, pre-clinical data is invaluable in elucidating drug effectiveness,

caution must be exerted when attempting to extrapolate animal antinociception and

neuronal responses to human analgesia where pain perception also involves its

intensity and unpleasantness. None the less, the results I present here are generally in

accordance with the observed clinical data surrounding the use of morphine and GBP

alone in neuropathic pain, and together in the limited clinical studies reported.

Alongside the favourable profile of GBP, especially in comparison with other

anticonvulsants, the clinical investigation of the additional administration of GBP in

patients suffering severe and chronic pain treated with morphine is warranted. The

limitations of opioids have led to the concept of combination therapy in the hope of

attaining a favourable balance o f analgesia and side-effects with a reduction in the

dose. I have demonstrated here the potential for the use of low dose morphine and

GBP combinational treatment after neuropathy. A GBP/morphine combination

187

should result both in better pain control, including types o f pain that respond poorly

to opioids, and a decrease in morphine dose and consequently a decrease in opioid-

mediated adverse reactions.

188

CHAPTER 8

FINAL DISCUSSION

189

Basic pain research in the preclinical setting has made many advances in

recent years in elucidating the neurobiology of neuropathic pain. However, clinical

interpretation and medical use o f this new-found knowledge has been relatively slow

and pharmacological treatment of neuropathic pain remains contentious. In terms of

aetiology, anatomical site o f nerve lesion and symptoms, neuropathic pain is

somewhat heterogeneous. Causes are diverse and range from trauma to metabolic

disorders, such as diabetes, to viral infections, such as herpes zoster. Attempts to

apply particular treatment regimens have previously been based upon a pathology-

based classification, however on the whole, the various clinically used drugs have

not clearly been seen to exert conspicuous differences in their effects on various

types of neuropathic pain (McQuay et aL, 1995; Sindrup & Jensen, 1999). Despite

such sources o f diversity many neuropathic conditions manifest common clinical

phenomena, including sensory deficits, allodynia and hyperalgesia, however these

are encountered by patients in various combinations and to differing extents, and

thus it may be at the level of experienced symptoms that underlying neuropathic

mechanisms differ. This concept is not encompassed by a pathology-based approach

to neuropathic pain management and thus may be the source o f much o f the

contention. A fresh perspective in neuropathic pain pharmacotherapy is now

suggested to be symptom-based (Woolf et a l , 1998). Although this has been put

forward as a new approach, already in practice, anticonvulsants are generally used to

control the lancinating or stabbing qualities of neuropathic pain whereas constant

burning pain components are tackled with antidepressants and Na" channel blockers

(MacFarlane et ah, 1997). However, this approach still requires a more definitive

appreciation o f the underlying pathophysiology. Improved knowledge of the changes

induced by nerve injury ascertained from pre-clinical studies together with clinical

information regarding quantification o f specific symptoms and particular drug

effectiveness in various neuropathic conditions should permit a judicious approach to

effective treatment.

Recent work with animal models has indicated that several key events

contribute towards the pathogenesis of abnormal pain states following peripheral

nerve injury. Dysfunctions in both the peripheral and central nervous systems are

implicated and interaction between the two pathogenic loci is critical. Peripheral

nerve damage has been shown to enhance expression and initiate redistribution of

190

channels including some of the newly discovered Na" channels from sensory

neurone cell bodies to the injury site (Novakovic et a l , 1998). Peripherally, this

contributes to the appearance o f ectopic discharge in damaged primary afferent

axons and the emergence o f abnormal evoked responses in intact primary afferent

axons within the damaged nerve, and these aberrant signals are communicated to the

CNS where further changes ensue (Wall & Devor, 1983). This is corroborated by the

observation that evoked pains are sustained after resolution o f the ectopic discharge

in some animal models of neuropathic pain (Scadding, 1981). Despite the fact that

these peripheral mechanisms appear to be the origin o f pathophysiological pain

states, many of the drugs that act to inhibit peripheral excitability via Na" channel

block have narrow therapeutic windows. Identification of novel drugs to target the

specific Na" channels responsible in these locations, that lack side-effects, are

awaited. The key role of peripheral mechanisms cannot be understated but the drugs

that block Na" channels, although shown to exert effects at peripheral sites, will also

reduce the excitability o f central neurones, and indeed are used exactly for this latter

effect in epilepsy.

In the CNS the key pathogenic event seems to be triggered in dorsal horn

neurones by C-fibre inputs, although a role for Aô-fibres cannot be excluded. Thus,

peripherally driven, high frequency ectopic discharge culminates in the generation of

central sensitization. This involves enhanced spinal neurotransmitter release,

expanded receptive fields and increased dorsal horn neuronal excitability, which are

indicative o f the associated central spinal plasticity possibly responsible for the

evoked pains of neuropathic conditions. Reorganization of receptive fields is an

established characteristic of nerve injury-induced neuronal plasticity, and as

discussed in chapter 3, this was reflected by the increased receptive fields of

convergent dorsal horn neurones mapped to low von Frey hair stimulation in the

animal model of SNL, For any given stimulus it is expected that an increased number

of spinal neurones would be recruited in the presence of an expanded receptive field

size, such that a greater afferent input and likely enhanced pain transmission would

result (Suzuki et al., 2000b). Peripheral nerve damage-induced plasticity of voltage-

dependent and ligand-gated ion channels is also apparent at the level of the spinal

cord and there are numerous drugs that act upon the various systems to reduce

central hyperexcitability. These agents produce reductions in transmission mediated

191

by modulation o f inhibitory systems, such as morphine, or blockade of excitatory

events such as ketamine at the NMDA-receptor complex. Given the convergence of

peripheral afferent input that occurs at the level o f the spinal cord, however

appealing the goal o f a symptom-based therapy for neuropathic pain, this might not

be attainable through modulation o f specific aspects o f the pain pathway. It appears

somewhat ambiguous that this would be symptom or modality-specific. More

workable and possibly beneficial would be the targeting of aberrant excitatory or

inhibitory systems at play in neuropathic pain.

Common to all neuropathic pain states is increased dorsal horn neuronal

excitability and at the root of this is Ca " entry into neurones via N- and P/Q-type

VDCCs, which is ultimately responsible for permitting neurotransmitter release.

Blockade o f VDCCs may simultaneously attenuate the release o f multiple

neurotransmitters that act at a number of receptor systems that participate in nearly

all states/models o f hyperalgesia and allodynia. This offers an advantage over

individual receptor antagonists as they would have a broad mechanism of action that

would not be modality, or possibly symptom-specific. Block of neurotransmitter

actions at their source o f release would possibly offer a treatment useful across the

spectrum of neuropathic syndromes. This description insinuates that blockade of

VDCCs is rather heavy-handed, however this approach does offer some beneficial

features, as has been clearly demonstrated and discussed in chapters 4, 5 and 6. Since

these channels are voltage-gated it should be possible to reduce aberrant transmission

leaving more physiological events unchanged. A predominant role for N-, over P/Q-,

T- and L-type VDCCs in sensory transmission has been established at the level of the

dorsal horn, in a manner that is enhanced after peripheral nerve damage, and thus

provides some specificity in mode o f action. This is substantiated by numerous

behavioural studies where block o f spinal N-type VDCCs is antiallodynic and

antihyperalgesic in models of neuropathy, and in contrast to opioids which may also

mediate a general inhibitory action including prohibiting neurotransmitter release,

block o f VDCCs does not appear to generate tolerance (Malmberg & Yaksh, 1995;

Bowersox et al., 1996). Furthermore, immunolocalization has demonstrated N-type

channels to be ideally located in the superficial dorsal horn where nociceptive fibre

afferents terminate. Spinal P/Q-type VDDCs are found primarily on the nerve

terminals of dorsal horn neurones in laminae II-VI, but unlike N-type channels they

192

are rarely found in neurones that contain substance P (Westenbroek et ah, 1998).

Intriguingly, this finding may indicate that N-type channels regulate the release of

neurotransmitters including glutamate, from peptide-containing C-fibres, whereas P-

type could possibly be localized to non-peptide, IB4-positive neurones. The naturally

occurring peptides that selectively block N-type VDCCs are unsuitable for clinical

application, but the synthetic analogue SNX-111 has shown encouraging analgesic

effects in clinical trials. Despite the generation of a potentially new drug for the

treatment o f neuropathic pain, SNX-111 still requires inconvenient intrathecal

administration and is not without unwanted side-effects (Brose et aL, 1997).

VDCCs may well be the critical link between the peripheral mechanisms of

neuropathic (and also inflammatory) pain to central events. The role of peripheral

Na^ channels is clear in both nerve and tissue damage-related pains. Changes in their

activity, distribution and expression may well lead to changes in action potential

characteristics. In fact, the TTX-resistant Na^ channels can have very different

biophysical characteristics from the TTX-sensitive channels. Changes in the

functional aspects o f the combinations o f these channels in peripheral nerve

terminals could either cause repetitive firing or altered action potential shape such

that the plateau phase becomes sustained (Waxman, 2001). Both could cause a

greater number of VDCCs to be open.

Recently there has been a reappraisal in the effectiveness o f traditional

analgesics such as morphine. Historically, neuropathic pain has been labelled

unresponsive to opioids, yet now it appears that this can be overcome in some

situations by dose escalation (Portenoy et al., 1990; Jadad et al., 1992) and thus the

opioids have been redeemed as useful for neuropathic pain control regimens.

Unfortunately, they are not always effective in all patients and dose titration can be

hindered by unpleasant side-effects. Treatment, more often than not, involves the

utilization o f other compounds that are not conventionally recognized as analgesics.

Compounds originally developed for uses other than pain management, such as the

antidepressants and anticonvulsants, have new-found roles as analgesics. The use of

meta-analyses of all currently used mainline drugs shows that have almost identical

efficacy (Sindrup & Jensen, 1999). Of current interest is the antiepileptic gabapentin.

193

thought to have a unique interaction with VDCCs through binding to the 0028 subunit,

which has been proven to provide relief from neuropathic pains (Backonja et al.,

1998; Rowbotham et a l, 1998) and is supported by a mounting body of preclinical

data. Although the pain relief experienced by neuropathic patients with gabapentin

appears to be no greater than that mediated by antidepressants or anticonvulsants,

gabapentin is renowned for its minimal side-effects and good tolerability, lack o f

toxicity and lack o f interaction with other drugs. Thus, although perhaps no more

efficacious than more established therapies, gabapentin’s worthy characteristic

profile makes it a preferred analgesic for what can often be a long-term management

of neuropathic pain. As yet, gabapentin can not considered a first-line treatment for

neuropathic pain (Hemstreet & Lapointe, 2001), however its lack of pharmacokinetic

interactions does make it an excellent candidate for add-on therapy where

monotherapy is inadequate. Chapter 7 clearly demonstrates the benefits o f

combination therapy, whereby utilizing low systemic doses o f both morphine and

gabapentin, that alone mediate negligible inhibitory effects on dorsal horn neuronal

responses, together mediated substantial inhibitions. In particular, in the presence o f

a neuropathic pain state, in contrast to gabapentin the inhibitory effects of morphine

alone were reduced in comparison to normal, yet in this pathological scenario in

conjunction with gabapentin, the most marked inhibitions were observed. Since

gabapentin and morphine are clinically licensed drugs, low doses and convenient

routes o f administration were used, possibly offering advantageous side-effect

profiles, translation of this combined therapy into the clinic deserves investigation

for the treatment of neuropathic pain.

Progress in the clinical management of neuropathic pain requires both more

detailed descriptions of experience o f the clinical symptoms and effects o f existing

drugs from the clinical setting, alongside improved knowledge o f the underlying

dysfunctional mechanisms derived from basic scientific research. This should help

better define mechanisms and drug targets involved in particular neuropathic pain

syndromes. Hopefully this thesis has substantiated or even illuminated potential

paths towards this goal.

194

REFERENCES

195

Abbadie C, Brown J L, Mantyh P W, Basbaum A I (1996) Spinal cord substance P receptor immunoreactivity increases in both inflammatory and nerve injury models of persistent pain. Neuroscience 70: 201-9.

Abdi S, Lee D H, Chung J M (1998) The anti-allodynic effects o f amitriptyline, gabapentin, and lidocaine in a rat model of neuropathic pain. Anesth Anal g 87: 1360- 6 .

Abdulla F A, Smith P A (1997) Ectopic alphal-adrenoceptors couple to N-type Ca2+ channels in axotomized rat sensory neurons. JNeurosci 17: 1633-41.

Abdulla F A, Smith P A (1998) Axotomy reduces the effect o f analgesic opioids yet increases the effect of nociceptin on dorsal root ganglion neurons. J Neurosci 18: 9685-94.

Ahlgren S C, Levine J D (1993) Mechanical hyperalgesia in streptozotocin-diabetic rats is not sympathetically maintained. Brain Res 616: 171-5.

Ahlgren S C, Levine J D (1994) Protein kinase C inhibitors decrease hyperalgesia and C-fiber hyperexcitability in the streptozotocin-diabetic rat. J Neurophysiol 72: 684-92.

Ahlijanian M K, Westenbroek R E, Catterall W A (1990) Subunit structure and localization o f dihydropyridine-sensitive calcium channels in mammalian brain, spinal cord, and retina. Neuron 4: 819-32.

Akaike N (1991) T-type calcium channel in mammalian CNS neurones. Comp Biochem Physiol A Physiol 98: 31-40.

Alden K J, Garcia J (2001) Differential effect of gabapentin on neuronal and muscle calcium currents. J Pharmacol Exp Ther 297: 727-35.

Alexander S P H (1998) Receptor and ion channel nomenclature supplement. Trends Pharmacol Sci: 23-27.

Alvarez F J, Kavookjian A M, Light A R (1992) Synaptic interactions between GABA-immunoreactive profiles and the terminals o f functionally defined myelinated nociceptors in the monkey and cat spinal cord. JNeurosci 12: 2901-17.

Amir R, Devor M (1992) Axonal cross-excitation in nerve-end neuromas: comparison o f A- and C-fibers. J Neurophysiol 68: 1160-6.

Amir R, Devor M (1997) Spike-evoked suppression and burst patterning in dorsal root ganglion neurons of the rat. J Physiol 501: 183-96.

Antkiewicz-Michaluk L, Michaluk J, Romanska I, Vetulani J (1993) Reduction of morphine dependence and potentiation of analgesia by chronic co-administration of nifedipine. Psychopharmacology (Berl) 111: 457-64.

196

Ardid D, Guilbaud G (1992) Antinociceptive effects o f acute and 'chronic' injections o f tricyclic antidepressant drugs in a new model of mononeuropathy in rats. Pain 49: 279-87.

Amer S, Meyerson B A (1988) Lack o f analgesic effect of opioids on neuropathic and idiopathic forms of pain. Pain 33: 11-23.

Attal N, Desmeules J, Kayser V (1991). Effects o f opioids in a rat model o f peripheral mononeuropathy. In Lesions ofprim ary afferent fibres as a tool for the study o f clinical pain. ed. Besson, J.M. & Guilbaud, G. pp. 245-258. Amsterdam: Elsevier.

Attal N, Jazat F, Kayser V, Guilbaud G (1990) Further evidence for 'pain-related' behaviours in a model of unilateral peripheral mononeuropathy. Pain 41: 235-51.

Averill S, McMahon S B, Clary D O, Reichardt L F, Priestley J V (1995) Immunocytochemical localization o f trkA receptors in chemically identified subgroups of adult rat sensory neurons. Eur JNeurosci 7: 1484-94.

Backonja M, Arndt G, Gombar K A, Check B, Zimmermann M (1994) Response o f chronic neuropathic pain syndromes to ketamine: a preliminary study. Pain 56: 51-7.

Backonja M, Beydoun A, Edwards K R, Schwartz S L, Fonseca V, Hes M, LaMoreaux L, Garofalo E (1998) Gabapentin for the symptomatic treatment of painful neuropathy in patients with diabetes mellitus: a randomized controlled trial. Jama 280: 1831-6.

Backonja M M, Miletic G, Miletic V (1995) The effect o f continuous morphine analgesia on chronic thermal hyperalgesia due to sciatic constriction injury in rats. Neurosci Lett 196: 61-4.

Bading H, Ginty D D, Greenberg M E (1993) Regulation o f gene expression in hippocampal neurons by distinct calcium signaling pathways. Science 260: 181-6.

Barhanin J, Schmid A, Lazdunski M (1988) Properties o f structure and interaction of the receptor for omega-conotoxin, a polypeptide active on Ca2+ channels. Biochem Biophys Res Commun 150: 1051-62.

Barrett C F, Rittenhouse A R (2000) Modulation of N-type calcium channel activity by G-proteins and protein kinase C .J Gen Physiol 115: 277-86.

Basbaum A I, Fields H L (1984) Endogenous pain control systems: brainstem spinal pathways and endorphin circuitry. Annu Rev Neurosci 7: 309-38.

Basbaum A I, Gautron M, Jazat F, Mayes M, Guilbaud G (1991) The spectrum of fiber loss in a model of neuropathic pain in the rat: an electron microscopic study. Pain 47: 359-67.

Battaglia G, Rustioni A (1988) Coexistence o f glutamate and substance P in dorsal root ganglion neurons of the rat and monkey. J Com/» Neurol 277: 302-12.

197

Bean B P (1984) Nitrendipine block o f cardiac calcium channels: high-afflnity binding to the inactivated state. Proc Natl Acad Sci U SA 81: 6388-92.

Bean B P (1989a) Classes of calcium channels in vertebrate cells. Annu Rev Physiol 51:367-84.

Bean B P (1989b) Multiple types o f calcium channels in heart muscle and neurons. Modulation by drugs and neurotransmitters. Ann N Y Acad Sci 560: 334-45.

Bean B P, Mintz I M, Regan L J, Sah D W Y, Adams M E (1993). Pharmacology of high threshold calcium channels in rat neurons. In Antagonists in the CNS. ed. Scriabine, A., Janis, R.A., Triggle, D.H. & Ferranye, G.K. pp. 61-69. Branford: Neva Press.

Bech-Hansen N T, Naylor M J, Maybaum T A, Pearce W G, Koop B, Fishman G A, Mets M, Musarella M A, Boycott K M (1998) Loss-of-function mutations in a calcium-channel alpha 1-subunit gene in Xp 11.23 cause incomplete X-linked congenital stationary night blindness. Nat Genet 19: 264-7.

Behbehani M M, Dollberg-Stolik O (1994) Partial sciatic nerve ligation results in an enlargement of the receptive field and enhancement of the response of dorsal horn neurons to noxious stimulation by an adenosine agonist. Pain 58: 421-8.

Benedetti F, Vighetti S, Amanzio M, Casadio C, Oliaro A, Bergamasco B, Maggi G (1998) Dose-response relationship o f opioids in nociceptive and neuropathic postoperative pain. Pain 74: 205-11.

Bennett D L, French J, Priestley J V, McMahon S B (1996) NGF but not NT-3 or BDNF prevents the A fiber sprouting into lamina II o f the spinal cord that occurs following axotomy. Mol Cell Neurosci 8:211-20.

Bennett G J, Xie Y K (1988) A peripheral mononeuropathy in rat that produces disorders o f pain sensation like those seen in man. Pain 33: 87-107.

Bernardi P S, Valtschanoff J G, Weinberg R J, Schmidt H H, Rustioni A (1995) Synaptic interactions between primary afferent terminals and GABA and nitric oxide-synthesizing neurons in superficial laminae o f the rat spinal cord. J Neurosci 15:1363-71.

Bernstein M A, Welch S P (1995) Alterations in L-type calcium channels in the brain and spinal cord of acutely treated and morphine-tolerant mice. Brain Res 696: 83-8.

Bertolino M, Llinas R R (1992) The central role of voltage-activated and receptor- operated calcium channels in neuronal cells. Annual Review o f Pharmacology & Toxicology 32: 399-421.

Besse D, Lombard M C, Besson J M (1991) Autoradiographic distribution o f mu, delta and kappa opioid binding sites in the superficial dorsal horn, over the rostrocaudal axis of the rat spinal cord. Brain Res 548: 287-91.

198

Besse D, Lombard M C, Zajac J M, Roques B P, Besson J M (1990a) Pre- and postsynaptic distribution o f mu, delta and kappa opioid receptors in the superficial layers o f the cervical dorsal horn of the rat spinal cord. Brain Res 521: 15-22.

Besse D, Lombard M C, Zajac J M, Roques B P, Besson J M (1990b) Pre- and postsynaptic location o f mu, delta and kappa opioid receptors in the superficial layers of the dorsal horn of the rat spinal cord. Prog Clin Biol Res 328: 183-6.

Besse D, Weil-Fugazza J, Lombard M C, Butler S H, Besson J M (1992) Monoarthritis induces complex changes in mu-, delta- and kappa-opioid binding sites in the superficial dorsal horn of the rat spinal cord. Eur J Pharmacol 223: 123-31.

Besson J M, Besse D, Lombard M C, Attal N, Desmeules J, Guilbaud G (1993). Opioids and neuropathic pain: experimental approaches in the rat. In Current and emerging issues in cancer pain, research and practice, ed. Chapman, C.R. & Foley, K.M. pp. 129-152. New York: Raven Press.

Besson J M, Chaouch A (1987) Peripheral and spinal mechanisms o f nociception. Physiol Rev 67: 67-186.

Bester H, Chapman V, Besson J M, Bernard J F (2000) Physiological properties of the lamina I spinoparabrachial neurons in the rat. J Neurophysiol 83: 2239-59.

Bezprozvanny I, Tsien R W (1995) Voltage-dependent blockade of diverse types of voltage-gated Ca2+ channels expressed in Xenopus oocytes by the Ca2+ channel antagonist mibefradil (Ro 40-5967). Mol Pharmacol 48: 540-9.

Bian D, Nichols M L, Ossipov M H, Lai J, Porreca F (1995) Characterization of the antiallodynic efficacy o f morphine in a model o f neuropathic pain in rats. Neuroreport 6\ 1981-4.

Birder L A, Perl E R (1999) Expression of alpha2-adrenergic receptors in rat primary afferent neurones after peripheral nerve injury or inflammation. J Physiol 515: 533- 42.

Bito H, Deisseroth K, Tsien R W (1996) CREB phosphorylation and dephosphorylation: a Ca(2+)- and stimulus duration-dependent switch for hippocampal gene expression. Cell 87: 1203-14.

Black J A, Cummins T R, Plumpton C, Chen Y H, Hormuzdiar W, Clare J J, Waxman S G (1999) Upregulation of a silent sodium channel after peripheral, but not central, nerve injury in DRG neurons. J Neurophysiol 82: 2776-85.

Black J L, 3rd, Lennon V A (1999) Identification and cloning o f putative human neuronal voltage-gated calcium channel gamma-2 and gamma-3 subunits: neurologic implications. Mayo Clin Proc 74: 357-61.

Boldt J, von Bormann B, Kling D, Russ W, Ratthey K, Hempelmann G (1987) Low- dose fentanyl analgesia modified by calcium channel blockers in cardiac surgery. Eur JAnaesthesiol 4: 387-94.

199

Bonfanti L, Bellardi S, Ghidella S, Gobetto A, Polak J M, Merighi A (1991) Distribution o f five peptides, three general neuroendocrine markers, and two synaptic-vesicle-associated proteins in the spinal cord and dorsal root ganglia of the adult and newborn dog: an immunocytochemical study. Am JAnat 191: 154-66.

Bonica J J (1979). Causalgia and other sympathetic dystrophies. In Advances in Pain Research and Therapy, ed. Bonica, J.J. pp. 146-166. New York: Raven Press.

Bonnot A, Corio M, Tramu G, Viala D (1996) Immunocytochemical distribution of ionotropic glutamate receptor subunits in the spinal cord of the rabbit. J Chem Neuroanat 11: 267-78.

Bossu J L, Feltz A (1986) Inactivation o f the low-threshold transient calcium current in rat sensory neurones: evidence for a dual process. J Physiol 376: 341-57.

Bossu J L, Feltz A, Thomann J M (1985) Depolarization elicits two distinct calcium currents in vertebrate sensory neurones. Pflugers Archiv - European Journal o f Physiology 403: 360-8.

Bourgoin S, Pohl M, Benoliel J J, Mauborgne A, Collin E, Hamon M, Cesselin F (1992) gamma-Aminobutyric acid, through GAB A A receptors, inhibits the potassium-stimulated release of calcitonin gene-related peptide- but not that o f substance P-like material from rat spinal cord slices. Brain Res 583: 344-8.

Bowersox S S, Gadbois T, Singh T, Pettus M, Wang Y X, Luther R R (1996) Selective N-type neuronal voltage-sensitive calcium channel blocker, SNX-111, produces spinal antinociception in rat models o f acute, persistent and neuropathic pain. J Pharmacol Exp Ther 279: 1243-9.

Bowsher D (1991) Neurogenic pain syndromes and their management. Br Med Bull 47: 644-66.

Branchek T A, Smith K E, Gerald C, Walker M W (2000) Galanin receptor subtypes. Trends Pharmacol Sci 21 ; 109-17.

Brenan A (1986) Collateral reinnervation of skin by C-fibres following nerve injury in the rat. Brain Res 385: 152-5.

Brickley K, Campbell V, Berrow N, Leach R, Norman R I, Wray D, Dolphin A C, Baldwin S A (1995) Use of site-directed antibodies to probe the topography of the alpha 2 subunit of voltage-gated Ca2+ channels. FEBS Lett 364: 129-33.

Brose W G, Gutlove D P, Luther R R, Bowersox S S, McGuire D (1997) Use of intrathecal SNX-111, a novel, N-type, voltage-sensitive, calcium channel blocker, in the management of intractable brachial plexus avulsion pain. Clin J Pain 13: 256-9.

Brose W G, Pfeiffer B L, Hassenbusch S J, Burchiel K J, Byas-Smith M, Krames E, McGuire D, Tich N, Luther R R (1996). Analgesia produced by SNX-111 in patients with morphine-resistant pain. In 15th Annual Scientific Meeting o f the Amereican Pain Society, pp. A122. Washington.

200

Brust P F, Simerson S, McCue A F, Deal C R, Schoonmaker S, Williams M E, Velicelebi G, Johnson E C, Harpold M M, Ellis S B (1993) Human neuronal voltage- dependent calcium channels: studies on subunit structure and role in channel assembly. Neuropharmacology 32: 1089-102.

Bryans J S, Davies N, Gee N S, Dissanayake V U, Ratcliffe G S, Horwell D C, Kneen C O, Morrell A I, Oles R J, O'Toole J C, Perkins G M, Singh L, Suman- Chauhan N, O'Neill J A (1998) Identification of novel ligands for the gabapentin binding site on the alpha2delta subunit o f a calcium channel and their evaluation as anticonvulsant agents. J Med Chem 41: 1838-45.

Burchiel K J (1980) Ectopic impulse generation in focally demyelinated trigeminal nerve. Exp Neurol 69: 423-9.

Burchiel K J (1984) Effects of electrical and mechanical stimulation on two foci of spontaneous activity which develop in primary afferent neurons after peripheral axotomy. Pain 18: 249-65.

Burke S P, Adams M E, Taylor C P (1993) Inhibition of endogenous glutamate release from hippocampal tissue by Ca2+ channel toxins. Eur J Pharmacol 238: 383- 6 .

Calcutt N A, Chaplan S R (1997) Spinal pharmacology o f tactile allodynia in diabetic rats. Br J Pharmacol 122: 1478-82.

Cameron A A, d iffer K D, Dougherty P M, Garrison C J, Willis W D, Carlton S M (1997) Time course of degenerative and regenerative changes in the dorsal horn in a rat model of peripheral neuropathy. J Comp Neurol 379: 428-42.

Caraceni A, Zecca E, Martini C, De Conno F (1999) Gabapentin as an adjuvant to opioid analgesia for neuropathic cancer pain. J Pain Symptom Manage 17: 441-5.

Carbone E, Lux H D (1984) A low voltage-activated, fully inactivating Ca channel in vertebrate sensory neurones. Nature 310: 501-2.

Carbone E, Lux H D (1987) Kinetics and selectivity of a low-voltage-activated calcium current in chick and rat sensory neurones. J Physiol 386: 547-70.

Carlton S M, Dougherty P M, Pover C M, Coggeshall R E (1991) Neuroma formation and numbers o f axons in a rat model o f experimental peripheral neuropathy. Neurosci Lett 131: 88-92.

Carpenter K J, Dickenson A H (1998) Evidence that [Phel psi(CH2- NH)Gly2]nociceptin-(l-13)-NH2, a peripheral ORL-1 receptor antagonist, acts as an agonist in the rat spinal cord. Br J Pharmacol 125: 949-51.

Carta F, Bianchi M, Argenton S, Cervi D, Marolla G, Tamburini M, Breda M, Fantoni A, Panerai A E (1990) Effect of nifedipine on morphine-induced analgesia. AnesthAnalg 70: 493-8.

201

Castillo P E, Weisskopf M G, Nicoll R A (1994) The role o f Ca2+ channels in hippocampal mossy fiber synaptic transmission and long-term potentiation. Neuron 12:261-9.

Castro-Lopes J M, Coimbra A, Grant G, Arvidsson J (1990) Ultrastructural changes o f the central scalloped (C l) primary afferent endings o f synaptic glomeruli in the substantia gelatinosa Rolandi of the rat after peripheral neurotomy. JNeurocytol 19: 329-37.

Castro-Lopes J M, Malcangio M, Pan B H, Bowery N G (1995) Complex changes o f GABAA and GABAB receptor binding in the spinal cord dorsal horn following peripheral inflammation or neurectomy. Brain Res 679: 289-97.

Castro-Lopes J M, Tavares I, Coimbra A (1993) G ABA decreases in the spinal cord dorsal horn after peripheral neurectomy. Brain Res 620: 287-91.

Catheline G, Le Guen S, Besson J M (2001) Intravenous morphine does not modify dorsal horn touch-evoked allodynia in the mononeuropathic rat: a Eos study. Pain 92: 389-98.

Cesselin F, Bourgoin S, Artaud F, Hamon M (1984) Basic and regulatory mechanisms o f in vitro release o f Met-enkephalin from the dorsal zone of the rat spinal cord. JNeurochem 43: 763-74.

Chaplan S R, Malmberg A B, Yaksh T L (1997) Efficacy of spinal NMDA receptor antagonism in formalin hyperalgesia and nerve injury evoked allodynia in the rat. J Pharmacol Exp Ther 280: 829-38.

Chaplan S R, Pogrel J W, Yaksh T L (1994) Role o f voltage-dependent calcium channel subtypes in experimental tactile allodynia. J Pharmacol Exp Ther 269: 1117- 23.

Chapman V (1999) The cannabinoid CBl receptor antagonist, SR141716A, selectively facilitates nociceptive responses o f dorsal horn neurones in the rat. Br J Pharmacol 127: 1765-7.

Chapman V, Dickenson A H (1992) The combination of NMDA antagonism and morphine produces profound antinociception in the rat dorsal horn. Brain Res 573: 321-3.

Chapman V, Suzuki R, Chamarette H L, Rygh L J, Dickenson A H (1998a) Effects of systemic carbamazepine and gabapentin on spinal neuronal responses in spinal nerve ligated rats. Pain 75: 261-72.

Chapman V, Suzuki R, Dickenson A H (1998b) Electrophysiological characterization of spinal neuronal response properties in anaesthetized rats after ligation o f spinal nerves L5-L6. J Physiol 507: 881-94.

Chemy N I, Thaler H T, Friedlander-Klar H, Lapin J, Foley K M, Houde R, Portenoy R K (1994) Opioid responsiveness of cancer pain syndromes caused by neuropathic

202

or nociceptive mechanisms: a combined analysis o f controlled, single-dose studies. Neurology 857-61.

Cho H J, Kim D S, Lee N H, Kim J K, Lee K M, Han K S, Kang Y N, Kim K J (1997) Changes in the alpha 2-adrenergic receptor subtypes gene expression in rat dorsal root ganglion in an experimental model o f neuropathic pain. Neuroreport 8: 3119-22.

Choi Y, Yoon Y W, Na H S, Kim S H, Chung J M (1994) Behavioral signs o f ongoing pain and cold allodynia in a rat model o f neuropathic pain. Pain 59: 369-76.

Christensen D, Idanpaan-Heikkila J J, Guilbaud G, Kayser V (1998) The antinociceptive effect o f combined systemic administration o f morphine and the glycine/NMDA receptor antagonist, (+)-HA966 in a rat model o f peripheral

. Br J Pharmacol 125: 1641-50.

Christie B R, Eliot L S, Ito K, Miyakawa H, Johnston D (1995) Different Ca2+ channels in soma and dendrites o f hippocampal pyramidal neurons mediate spike- induced Ca2+ influx. J Neurophysiol 73: 2553-7.

Chung J M, Huguenard J R, Prince D A (1993) Transient enhancement of low- threshold calcium current in thalamic relay neurons after corticectomy. J Neurophysiol 70: 20-7.

Chung K, Lee B H, Yoon Y W, Chung J M (1996) Sympathetic sprouting in the dorsal root ganglia of the injured peripheral nerve in a rat neuropathic pain model. J Comp Neurol 376: 241-52.

Chung K, Yoon Y W, Chung J M (1997) Sprouting sympathetic fibers form synaptic varicosities in the dorsal root ganglion of the rat with neuropathic injury. Brain Res 751:275-80.

Cizkova D, Marsala M, Stauderman K, Yaksh T L (1999). Calcium channel alB subunit in spinal cord/DRG of normal and nerve-injured rats. In 9th World Congress on Pain. pp. 134. Vienna: IAS? Press.

Coderre T J (1992) Contribution o f protein kinase C to central sensitization and persistent pain following tissue injury. Neurosci Lett 140: 181-4.

Coderre T J, Grimes R W, Melzack R (1986) Deafferentation and chronic pain in animals: an evaluation o f evidence suggesting autotomy is related to pain. Pain 26: 61-84.

Coderre T J, Katz J, Vaccarino A L, Melzack R (1993) Contribution of central neuroplasticity to pathological pain: review o f clinical and experimental evidence. Pain 52: 259-85.

Coderre T J, Vaccarino A L, Melzack R (1990) Central nervous system plasticity in the tonic pain response to subcutaneous formalin injection. Brain Res 535: 155-8.

203

Contreras E, Tamayo L, Amigo M (1988) Calcium channel antagonists increase morphine-induced analgesia and antagonize morphine tolerance. Eur J Pharmacol 148: 463-6.

Cooper C B, Amot M I, Feng Z P, Jarvis S E, Hamid J, Zamponi G W (2000) Cross­talk between G-protein and protein kinase C modulation o f N-type calcium channels is dependent on the G-protein beta subunit isoform. J Biol Chem 275: 40777-81.

Coulter D A, Huguenard J R, Prince D A (1989a) Calcium currents in rat thalamocortical relay neurones: kinetic properties of the transient, low-threshold current. J Physiol 414: 587-604.

Coulter D A, Huguenard J R, Prince D A (1989b) Characterization of ethosuximide reduction o f low-threshold calcium current in thalamic neurons. Ann Neurol 25: 582- 93.

Coulter D A, Huguenard J R, Prince D A (1989c) Specific petit mal anticonvulsants reduce calcium currents in thalamic neurons. Neurosci Lett 98: 74-8.

Craig A D, Dostrovsky J O (1997) Processing o f nociceptive information at supraspinal levels. In Anesthesia: Biologic Foundations, ed. Yaksh, T.L., Lynch III, C., Zapol, W.M., Maze, M., Biebuyck, J.F. & Saidman, L.J. pp. 625-642. Philadelphia: Lippincott-Raven Publishers.

Craig P J, Beattie R E, Folly E A, Banerjee M D, Reeves M B, Priestley J V, Carney S L, Sher E, Perez-Reyes E, Volsen S G (1999) Distribution o f the voltage- dependent calcium channel alpha 1G subunit mRNA and protein throughout the mature rat brain. Eur J Neurosci 11: 2949-64.

Cribbs L L, Lee J H, Yang J, Satin J, Zhang Y, Daud A, Barclay J, Williamson M P, Fox M, Rees M, Perez-Reyes E (1998) Cloning and characterization of alphalH from human heart, a member of the T-type Ca2+ channel gene family. Circ Res 83: 103-9.

Croul S, Radzievsky A, Sverstiuk A, Murray M (1998) N K l, NMDA, 5HTla, and 5HT2 receptor binding sites in the rat lumbar spinal cord: modulation following sciatic nerve crush. Exp Neurol 154: 66-79.

Curtis B M, Catterall W A (1984) Purification of the calcium antagonist receptor of the voltage-sensitive calcium channel from skeletal muscle transverse tubules. Biochemistry 23: 2113-8.

Dallocchio C, Buffa C, Mazzarello P, Chiroli S (2000) Gabapentin vs. amitriptyline in painful diabetic neuropathy: an open-label pilot study. J Pain Symptom Manage 20: 280-5.

Daniell L C, Barr E M, Leslie S W (1983) 45Ca2+ uptake into rat whole brain synaptosomes unaltered by dihydropyridine calcium antagonists. J Neurochem 41: 1455-9.

204

Darland T, Heinricher M M, Grandy D K (1998) Orphanin FQ/nociceptin: a role in pain and analgesia, but so much more. Trends Neurosci 21: 215-21.

Davies J, Watkins J C (1983) Role of excitatory amino acid receptors in mono- and polysynaptic excitation in the cat spinal cord. Exp Brain Res 49: 280-90.

deOroot J F, Coggeshall R E, Carlton S M (1997) The reorganization of mu opioid receptors in the rat dorsal horn following peripheral axotomy. Neurosci Lett 233: 113-6.

Deisseroth K, Bito H, Tsien R W (1996) Signaling from synapse to nucleus: postsynaptic CREB phosphorylation during multiple forms of hippocampal synaptic plasticity. Neuron 16: 89-101.

Del Pozo E, Caro G, Baeyens J M (1987) Analgesic effects o f several calcium channel blockers in mice. Eur J Pharmacol 137: 155-60.

Dellemijn P (1999) Are opioids effective in relieving neuropathic pain? Pain 80: 453-62.

Desarmenien M, Feltz P, Occhipinti G, Santangelo F, Schlichter R (1984) Coexistence o f GABAA and GABAB receptors on A delta and C primary afferents. Br J Pharmacol 81: 327-33.

Desmeules J A, Kayser V, Guilbaud G (1993) Selective opioid receptor agonists modulate mechanical allodynia in an animal model o f neuropathic pain. Pain 53: 277-85.

Devor M (1991) Neuropathic pain and injured nerve: peripheral mechanisms. Br M edBullAl: 619-30.

Devor M, Seltzer Z (1999). Pathophysiology o f damaged nerves in relation to chronic pain. In Textbook o f Pain. ed. Wall, P.D. & Melzack, R. pp. 129-164. London: Churchill Livingston.

Devor M, White D M, Goetzl E J, Levine J D (1992) Eicosanoids, but not tachykinins, excite C-fiber endings in rat sciatic nerve-end neuromas. Neuroreport 3: 21-4.

Diaz A, Dickenson A H (1997) Blockade of spinal N- and P-type, but not L-type, calcium channels inhibits the excitability o f rat dorsal horn neurones produced by subcutaneous formalin inflammation. Pain 69: 93-100.

Diaz A, Ruiz F, Florez J, Pazos A, Hurle M A (1995) Regulation of dihydropyridine- sensitive Ca++ channels during opioid tolerance and supersensitivity in rats. J Pharmacol Exp Ther 274: 1538-44.

Dib-Hajj S D, Fjell J, Cummins T R, Zheng Z, Fried K, LaMotte R, Black J A, Waxman S G (1999) Plasticity of sodium channel expression in DRG neurons in the chronic constriction injury model of neuropathic pain. Pain 83: 591-600.

205

Dickenson A H (1994). NMDA receptor antagonists as analgesics. In Pharmacological approaches to the treatment o f chronic pain: New concepts and critical issues, ed. Fields, H.L. & Liebeskind, J.C. pp. 173-187. Seattle: lASP.

Dickenson A H (1995) Spinal cord pharmacology of pain. Br J Anaesth 75: 193-200.

Dickenson A H (1997) NMDA receptor antagonists: interactions with opioids. Acta Anaesthesiol Scand 41:112-5.

Dickenson A H, Chapman V, Green G M (1997a) The pharmacology of excitatory and inhibitory amino acid-mediated events in the transmission and modulation of pain in the spinal cord. Gen Pharmacol 28: 633-8.

Dickenson A H, Matthews E A, Suzuki R (2001). Central nervous system mechanisms o f pain in peripheral neuropathy. In Progress in Pain Research and Management. Neuropathic pain: pathophysiology and treatment, ed. Hansson, P.T., Fields, H.L., Hill, R.G. & Marchettini, P. pp. In press. Seattle: lASP Press.

Dickenson A H, Stanfa L C, Chapman V, Yaksh T L (1997b). Response properties of dorsal horn neurons: pharmacology o f the dorsal horn. In Anesthesia: Biologic Foundations, ed. Yaksh, T.L., Lynch III, C., Zapol, W.M., Maze, M., Biebuyck, J.F. & Saidman, L.J. pp. 611-624. Philadelphia: Lippincott-Raven Publishers.

Dickenson A H, Sullivan A F (1986) Electrophysiological studies on the effects of intrathecal morphine on nociceptive neurones in the rat dorsal horn. Pain 24: 211-22.

Dickenson A H, Sullivan A F (1987) Evidence for a role of the NMDA receptor in the frequency dependent potentiation o f deep rat dorsal horn nociceptive neurones following C fibre stimulation. Neuropharmacology 26: 1235-8.

Dickenson A H, Suzuki R (1999). Function and dysfunction of opioid receptors in the spinal cord. In Progress in pain research and management, ed. Kalso, E., McQuay, H. & Wiesenfeld-Hallin, Z. pp. 17-44. Seattle: lASP Press.

Dickenson A H, Suzuki R, Reeve A J (2000) Adenosine as a potential analgesic target in inflammatory and neuropathic pains. CNSDrugs 13: 77-85.

Dickie B G, Davies J A (1992) Calcium channel blocking agents and potassium- stimulated release of glutamate from cerebellar slices. Eur J Pharmacol 229: 97-9.

Dickins'on T, Fleetwood-Walker S M (1999) VIP and PACAP: very important in pain? Trends Pharmacol Sci 20: 324-9.

Doubelll T P, Mannion R J, W oolf C J (1997) Intact sciatic myelinated primary afferent terminals collaterally sprout in the adult rat dorsal horn following section of a neighbouring peripheral nerve. J Comp Neurol 380: 95-104.

Drew L J, Harris J, Millns P J, Kendall D A, Chapman V (2000) Activation of spinal cannabimoid 1 receptors inhibits C-fibre driven hyperexcitable neuronal responses and increases [3 5 S]GTPgammaS binding in the dorsal horn o f the spinal cord of noninflaimed and inflamed rats. Eur J Neurosci 12: 2079-86.

206

Dubner R (1991). Neuronal plasticity in the spinal and medullary dorsal horns: a possible role in central pain mechanisms. In Pain and Central Nervous Disease: the Central Pain Syndromes, ed. Caser, K.L. pp. 143-155. New York: Raven Press.

Dubner R, Basbaum A (1994). Spinal dorsal horn plasticity following tissue or nerve injury. In Textbook o f Pain. ed. Wall, P.D. & Melzack, R. pp. 225-241. Edinburgh: Churchill Livingstone.

Dubner R, Bennett G J (1983) Spinal and trigeminal mechanisms of nociception. Annu Rev Neurosci 6: 381-418.

Dubner R, Kenshalo D R, Jr., Maixner W, Bushnell M C, Oliveras J L (1989) The correlation o f monkey medullary dorsal horn neuronal activity and the perceived intensity of noxious heat stimuli. J Neurophysiol 62: 450-7.

Duggan A W, Hendry I A, Morton C R, Hutchison W D, Zhao Z Q (1988) Cutaneous stimuli releasing immunoreactive substance P in the dorsal horn o f the cat. Brain Res 451: 261-73.

Duggan A W, North R A (1984) Electrophysiology o f opioids. Pharmacology Review 35: 219-281.

Eberst R, Dai S, Klugbauer N, Hofmann F (1997) Identification and functional characterization of a calcium channel gamma subunit. Pflugers Arch 433: 633-7.

Eckhardt K, Ammon S, Hofmann U, Riebe A, Gugeler N, Mikus G (2000) Gabapentin enhances the analgesic effect o f morphine in healthy volunteers. Anesth Analg 91: 185-91.

Eide P K, Jorum E, Stubhaug A, Bremnes J, Breivik H (1994) Relief of post-herpetic neuralgia with the N-methyl-D-aspartic acid receptor antagonist ketamine: a double­blind, cross-over comparison with morphine and placebo. Pain 58: 347-54.

Eisenach J C, DuPen S, Dubois M, Miguel R, Allin D (1995) Epidural clonidine analgesia for intractable cancer pain. The Epidural Clonidine Study Group. Pain 61: 391-9.

Eisenberg E, Alon N, Ishay A, Daoud D, Yarnitsky D (1998) Lamotrigine in the treatment of painful diabetic neuropathy. Eur J Neurol 5: 167-173.

Ellis S B, Williams M E, Ways N R, Brenner R, Sharp A H, Leung A T, Campbell K P, McKenna E, Koch W J, Hui A, et al. (1988) Sequence and expression of mRNAs encoding the alpha 1 and alpha 2 subunits o f a DHP-sensitive calcium channel. Science 241: 1661-4.

Ertel E A, Campbell K P, Harpold M M, Hofmann F, Mori Y, Perez-Reyes E, Schwartz A, Snutch T P, Tanabe T, Bimbaumer L, Tsien R W, Catterall W A (2000) Nomenclature of voltage-gated calcium channels. Neuron 25: 533-5.

207

Fedulova S A, Kostyuk P G, Veselovsky N S (1985) Two types of calcium channels in the somatic membrane of new-born rat dorsal root ganglion neurones. J Physiol 359:431-46.

Field M J, Bramwell S, Hughes J, Singh L (1999a) Detection o f static and dynamic components o f mechanical allodynia in rat models o f neuropathic pain: are they signalled by distinct primary sensory neurones? Pain 83: 303-11.

Field M J, Holloman E F, McCleary S, Hughes J, Singh L (1997a) Evaluation of gabapentin and S-(+)-3-isobutylgaba in a rat model o f postoperative pain. J Pharmacol Exp Ther 282: 1242-6.

Field M J, Hughes J, Singh L (2000) Further evidence for the role o f the alpha(2)delta subunit o f voltage dependent calcium channels in models o f neuropathic pain. Br J Pharmacol 131: 282-6.

Field M J, McCleary S, Hughes J, Singh L (1999b) Gabapentin and pregabalin, but not morphine and amitriptyline, block both static and dynamic components of mechanical allodynia induced by streptozocin in the rat. Pain 80: 391-8.

Field M J, Oles R J, Lewis A S, McCleary S, Hughes J, Singh L (1997b) Gabapentin (neurontin) and S-(+)-3-isobutylgaba represent a novel class o f selective antihyperalgesic agents. .SrJP/zarwûfco/121: 1513-22.

Fields H L, Rowbotham M, Baron R (1998) Postherpetic neuralgia: irritable nociceptors and deafferentation. Neurobiol Dis 5: 209-27.

Filos K S, Goudas L C, Patroni O, Tassoudis V (1993) Analgesia with epidural nimodipine [letter]. Lancet 342: 1047.

Fink K, Meder W, Dooley D J, Gothert M (2000) Inhibition of neuronal Ca(2+) influx by gabapentin and subsequent reduction o f neurotransmitter release from rat neocortical slices. Br J Pharmacol 130: 900-6.

Foley K M, Inturrisi C E (1987) Analgesic drug therapy in cancer pain: principles and practice. Med Clin North Am 71: 207-32.

Foley K M, Portenoy R K (1991). The role of opioid analgesics in neuropathic pain. In Lesions o f primary afferent fibres as a tool for the study o f the clinical pain. ed. Besson, J.M. & Guilbaud, G. pp. 277-292. Amsterdam: Elsevier.

Fox A, Kesingland A, Gentry C, McNair K, Patel S, Urban L, James I (2001) The role of central and peripheral Cannabinoid 1 receptors in the antihyperalgesic activity of cannabinoids in a model of neuropathic pain. Pain 92: 91-100.

Fox A J (1999). A comparative discussion of Ad and C fibres in different tissues. In Pain and Neurogenic Inflammation, ed. Brain, S.D. & Moore, P.K. pp. 1-22. Basel: Birkhauser Verlag.

Frayer S M, Barber L A, Vasko M R (1999) Activation of protein kinase C enhances peptide release from rat spinal cord slices. Neurosci Lett 265: 17-20.

208

Fundin B T, Silos-Santiago I, Emfors P, Fagan A M, Aldskogius H, DeChiara T M, Phillips H S, Barbacid M, Yancopoulos G D, Rice F L (1997) Differential dependency o f cutaneous mechanoreceptors on neurotrophins, trk receptors, and P75 LNGFR. Dev Biol 190: 94-116.

Garry M G, Tanelian D.L. (1997). Afferent activity in injured afferent nerves. In Anesthesia: Biologic Foundations, ed. Yaksh, T.L., Lynch III, C., Zapol, W.M., Maze, M., Biebuyck, J.F. & Saidman, L.J. pp. 531-542. Philadelphia: Lippincott- Raven Publishers.

Gee N S, Brown J P, Dissanayake V U, Offord J, Thurlow R, Woodruff G N (1996) The novel anticonvulsant drug, gabapentin (Neurontin), binds to the alpha2delta subunit of a calcium channel. JB iol Chem 271: 5768-76.

Ghilardi J R, Allen C J, Vigna S R, McVey D C, Mantyh P W (1992) Trigeminal and dorsal root ganglion neurons express CCK receptor binding sites in the rat, rabbit, and monkey: possible site o f opiate-CCK analgesic interactions. J Neurosci 12: 4854-66.

Gibson S J, Polak J M, Bloom S R, Wall P D (1981) The distribution of nine peptides in rat spinal cord with special emphasis on the substantia gelatinosa and on the area around the central canal (lamina X). J Comp Neurol 201: 65-79.

Glasser S P (1998) The relevance of T-type calcium antagonists: a profile of mibefradil. J Clin Pharmacol 38: 659-69.

Go V L, Yaksh T L (1987) Release of substance P from the cat spinal cord. J Physiol 391: 141-67.

Goa K L, Sorkin E M (1993) Gabapentin. A review o f its pharmacological properties and clinical potential in epilepsy. Drugs 46: 409-27.

Goadsby P J (2000) The pharmacology of headache. Prog Neurobiol 62: 509-25.

Gohil K, Bell J R, Ramachandran J, Miljanich G P (1994) Neuroanatomical distribution o f receptors for a novel voltage-sensitive calcium-channel antagonist, SNX-230 (omega-conopeptide MVIIC). Brain Res 653: 258-66.

Goldlust A, Su T Z, Welty D F, Taylor C P, Oxender D L (1995) Effects of anticonvulsant drug gabapentin on the enzymes in metabolic pathways o f glutamate and GAB A. Epilepsy Res 22: 1-11.

Gotz E, Feuerstein T J, Lais A, Meyer D K (1993) Effects of gabapentin on release of gamma-aminobutyric acid from slices of rat neostriatum. Arzneimittelforschung 43:636-8.

Govrin-Lippmann R, Devor M (1978) Ongoing activity in severed nerves: source and variation with time. Brain Res 159: 406-10.

209

Gracely R H, Lynch S A, Bennett G J (1992) Painful neuropathy: altered central processing maintained dynamically by peripheral input. Pain 51: 175-94.

Gruner W, Silva L R (1994) Omega-conotoxin sensitivity and presynaptic inhibition of glutamatergic sensory neurotransmission in vitro. J Neurosci 14: 2800-8.

Gumett C A, De Waard M, Campbell K P (1996) Dual function o f the voltage- dependent Ca2+ channel alpha 2 delta subunit in current stimulation and subunit interaction. Neuron 16: 431-40.

Hagiwara N, Irisawa H, Kameyama M (1988) Contribution of two types of calcium currents to the pacemaker potentials of rabbit sino-atrial node cells. J Physiol 395: 233-53.

Haley J E, Sullivan A F, Dickenson A H (1990) Evidence for spinal N-methyl-D- aspartate receptor involvement in prolonged chemical nociception in the rat. Brain Rgj 518: 218-26.

Hamid J, Nelson D, Spaetgens R, Dubel S J, Snutch T P, Zamponi G W (1999) Identification o f an integration center for cross-talk between protein kinase C and G protein modulation of N-type calcium channels. J Biol Chem 274: 6195-202.

Hao J X, Xu X J, Aldskogius H, Seiger A, Wiesenfeld-Hallin Z (1992a) Photochem ically induced transient spinal ischem ia induces behavioral hypersensitivity to mechanical and cold stimuli, but not to noxious-heat stimuli, in the rat. Exp Neurol 118:187-94.

Hao J X, Xu X J, Yu Y X, Seiger A, Wiesenfeld-Hallin Z (1992b) Baclofen reverses the hypersensitivity of dorsal horn wide dynamic range neurons to mechanical stimulation after transient spinal cord ischemia; implications for a tonic GABAergic inhibitory control of myelinated fiber input. J Neurophysiol 68: 392-6.

Harris J A, Corsi M, Quartaroli M, Arban R, Bentivoglio M (1996) Upregulation of spinal glutamate receptors in chronic pain. Neuroscience 74: 7-12.

Hasegawa A E, Zacny J P (1997a) The influence of three L-type calcium channel blockers on morphine effects in healthy volunteers. Anesth Analg 85: 633-8.

Hasegawa A E, Zacny J P (1997b) The influence of three L-type calcium channel blockers on morphine effects in healthy volunteers. Anesth Analg 85: 633-8.

Hedley L R, Martin B, Waterbury L D, Clarke D E, Hunter J C (1995) A comparison of the action o f mexiletine and morphine in rodent models o f acute and chronic pain. Proc West Pharmacol Soc 38: 103-4.

Heinricher M (1997). Organizational characteristics o supraspinally mediated responses to nociceptive input. In Anesthesia: Biologic Foundations, ed. Yaksh, T.L., Lynch III, C., Zapol, W.M., Maze, M., Biebuyck, J.F. & Saidman, L.J. pp. 643- 661. Philadelphia: Lippincott-Raven Publishers.

210

Helke C J, Krause J E, Mantyh P W, Couture R, Bannon M J (1990) Diversity in mammalian tachykinin peptidergic neurons: multiple peptides, receptors, and regulatory mechanisms. FASEBJ4: 1606-15.

Hell J W, Westenbroek R E, Warner C, Ahlijanian M K, Prystay W, Gilbert M M, Snutch T P, Catterall W A (1993) Identification and differential subcellular localization of the neuronal class C and class D L-type calcium channel alpha 1 subunits. J Cell Biol 123: 949-62.

Hemstreet B, Lapointe M (2001) Evidence for the use of gabapentin in the treatment of diabetic peripheral neuropathy. Clin Ther 23: 520-31.

Herrington J, Lingle C J (1992) Kinetic and pharmacological properties o f low voltage-activated Ca2+ current in rat clonal (GH3) pituitary cells. J Neurophysiol 68: 213-32.

Hill D R, Shaw T M, Graham W, Woodruff G N (1990) Autoradiographical detection of cholecystokinin-A receptors in primate brain using 1251-Bolton Hunter CCK-8 and 3H-MK-329. J Neurosci 10: 1070-81.

Hobom M, Dai S, Marais E, Lacinova L, Hofmann F, Klugbauer N (2000) Neuronal distribution and functional characterization of the calcium channel OC26-2 subunit.Eur J Neurosci \2\ 1217-1226.

Hoffmeister F, Tettenborn D (1986) Calcium agonists and antagonists o f the dihydropyridine type: antinociceptive effects, interference with opiate-mu-receptor agonists and neuropharmacological actions in rodents. Psychopharmacology (Berl) 90: 299-307.

Hofmann F, Biel M, Flockerzi V (1994) Molecular basis for Ca2+ channel diversity. Annu Rev Neurosci 17: 399-418.

Hofmann F, Lacinova L, Klugbauer N (1999) Voltage-dependent calcium channels: from structure to function. Rev Physiol Biochem Pharmacol 139: 33-87.

Hohmann A G, Briley E M, Herkenham M (1999) Pre- and postsynaptic distribution of cannabinoid and mu opioid receptors in rat spinal cord. Brain Res 822: 17-25.

Hohmann A G, Herkenham M (1999) Localization o f central cannabinoid CBl receptor messenger RNA in neuronal subpopulations o f rat dorsal root ganglia: a double-label in situ hybridization study. Neuroscience 90: 923-31.

Holz G G T, Dunlap K, Kream R M (1988) Characterization o f the electrically evoked release of substance P from dorsal root ganglion neurons: methods and dihydropyridine sensitivity. J Neurosci 8: 463-71.

Holzer P (1988) Local effector functions o f capsaicin-sensitive sensory nerve endings: involvement o f tachykinins, calcitonin gene-related peptide and other neuropeptides. Neuroscience 24: 739-68.

211

Honda C N, Rethelyi M, Petrusz P (1983) Preferential immunohistochemical localization o f vasoactive intestinal polypeptide (VIP) in the sacral spinal cord o f the cat: light and electron microscopic observations. J Neurosci 3: 2183-96.

Houser S J, Eads M, Embrey J P, Welch S P (2000) Dynorphin B and spinal analgesia: induction of antinociception by the cannabinoids CP55,940, Delta(9)-THC and anandamide. Brain Res 857: 337-42.

Hua X Y, Boublik J H, Spicer M A, Rivier J E, Brown M R, Yaksh T L (1991) The antinociceptive effects of spinally administered neuropeptide Y in the rat: systematic studies on structure-activity relationship. J Pharmacol Exp Ther 258: 243-8.

Hua X Y, Chen P, Yaksh T L (1999) Inhibition o f spinal protein kinase C reduces nerve injury-induced tactile allodynia in neuropathic rats. Neurosci Lett 276: 99-102.

Huguenard J R (1996) Low-threshold calcium currents in central nervous system neurons. Annu Rev Physiol 58: 329-48.

Hunt S P, Kelly J S, Emson P C, Kimmel J R, Miller R J, Wu J Y (1981) An immunohistochemical study of neuronal populations containing neuropeptides or gamma-aminobutyrate within the superficial layers o f the rat dorsal horn. Neuroscience 6: 1883-98.

Hunter J C, Gogas K R, Hedley L R, Jacobson L O, Kassotakis L, Thompson J, Fontana D J (1997) The effect o f novel anti-epileptic drugs in rat experimental models of acute and chronic pain. Eur J Pharmacol 324: 153-60.

Hwang J H, Hwang K S, Choi Y, Park P H, Han S M, Lee D M (2000) An analysis of drug interaction between morphine and neostigmine in rats with nerve-ligation injury. Anesth Analg 90: 421-6.

Hwang J H, Yaksh T L (1997) Effect o f subarachnoid gabapentin on tactile-evoked allodynia in a surgically induced neuropathic pain model in the rat. Reg Anesth 22: 249-56.

Ibuki T, Hama A T, Wang X T, Pappas G D, Sagen J (1997) Loss o f GABA- immunoreactivity in the spinal dorsal horn of rats with peripheral nerve injury and promotion o f recovery by adrenal medullary grafts. Neuroscience 76: 845-58.

Igelmund P, Zhao Y Q, Heinemann U (1996) Effects of T-type, L-type, N-type, P- type, and Q-type calcium channel blockers on stimulus-induced pre- and postsynaptic calcium fluxes in rat hippocampal slices. Exp Brain Res 109: 22-32.

Ishibashi H, Rhee J S, Akaike N (1995) Regional difference o f high voltage- activated Ca2+ channels in rat CNS neurones. Neuroreport 6: 1621-4.

Isom L L, De Jongh K S, Catterall W A (1994) Auxiliary subunits of voltage-gated ion charmels. Neuron 12: 1183-94.

212

Jadad A R, Carroll D, Glynn C J, Moore R A, McQuay H J (1992) Morphine responsiveness of chronic pain: double-blind randomised crossover study with patient-controlled analgesia. Lancet 339: 1367-71.

Jahnsen H, Llinas R (1984) Electrophysiological properties o f guinea-pig thalamic neurones: an in vitro study. J Physiol 349: 205-26.

Jasmin L, Kohan L, Franssen M, Janni G, Goff J R (1998) The cold plate as a test o f nociceptive behaviors: description and application to the study o f chronic neuropathic and inflammatory pain models. Pain 75: 367-82.

Jay S D, Sharp A H, Kahl S D, Vedvick T S, Harpold M M, Campbell K P (1991) Structural characterization o f the dihydropyridine-sensitive calcium channel alpha 2- subunit and the associated delta peptides. JB iol Chem 266: 3287-93.

Jazat F, Guilbaud G (1991) The 'tonic' pain-related behaviour seen in mononeuropathic rats is modulated by morphine and naloxone. Pain 44: 97-102.

Jensen T S, Rasmussen P (1994). Phantom pain and related phenomena in amputation. In Textbook o f Pain. ed. Wall, P.D. & Melzack, R. pp. 651-665. Edinburgh: Churchill Livingstone.

Jones R S, Heinemann U H (1987) Differential effects o f calcium entry blockers on pre- and postsynaptic influx o f calcium in the rat hippocampus in vitro. Brain Res 416:257-66.

Jun J H, Yaksh T L (1998) The effect o f intrathecal gabapentin and 3-isobutyl gamma-aminobutyric acid on the hyperalgesia observed after thermal injury in the rat. Anesth Analg 86: 348-54.

Kajander K C, Bennett G J (1992) Onset o f a painful peripheral neuropathy in rat: a partial and differential deafferentation and spontaneous discharge in A beta and A delta primary afferent neurons. J Neurophysiol 68: 734-44.

Kajander K C, Wakisaka S, Bennett G J (1992) Spontaneous discharge originates in the dorsal root ganglion at the onset o f a painful peripheral neuropathy in the rat. Neurosci Lett 138: 225-8.

Kamiya H, Sawada S, Yamamoto C (1988) Synthetic omega-conotoxin blocks synaptic transmission in the hippocampus in vitro. Neurosci Lett 91: 84-8.

Kaneda M, Wakamori M, Ito C, Akaike N (1990) Low-threshold calcium current in isolated Purkinje cell bodies o f rat cerebellum. J Neurophysiol 63: 1046-51.

Karst H, Joels M, Wadman W J (1993) Low-threshold calcium current in dendrites o f the adult rat hippocampus. Neurosci Lett 164: 154-8.

Kasai H (1991) Tonic inhibition and rebound facilitation o f a neuronal calcium channel by a GTP-binding protein. Proc Natl Acad Sci U SA 88: 8855-9.

213

Kasai H (1992) Voltage- and time-dependent inhibition of neuronal calcium channels by a GTP-binding protein in a mammalian cell line. J Physiol 448: 189-209.

Kauppila T, Kontinen V K, Pertovaara A (1998a) Weight bearing o f the limb as a confounding factor in assessment of mechanical allodynia in the rat. Pain 74: 55-9.

Kauppila T, Xu X J, Yu W, Wiesenfeld-Hallin Z (1998b) Dextromethorphan potentiates the effect of morphine in rats with peripheral neuropathy. Neuroreport 9: 1071-4.

Kavalali E T, Zhuo M, Bito H, Tsien R W (1997) Dendritic Ca2+ channels characterized by recordings from isolated hippocampal dendritic segments. Neuron 18: 651-63.

Kawamata M, Omote K (1996) Involvement of increased excitatory amino acids and intracellular Ca2+ concentration in the spinal dorsal horn in an animal model o f neuropathic pain. Pain 68: 85-96.

Kayser V, Christensen D (2000) Antinociceptive effect of systemic gabapentin in mononeuropathic rats, depends on stimulus characteristics and level o f test integration. Pain 88: 53-60.

Kayser V, Lee S H, Guilbaud G (1995) Evidence for a peripheral component in the enhanced antinociceptive effect of a low dose of systemic morphine in rats with peripheral mononeuropathy. Neuroscience 64: 537-45.

Kerr L M, Filloux F, Olivera B M, Jackson H, Wamsley J K (1988) Autoradiographic localization of calcium channels with [125I]omega-conotoxin in rat brain. Eur J Pharmacol 146: 181-3.

Kieffer B L (1997). Molecular aspect of opioid receptors. In The pharmacology o f pain: Handbook o f experimental pharmacology, ed. Dickenson, A.H. & Besson, J.M. pp. 281-304. Berlin: Springer-Verlag.

Kim C, Jun K, Lee T, Kim S S, McEnery M W, Chin H, Kim H L, Park J M, Kim D K, Jung S J, Kim J, Shin H S (2001) Altered Nociceptive Response in Mice Deficient in the alpha(lB) Subunit of the Voltage-Dependent Calcium Channel. Mol Cell Neurosci 18: 235-45.

Kim H J, Na H S, Sung B, Hong S K (1998) Amount of sympathetic sprouting in the dorsal root ganglia is not correlated to the level o f sympathetic dependence of neuropathic pain in a rat model. Neurosci Lett 245: 21-4.

Kim K J, Yoon Y W, Chung J M (1997) Comparison o f three rodent neuropathic pain models. Exp Brain Res 113: 200-6.

Kim S H, Chung J M (1992) An experimental model for peripheral neuropathy produced by segmental spinal nerve ligation in the rat. Pain 50: 355-63.

Kim S H, Na H S, Sheen K, Chung J M (1993) Effects of sympathectomy on a rat model o f peripheral neuropathy. Pain 55: 85-92.

214

Kimura M, Yamanishi Y, Hanada T, Kagaya T, Kuwada M, Watanabe T, Katayama K, Nishizawa Y (1995) Involvement o f P-type calcium channels in high potassium- elicited release of neurotransmitters from rat brain slices. Neuroscience 66: 609-15.

Kingery W S (1997) A critical review o f controlled clinical trials for peripheral neuropathic pain and complex regional pain syndromes. Pain 73: 123-39.

Kingery W S, Vallin J A (1989) The development o f chronic mechanical hyperalgesia, autotomy and collateral sprouting following sciatic nerve section in rat. Pam 38: 321-32.

Klugbauer N, Lacinova L, Marais E, Hobom M, Hofmann F (1999) Molecular diversity of the calcium channel alpha2delta subunit. J Neurosci 19: 684-91.

Koch B D, Faurot G F, McGiurk J R, Clarke D E, Hunter J C (1996) Modulation of mechano-hyperalgesia by clinically effective analgesics in rats with a peripheral mononeuropathy. Analgesia 2: 157-164.

Koerber H R, Seymour A W, Mendell L M (1989) Mismatches between peripheral receptor type and central projections after peripheral nerve regeneration. Neurosci Lett 99: 67-72.

Kohama I, Ishikawa K, Kocsis J D (2000) Synaptic reorganization in the substantia gelatinosa after peripheral nerve neuroma formation: aberrant innervation of lamina II neurons by Abeta afferents. J Neurosci 20: 1538-49.

Kohno T, Kumamoto E, Higashi H, Shimoji K, Yoshimura M (1999) Actions of opioids on excitatory and inhibitory transmission in substantia gelatinosa of adult rat spinal cord. J Physiol 518: 803-13.

Koltzenburg M, Kees S, Budweiser S, Ochs G, Toyka K V (1994a). The properties of unmyelinated nociceptive afferents change in a painful chronic constriction neuropathy. In Proceeding o f the 7th world congress on pain. Progress in pain research and management, ed. Gebhart, G.F., Hammond, D.L. & Jensen, T.S. pp. 511-522. Seattle: lASP Press.

Koltzenburg M, Lundberg L E, Torebjork H E (1992) Dynamic and static components o f mechanical hyperalgesia in human hairy skin. Pain 51: 207-19.

Koltzenburg M, Torebjork H E, Wahren L K (1994b) Nociceptor modulated central sensitization causes mechanical hyperalgesia in acute chemogenic and chronic neuropathic pain. Brain 117: 579-91.

Kontinen V K, Dickenson A H (2000) Effects o f midazolam in the spinal nerve ligation model of neuropathic pain in rats. Pain 85: 425-31.

Kontinen V K, Kauppila T, Paananen S, Pertovaara A, Kalso E (1999) Behavioural measures of depression and anxiety in rats with spinal nerve ligation-induced neuropathy. Pain 80: 341-6.

215

Kontinen V K, Paananen S, Kalso E (1998) Systemic morphine in the prevention of allodynia in the rat spinal nerve ligation model of neuropathic pain. Eur J Pain 2:35- 42.

Kostyuk P G (1999) Low-voltage activated calcium channels: achievements and problems. Neuroscience 92: 1157-63.

Kostyuk P G, Molokanova E A, Pronchuk N F, Savchenko A N, Verkhratsky A N(1992) Different action of ethosuximide on low- and high-threshold calcium currents in rat sensory neurons. Neuroscience 51: 755-8.

Kostyuk P G, Shirokov R E (1989) Deactivation kinetics of different components of calcium inward current in the membrane o f mice sensory neurones. J Physiol 409: 343-55.

Kupers R C, Konings H, Adriaensen H, Gybels J M (1991) Morphine differentially affects the sensory and affective pain ratings in neurogenic and idiopathic forms of pain. Pain 47: 5-12.

Lai J, Hunter J C, Ossipov M H, Porreca F (2000) Blockade of neuropathic pain by antisense targeting of tetrodotoxin-resistant sodium channels in sensory neurons. Methods Enzymol 314: 201-13.

Laird J M, Bennett G J (1993) An electrophysiological study of dorsal horn neurons in the spinal cord of rats with an experimental peripheral neuropathy. J Neurophysiol 69: 2072-85.

Lawson S N, Perry M J, Prabhakar E, McCarthy P W (1993) Primary sensory neurones: neurofilament, neuropeptides, and conduction velocity. Brain Res Bull 30: 239-43.

Le Bars D, Guilbaud G, Jurna I, Besson J M (1976) Differential effects of morphine on responses of dorsal horn lamina V type cells elicited by A and C fibre stimulation in the spinal cat. Brain Res 115:518-24.

Lee J H, Daud A N, Cribbs L L, Lacerda A E, Pereverzev A, Klockner U, Schneider T, Perez-Reyes E (1999) Cloning and expression o f a novel member o f the low voltage-activated T-type calcium channel family. J Neurosci 19: 1912-21.

Lee S H, Kayser V, Desmeules J, Guilbaud G (1994) Differential action of morphine and various opioid agonists on thermal allodynia and hyperalgesia in mononeuropathic rats. Pain 57: 233-40.

Lee Y W, Chaplan S R, Yaksh T L (1995) Systemic and supraspinal, but not spinal, opiates suppress allodynia in a rat neuropathic pain model. Neurosci Lett 199: 111-4.

Lehmann K A, Kriegel R, Ueki M (1989) The clinical significance o f drug interactions between opiates and calcium antagonists. A randomized double-blind study using fentanyl and nimodipine within the framework o f postoperative intravenous on-demand analgesia. Anaesthesist 38: 110-5.

216

Lekan H A, Carlton S M, Coggeshall R E (1996) Sprouting of A beta fibers into lamina II o f the rat dorsal horn in peripheral neuropathy. Neurosci Lett 208: 147-50.

Letts V A, Felix R, Biddlecome G H, Arikkath J, Mahaffey C L, Valenzuela A, Bartlett F S, 2nd, Mori Y, Campbell K P, Frankel W N (1998) The mouse stargazer gene encodes a neuronal Ca2+-channel gamma subunit. Nat Genet 19: 340-7.

Levine L L, Winter P M, Nemoto E M, Uram M, Lin M R (1986) Naloxone does not antagonize the analgesic effects of inhalation anesthetics. Anesth Analg 65: 330-2.

Likar R, Spendel M C, Amberger W, Kepplinger B, Supanz S, Sadjak A (1999) Long-term intraspinal infusions of opioids with a new implantable medication pump. Arzneimittelforschung A9\ 489-93.

Lindblom U (1994). Analysis o f abnormal touch, pain and temperature sensation in patients. In Touch, temperature and pain in health and disease: mechanisms and assessments, Progress in pain research and management, ed. Boivie, J., Hansson, P. & Lindblom, U. pp. 63-84. Seattle: lASP Press.

Liu C-N, Michaelis M, Amir R, Devor M (2000) Spinal Nerve Injury Enhances Subthreshold Membrane Potential Oscillations in DRG Neurons: Relation to Neuropathic Pain. J Neurophysiol 84: 205-215.

Liu H, De Waard M, Scott V E, Gurnett C A, Lennon V A, Campbell K P (1996) Identification o f three subunits o f the high affinity omega-conotoxin MVIIC- sensitive Ca2+ channel. JB iol Chem 271: 13804-10.

Llinas R R, Sugimori M, Cherksey B (1989) Voltage-dependent calcium conductances in mammalian neurons. The P channel. Ann NY Acad Sci 560:103-11.

Lombard M C, Besse D, Chen Y L, Besson J M (1990) mu-opioid receptors in the superficial spinal dorsal horn in an experimental model o f periperal mononeuropathy in the rat. Eur J Pharmacol 181: 2308-2318.

Lombard M C, Besson J M (1989) Attempts to gauge the relative importance of pre- and postsynaptic effects of morphine on the transmission o f noxious messages in the dorsal horn of the rat spinal cord Pain 37: 335-45.

Lucas D, Yaksh T L (1990) Release in vivo o f Met-enkephalin and encrypted forms of Met-enkephalin from brain and spinal cord of the anesthetized cat. Peptides 11 : 1119-25.

Luebke J I, Dunlap K, Turner T J (1993) Multiple calcium channel types control glutamatergic synaptic transmission in the hippocampus. Neuron 11: 895-902.

Luo L, Wiesenfeld-Hallin Z (1995) The effects o f pretreatment with tachykinin antagonists and galanin on the development o f spinal cord hyperexcitability following sciatic nerve section in the rat. Neuropeptides 28: 161-6.

Luo Z D (2000) Rat dorsal root ganglia express distinctive forms o f the alpha2 calcium channel subunit. Neuroreport 11: 3449-52.

217

Luo Z D, Chaplan S R, Higuera E S, Sorkin L S, Stauderman K A, Williams M E, Yaksh r L (2001) Upregulation of dorsal root ganglion azô calcium channel subunit and its correlation with allodynia in spinal nerve-injured rats. J Neurosci 21: 1868- 75.

Luo Z D, Chaplan S R, Scott B P, Cizkova D, Calcutt N A, Yaksh T L (1999) Neuronal nitric oxide synthase mRNA upregulation in rat sensory neurons after spinal nerve ligation; lack o f a role in allodynia development. J Neurosci 19: 9201-8.

Luukko M, Konttinen Y, Kemppinen P, Pertovaara A (1994) Influence o f various experimental parameters on the incidence of thermal and mechanical hyperalgesia induced by a constriction mononeuropathy of the sciatic nerve in lightly anesthetized mts. Exp Neurol 128: 143-54.

Lynch III C (1997). Voltage-gated calcium channels. In Anesthesia: Biologic Foundations, ed. Yaksh, T.L., Lynch III, C., Zapol, W.M., Maze, M., Biebuyck, J.F. & Saidman, L.J. pp. 163-196. Philadelphia: Lippincott-Raven Publishers.

Ma Q P, W oolf C J (1995) Noxious stimuli induce an N-methyl-D-aspartate receptor-dependent hypersensitivity o f the flexion withdrawal reflex to touch: implications for the treatment of mechanical allodynia. Pain 61: 383-90.

Ma W, Bisby M A (1997) Differential expression o f galanin immunoreactivities in the primary sensory neurons following partial and complete sciatic nerve injuries. Neuroscience 79: 1183-95.

Macdonald R L, McLean M J (1986) Anticonvulsant drugs: mechanisms o f action. Adv Neurol 44: 713-36.

MacFarlane B V, Wright A, O'Callaghan J, Benson H A (1997) Chronic neuropathic pain and its control by drugs. Pharmacology & Therapeutics 75: 1-19.

Magee J C, Johnston D (1995) Synaptic activation o f voltage-gated channels in the dendrites o f hippocampal pyramidal neurons. Science 268: 301-4.

Maggi C A, Tramontana M, Cecconi R, Santicioli P (1990) Neurochemical evidence for the involvement o f N-type calcium channels in transmitter secretion from peripheral endings o f sensory nerves in guinea pigs. Neurosci Lett 114: 203-6.

Magoul R, Onteniente B, Geffard M, Galas A (1987) Anatomical distribution and ultrastructural organization o f the GABAergic system in the rat spinal cord. An immunocytochemical study using anti-GABA antibodies. Neuroscience 20: 1001-9.

Maixner W, Dubner R, Bushnell M C, Kenshalo D R, Jr., Oliveras J L (1986) Wide- dynamic-range dorsal horn neurons participate in the encoding process by which monkeys perceive the intensity o f noxious heat stimuli. Brain Res 374: 385-8.

Malcangio M, Bowery N G (1994) Spinal cord SP release and hyperalgesia in monoarthritic rats: involvement o f the GABAB receptor system. Br J Pharmacol 113:1561-6.

218

Malmberg A B, Chen C, Tonegawa S, Basbaum A I (1997) Preserved acute pain and reduced neuropathic pain in mice lacking PKCgamma. Science 278: 279-83.

Malmberg A B, Yaksh T L (1994) Voltage-sensitive calcium channels in spinal nociceptive processing: blockade o f N- and P-type channels inhibits formalin- induced nociception. J Neurosci 14: 4882-90.

Malmberg A B, Yaksh T L (1995) Effect of continuous intrathecal infusion of omega-conopeptides, N-type calcium-channel blockers, on behavior and antinociception in the formalin and hot-plate tests in rats. Pain 60: 83-90.

Malmberg A B, Yaksh T L, Calcutt N A (1993) Anti-nociceptive effects of the GMl ganglioside derivative AGF 44 on the formalin test in normal and streptozotocin- diabetic rats. Neurosci Lett 161: 45-8.

Mannion R J, Doubell T P, Gill H, Woolf C J (1998) Deafferentation is insufficient to induce sprouting of A-fibre central terminals in the rat dorsal horn. J Comp Neurol 393: 135-44.

Mansikka H, Sheth R N, DeVries C, Lee H, Winchurch R, Raja S N (2000) Nerve injury-induced mechanical but not thermal hyperalgesia is attenuated in neurokinin-1 receptor knockout mice. Exp Neurol 162: 343-9.

Mansour A, Fox C A, Akil H, Watson S J (1995) Opioid-receptor mRNA expression in the rat CNS: anatomical and functional implications. Trends Neurosci 18: 22-9.

Mao J, Price D D, Hayes R L, Lu J, Mayer D J, Frenk H (1993) Intrathecal treatment with dextrorphan or ketamine potently reduces pain-related behaviors in a rat model of peripheral mononeuropathy. Brain Res 605: 164-8.

Mao J, Price D D, Mayer D J (1995) Experimental mononeuropathy reduces the antinociceptive effects o f morphine: implications for common intracellular mechanisms involved in morphine tolerance and neuropathic pain. Pain 61: 353-64.

Mao J, Price D D, Mayer D J, Hayes R L (1992) Pain-related increases in spinal cord membrane-bound protein kinase C following peripheral nerve injury. Brain Res 588: 144-9.

Marchand J E, Cepeda M S, Carr D B, Wurm W H, Kream R M (1999) Alterations in neuropeptide Y, tyrosine hydroxylase, and Y-receptor subtype distribution following spinal nerve injury to rats. Pain 79: 187-200.

Marchand J E, Wurm W H, Kato T, Kream R M (1994) Altered tachykinin expression by dorsal root ganglion neurons in a rat model of neuropathic pain. Pain 58:219-31.

Markram H, Sakmarm B (1994) Calcium transients in dendrites o f neocortical neurons evoked by single subthreshold excitatory postsynaptic potentials via low- voltage-activated calcium channels. Proc Natl Acad Sci USA 91: 5207-11.

219

Markus H, Pomeranz B (1987) Saphenous has weak ineffective synapses in sciatic territory of rat spinal cord: electrical stimulation o f the saphenous or application of drugs reveal these somatotopically inappropriate synapses. Brain Res 416: 315-21.

Markus H, Pomeranz B, Krushelnycky D (1984) Spread o f saphenous somatotopic projection map in spinal cord and hypersensitivity o f the foot after chronic sciatic denervation in adult rat. Brain Res 296: 27-39.

Marlier L, Poulat P, Rajaofetra N, Sandillon F, Privât A (1992) Plasticity o f the serotonergic innervation o f the dorsal horn o f the rat spinal cord following neonatal capsaicin treatment. J Neurosci Res 31: 346-58.

Marsala M, Yaksh T L (1994) Transient spinal ischemia in the rat: characterization of behavioral and histopathological consequences as a function of the duration of aortic occlusion. J Cereb Blood Flow Metab 14: 526-35.

Martin W J, Liu H, Wang H, Malmberg A B, Basbaum A I (1999a) Inflammation- induced up-regulation of protein kinase Cgamma immunoreactivity in rat spinal cord correlates with enhanced nociceptive processing. Neuroscience 88: 1267-74.

Martin W J, Loo C M, Basbaum A I (1999b) Spinal cannabinoids are anti-allodynic in rats with persistent inflammation. Pain 82: 199-205.

Martin W J, Malmberg A B, Basbaum A I (2001) PKCgamma contributes to a subset of the NMDA-dependent spinal circuits that underlie injury-induced persistent pain. J Neurosci 21: 5321-7.

Mathur V S, McGuire D, Bowersox S, Miljanich G P, Luther R R (1998) Neuronal N-type Calcium Channels: New Prospect in Pain Therapy. Pharmaceutical News 5: 25-29.

Maxwell D J, Christie W M, Ottersen O P, Storm-Mathisen J (1990) Terminals of group la primary afferent fibres in Clarke's column are enriched with L-glutamate- like immunoreactivity. Brain Res 510: 346-50.

Mayer D J, Price D D, Becker D P (1975) Neurophysiological characterization of the anterolateral spinal cord neurons contributing to pain perception in man. Pain 1:51- 8 .

McCarthy P W, Lawson S N (1989) Cell type and conduction velocity of rat primary sensory neurons with substance P-like immunoreactivity. Neuroscience 28: 745-53.

McCarthy P W, Lawson S N (1990) Cell type and conduction velocity of rat primary sensory neurons with calcitonin gene-related peptide-like immunoreactivity. Neuroscience 34: 623-32.

McDowell T S, Pancrazio J J, Barrett P Q, Lynch C, 3rd (1999) Volatile anesthetic sensitivity of T-type calcium currents in various cell types. Anesth Analg 88: 168-73.

McMahon S B, Lewin G R, Wall P D (1993) Central hyperexcitability triggered by noxious inputs. Curr Opin Neurobiol 3: 602-10.

220

McQuay H, Carroll D, Jadad A R, Wiffen P, Moore A (1995) Anticonvulsant drugs for management of pain: a systematic review. Br Med J 311: 1047-52.

McQuay H J, Jadad A R, Carroll D, Faura C, Glynn C J, Moore R A, Liu Y (1992) Opioid sensitivity o f chronic pain: a patient-controlled analgesia method. Anaesthesia 47: 757-67.

McQuay H J, Tramer M, Nye B A, Carroll D, Wiffen P J, Moore R A (1996) A systematic review o f antidepressants in neuropathic pain. Pain 68: 217-27.

Meder W P, Dooley D J (2000) Modulation of K(+)-induced synaptosomal calcium influx by gabapentin. Brain Res 875: 157-9.

Meller S T, Pechman P S, Gebhart G F, Maves T J (1992) Nitric oxide mediates the thermal hyperalgesia produced in a model o f neuropathic pain in the rat. Neuroscience 50: 7-10.

Merighi A, Kar S, Gibson S J, Ghidella S, Gobetto A, Peirone S M, Polak J M (1990) The immunocytochemical distribution of seven peptides in the spinal cord and dorsal root ganglia of horse and pig. Anat Embryol (Berl) 181: 271-80.

Merighi A, Polak J M, Gibson S J, Gulbenkian S, Valentino K L, Peirone S M (1988) Ultrastructural studies on calcitonin gene-related peptide-, tachykinins- and somatostatin-immunoreactive neurones in rat dorsal root ganglia: evidence for the colocalization o f different peptides in single secretory granules. Cell Tissue Res 254: 101-9.

Merighi A, Polak J M, Theodosis D T (1991) Ultrastructural visualization of glutamate and aspartate immunoreactivities in the rat dorsal horn, with special reference to the co-localization o f glutamate, substance P and calcitonin-gene related peptide. Neuroscience 40: 67-80.

Mertens P, Ghaemmaghami C, Bert L, Perret-Liaudet A, Sindou M, Renaud B (2000) Amino acids in spinal dorsal horn of patients during surgery for neuropathic pain or spasticity. Neuroreport 11: 1795-8.

Michaelis M, Vogel C, Blenk K H, Janig W (1997) Algesics excite axotomised afferent nerve fibres within the first hours following nerve transection in rats. Pain 72: 347-54.

Mikami A, Imoto K, Tanabe T, Niidome T, Mori Y, Takeshima H, Narumiya S, Numa S (1989) Primary structure and functional expression o f the cardiac dihydropyridine-sensitive calcium channel. Nature 340: 230-3.

Miki K, Fukuoka T, Tokunaga A, Noguchi K (1998) Calcitonin gene-related peptide increase in the rat spinal dorsal horn and dorsal column nucleus following peripheral nerve injury: up-regulation in a subpopulation o f primary afferent sensory neurons. Neuroscience 82: 1243-52.

221

Milani D, Malgaroli A, Guidolin D, Fasolato C, Skaper S D, Meldolesi J, Pozzan T (1990) Ca2+ channels and intracellular Ca2+ stores in neuronal and neuroendocrine cells. Cell Calcium 11: 191-9.

Miletic V, Tan H (1988) lontophoretic application of calcitonin gene-related peptide produces a slow and prolonged excitation of neurons in the cat lumbar dorsal horn. Brain Res 446: 169-72.

Miljanich G P, Ramachandran J (1995) Antagonists o f neuronal calcium channels: structure, function, and therapeutic implications. Annual Review o f Pharmacology & Toxicology 35: 707-34.

Millan M J (1997). The role of descending noradrenergic and serotonergic pathways in the modulation o f nociception: focus on receptor multiplicity. In The Pharmacology o f Pain. Handbook o f Experimental Pharmacology, ed. Dickenson, A. & Besson, J.-M. pp. 385-446. Berlin: Springer-Verlag.

Millhom D E, Hokfelt T, Seroogy K, Oertel W, Verhofstad A A, Wu J Y (1987) Immunohistochemical evidence for colocalization o f gamma-aminobutyric acid and serotonin in neurons of the ventral medulla oblongata projecting to the spinal cord. Brain Res 4\0: 179-85.

Mintz I M, Adams M E, Bean B P (1992) P-type calcium channels in rat central and peripheral neurons. Neuron 9: 85-95.

Mintz I M, Bean B P (1993) GABAB receptor inhibition of P-type Ca2+ channels in central neurons. Neuron 10: 889-98.

Miranda H F, Bustamante D, Kramer V, Pelissier T, Saavedra H, Paeile C, Fernandez E, Pinardi G (1992) Antinociceptive effects of Ca2+ channel blockers. Eur J Pharmacol 2 \1 : 137-41.

Molander C, Grant G (1986) Laminar distribution and somatotopic organization of primary afferent fibers from hindlimb nerves in the dorsal horn. A study by transganglionic transport o f horseradish peroxidase in the rat. Neuroscience 19: 297- 312.

Molliver D C, Radeke M J, Feinstein S C, Snider W D (1995) Presence or absence of Trk A protein distinguishes subsets o f small sensory neurons with unique cytochemical characteristics and dorsal horn projections. J Comp Neurol 361: 404- 16.

Molliver D C, Wright D E, Leitner M L, Parsadanian A S, Doster K, Wen D, Yan Q, Snider W D (1997) IB4-binding DRG neurons switch from NGF to GDNF dependence in early postnatal life. Neuron 19: 849-61.

Morisset V, Nagy F (1996) Modulation of regenerative membrane properties by stimulation of metabotropic glutamate receptors in rat deep dorsal horn neurons. J Neurophysiol 76: 2794-8.

2 2 2

Morisset V, Nagy F (1998) Nociceptive integration in the rat spinal cord: role of non-linear membrane properties o f deep dorsal horn neurons. Eur J Neurosci 10: 3642-52.

Morisset V, Nagy F (2000) Plateau potential-dependent windup o f the response to primary afferent stimuli in rat dorsal horn neurons. Eur J Neurosci 12: 3087-95.

Morton C R, Hutchison W D (1989) Release o f sensory neuropeptides in the spinal cord: studies with calcitonin gene-related peptide and galanin. Neuroscience 31: 807- 15.

Mouginot D, Bossu J L, Gahwiler B H (1997) Low-threshold Ca2+ currents in dendritic recordings from Purkinje cells in rat cerebellar slice cultures. J Neurosci 17: 160-70.

Moulin D E, lezzi A, Amireh R, Sharpe W K, Boyd D, Merskey H (1996) Randomised trial of oral morphine for chronic non-cancer pain. Lancet 347: 143-7.

Mulderry P K (1994) Neuropeptide expression by newborn and adult rat sensory neurons in culture: effects o f nerve growth factor and other neurotrophic factors. Neuroscience 59: 673-88.

Munglani R, Hudspith M J, Hunt S P (1996) The therapeutic potential of neuropeptide Y. Analgesic, anxiolytic and antihypertensive. Drugs 52: 371-89.

Murakami M, Suzuki T, Nakagawasai O, Murakami H, Murakami S, Esashi A, Taniguchi R, Yanagisawa T, Tan-No K, Miyoshi I, Sasano H, Tadano T (2001) Distribution of various calcium channel alpha(l) subunits in murine DRG neurons and antinociceptive effect of omega-conotoxin SVIB in mice. Brain Res 903: 231-6.

Murphy T H, Worley P F, Baraban J M (1991) L-type voltage-sensitive calcium channels mediate synaptic activation of immediate early genes. Neuron 7: 625-35.

Na H S, Kim H J, Sung B, Back S K, Kim D Y, Kim J S, Hong S K (2001) Decrease in spinal CGRP and substance P is not related to neuropathic pain in a rat model. Neuroreport 12: 175-8.

Nakamura S, Myers R R (1999) Myelinated afferents sprout into lamina II of L3-5 dorsal horn following chronic constriction nerve injury in rats. Brain Res 818: 285- 90.

Navarro X, Verdu E, Buti M (1994) Comparison of regenerative and reinnervating capabilities of different functional types o f nerve fibers. Exp Neurol 129: 217-24.

Navarro X, Verdu E, Wendelschafer-Crabb G, Kennedy W R (1997) Immunohistochemical study o f skin reinnervation by regenerative axons. J Comp Neurol 380: 164-74.

Nebe J, Vanegas H, Neugebauer V, Schaible H G (1997) Omega-agatoxin IVA, a P- type calcium channel antagonist, reduces nociceptive processing in spinal cord

223

neurons with input from the inflamed but not from the normal knee joint—an electrophysiological study in the rat in vivo. Eur J Neurosci 9: 2193-201.

Nebe J, Vanegas H, Schaible H G (1998) Spinal application of omega-conotoxin G VIA, an N-type calcium channel antagonist, attenuates enhancement o f dorsal spinal neuronal responses caused by intra-articular injection of mustard oil in the rat. Exp Brain Res 120: 61-9.

Neugebauer V, Vanegas H, Nebe J, Rumenapp P, Schaible H G (1996) Effects o fN - and L-type calcium channel antagonists on the responses of nociceptive spinal cord neurons to mechanical stimulation o f the normal and the inflamed knee joint. J Neurophysiol 76: 3740-9.

Nicholas A P, Pieribone V, Hokfelt T (1993) Distributions of mRNAs for alpha-2 adrenergic receptor subtypes in rat brain: an in situ hybridization study. J Comp Neurol 328: 575-94.

Nichols M L, Bian D, Ossipov M H, Lai J, Porreca F (1995) Regulation of morphine antiallodynic efficacy by cholecystokinin in a model o f neuropathic pain in rats. J Pharmacol Exp Ther215\ 1339-45.

Nichols M L, Lopez Y, Ossipov M H, Bian D, Porreca F (1997) Enhancement o f the antiallodynic and antinociceptive efficacy o f spinal morphine by antisera to dynorphin A (1-13) or MK-801 in a nerve-ligation model of peripheral neuropathy. Pam 69: 317-22.

Nicholson B (2000) Gabapentin use in neuropathic pain syndromes. Acta Neurol S ca n d \Q \\359-71.

Nielsch U, Bisby M A, Keen P (1987) Effect of cutting or crushing the rat sciatic nerve on synthesis o f substance P by isolated L5 dorsal root ganglia. Neuropeptides 10: 137-45.

Nielsch U, Keen P (1989) Reciprocal regulation o f tachykinin- and vasoactive intestinal peptide-gene expression in rat sensory neurones following cut and crush injury. Brain Res 481: 25-30.

Noguchi K, De Leon M, Nahin R L, Senba E, Ruda M A (1993) Quantification of axotomy-induced alteration of neuropeptide mRNAs in dorsal root ganglion neurons with special reference to neuropeptide Y mRNA and the effects of neonatal capsaicin treatment. J Neurosci Res 35: 54-66.

Noguchi K, Dubner R, De Leon M, Senba E, Ruda M A (1994) Axotomy induces preprotachykinin gene expression in a subpopulation of dorsal root ganglion neurons. J Neurosci Res 37: 596-603.

Noguchi K, Kawai Y, Fukuoka T, Senba E, Miki K (1995) Substance P induced by peripheral nerve injury in primary afferent sensory neurons and its effect on dorsal column nucleus neurons. J Neurosci 15: 7633-43.

224

Nothias F, Tessler A, Murray M (1993) Restoration of substance P and calcitonin gene-related peptide in dorsal root ganglia and dorsal horn after neonatal sciatic nerve lesion. J Neurol 334: 370-84.

Novakovic S D, Tzoumaka E, McGivern J G, Haraguchi M, Sangameswaran L, Gogas K R, Eglen R M, Hunter J C (1998) Distribution of the tetrodotoxin-resistant sodium channel PN3 in rat sensory neurons in normal and neuropathic conditions. J Neurosci 18: 2174-87.

Nowycky M C, Fox A P, Tsien R W (1985) Three types of neuronal calcium channel with different calcium agonist sensitivity. Nature 316: 440-3.

Nystrom B, Hagbarth K E (1981) Microelectrode recordings from transected nerves in amputees with phantom limb pain. Neurosci Lett 27: 211-6.

Ochoa J L, Yarnitsky D (1993) Mechanical hyperalgesias in neuropathic pain patients: dynamic and static subtypes. Ann Neurol 33: 465-72.

Ohsawa M, Narita M, Mizoguchi H, Suzuki T, Tseng L F (2000) Involvement of spinal protein kinase C in thermal hyperalgesia evoked by partial sciatic nerve ligation, but not by inflammation in the mouse. Eur J Pharmacol 403: 81-5.

Oku R, Nanayama T, Satoh M (1988) Calcitonin gene-related peptide modulates calcium mobilization in synaptosomes of rat spinal dorsal horn. Brain Res 475: 356- 60.

Olivera B M, Miljanich G P, Ramachandran J, Adams M E (1994) Calcium channel diversity and neurotransmitter release: the omega-conotoxins and omega-agatoxins. Annu Rev Biochem 63: 823-67.

Omote K, Iwasaki H, Kawamata M, Satoh O, Namiki A (1995) Effects of verapamil on spinal anesthesia with local anesthetics. Anesth Analg 80: 444-8.

Ossipov M H, Lopez Y, Bian D, Nichols M L, Porreca F (1997) Synergistic antinociceptive interactions o f morphine and clonidine in rats with nerve-ligation injury. Anesthesiology S6: 196-204.

Ossipov M H, Lopez Y, Nichols M L, Bian D, Porreca F (1995a) Inhibition by spinal morphine o f the tail-flick response is attenuated in rats with nerve ligation injury. Neurosci Lett 199: 83-6.

Ossipov M H, Lopez Y, Nichols M L, Bian D, Porreca F (1995b) The loss of antinociceptive efficacy o f spinal morphine in rats with nerve ligation injury is prevented by reducing spinal afferent drive. Neurosci Lett 199: 87-90.

Ozawa S, Kamiya H, Tsuzuki K (1998) Glutamate receptors in the mammalian central nervous system. ProgNeurobiol 54: 581-618.

Palecek J, Dougherty P M, Kim S H, Paleckova V, Lekan H, Chung J M, Carlton S M, Willis W D (1992a) Responses of spinothalamic tract neurons to mechanical and

225

thermal stimuli in an experimental model of peripheral neuropathy in primates. J Neurophysiol 6^’. 1951-66.

Palecek J, Paleckova V, Dougherty P M, Carlton S M, Willis W D (1992b) Responses o f spinothalamic tract cells to mechanical and thermal stimulation of skin in rats vdth experimental peripheral neuropathy. J Neurophysiol 67: 1562-73.

Pan H L, Eisenach J C, Chen S R (1999) Gabapentin suppresses ectopic nerve discharges and reverses allodynia in neuropathic rats. J Pharmacol Exp Ther 288: 1026-30.

Pang I H, Vasko M R (1986) Morphine and norepinephrine but not 5- hydroxytryptamine and gamma-aminobutyric acid inhibit the potassium-stimulated release o f substance P from rat spinal cord slices. Brain Res 376: 268-79.

Pappagallo M, Oaklander A L, Quatrano-Piacentini A L, Clark M R, Raja S N (2000) Heterogenous patterns of sensory dysfunction in postherpetic neuralgia suggest multiple pathophysiologic mechanisms. Anesthesiology 92: 691-8.

Patel M K, Gonzalez M I, Bramwell S, Pinnock R D, Lee K (2000) Gabapentin inhibits excitatory synaptic transmission in the hyperalgesic spinal cord. Br J Pharmacol 130: 1731-4.

Patel S, Naeem S, Kesingland A, Froestl W, Capogna M, Urban L, Fox A (2001) The effects o f GABA(B) agonists and gabapentin on mechanical hyperalgesia in models o f neuropathic and inflammatory pain in the rat. Pain 90: 217-26.

Patsalos P N, Duncan J S (1993) Antiepileptic drugs. A review o f clinically significant drug interactions. DrugSaf9: 156-84.

Peluso J, LaForge K S, Matthes H W, Kreek M J, Kieffer B L, Gaveriaux-Ruff C (1998) Distribution of nociceptin/orphanin FQ receptor transcript in human central nervous system and immune cells. JNeuroimmunol 81: 184-92.

Penn R D, Paice J A (2000) Adverse effects associated with the intrathecal administration o f zinconotide. Pain 85: 291-296.

Perez-Reyes E, Cribbs L L, Daud A, Lacerda A E, Barclay J, Williamson M P, Fox M, Rees M, Lee J H (1998) Molecular characterization of a neuronal low-voltage- activated T-type calcium channel. Nature 391: 896-900.

Pemey T M, Himing L D, Leeman S E, Miller R J (1986) Multiple calcium chaimels mediate neurotransmitter release from peripheral neurons. Proc Natl Acad Sci U S A 83:6656-9.

Perrot S, Attal N, Ardid D, Guilbaud G (1993) Are mechanical and cold allodynia in mononeuropathic and arthritic rats relieved by systemic treatment with calcitonin or guanethidine? Pain 52: 41-7.

Pertovaara A, Kontinen V K, Kalso E A (1997) Chronic spinal nerve ligation induces changes in response characteristics of nociceptive spinal dorsal horn neurons and in

226

their descending regulation originating in the periaqueductal gray in the rat. Exp Neurol 147: 428-36.

Peterson D O, Drummond J C, Todd M M (1986) Effects o f halothane, enflurane, isoflurane, and nitrous oxide on somatosensory evoked potentials in humans. Anesthesiology 65: 35-40.

Pfrieger F W, Veselovsky N S, Gottmarm K, Lux H D (1992) Pharmacological characterization o f calcium currents and synaptic transmission between thalamic neurons in vitro. J Neurosci 12: 4347-57.

Pirec V, Laurito C E, Lu Y, Yeomans D C (2001) The combined effects of N-type calcium channel blockers and morphine on A delta versus C fiber mediated nociception. Anesth Analg 92: 239-43.

Pocock J M, Nicholls D G (1992) A toxin (Aga-GI) from the venom of the spider Agelenopsis aperta inhibits the mammalian presynaptic Ca2+ channel coupled to glutamate exocytosis. Eur J Pharmacol 226: 343-50.

Porreca F, Lai J, Bian D, Wegert S, Ossipov M H, Eglen R M, Kassotakis L, Novakovic S, Rabert D K, Sangameswaran L, Hunter J C (1999) A comparison of the potential role o f the tetrodotoxin-insensitive sodium channels, PN3/SNS and NaN/SNS2, in rat models of chronic pain. Proc Natl Acad Sci U S A 96: 7640-4.

Porreca F, Tang Q B, Bian D, Riedl M, Elde R, Lai J (1998) Spinal opioid mu receptor expression in lumbar spinal cord o f rats following nerve injury. Brain Res 795: 197-203.

Porro C A, Cavazzuti M (1993) Spatial and temporal aspects o f spinal cord and brainstem activation in the formalin pain model. Prog Neurobiol 41: 565-607.

Portenoy R K (1994) Tolerance to opioid analgesics: clinical aspects. Cancer Surv 21:49-65.

Portenoy R K, Foley K M, Inturrisi C E (1990) The nature o f opioid responsiveness and its implications for neuropathic pain: new hypotheses derived from studies of opioid infusions. Pain 43: 273-86.

Price D D, Mao J, Frenk H, Mayer D J (1994) The N-methyl-D-aspartate receptor antagonist dextromethorphan selectively reduces temporal summation of second pain in man. Pain 59: 165-74.

Price D D, Mayer D J (1975) Neurophysiological characterization o f the anterolateral quadrant neurons subserving pain in M. mulatta. Pain 1: 59-72.

Radhakrishnan V, Yashpal K, Hui-Chan C W, Henry J L (1995) Implication of a nitric oxide synthase mechanism in the action o f substance P: L-NAME blocks thermal hyperalgesia induced by endogenous and exogenous substance P in the rat. Eur J Neuroscil: 1920-5.

227

Rahman W, Dashwood M R, Fitzgerald M, Aynsley-Green A, Dickenson A H (1998) Postnatal development o f multiple opioid receptors in the spinal cord and development of spinal morphine analgesia. Brain Res Dev Brain Res 108: 239-54.

Ralston D (1998) Present models of neuropathic pain. Pain Reviews 5: 83-100.

Ramer M S, Bisby M A (1997) Rapid sprouting of sympathetic axons in dorsal root ganglia of rats with a chronic constriction injury. Pain 70: 237-44.

Ramer M S, French G D, Bisby M A (1997) Wallerian degeneration is required for both neuropathic pain and sympathetic sprouting into the DRG. Pain 72: 71-8.

Ramkumar V, el-Fakahany E E (1988) Prolonged morphine treatment increases rat brain dihydropyridine binding sites: possible involvement in development o f morphine dependence. Eur J Pharmacol 146: 73-83.

Randall A, Tsien R W (1995) Pharmacological dissection of multiple types of Ca channel currents in rat cerebellar granule neurones. J Neurosci 15:2995-3012.

Rane S G, Holz G G t, Dunlap K (1987) Dihydropyridine inhibition of neuronal calcium current and substance P release. Pflugers Arch 409: 361-6.

Reddy S V, Yaksh T L (1980) Antinociceptive effects of lanthanum neodymium and europium following intrathecal administration. Neuropharmacology 19: 181-5.

Regan L J, Sah D W, Bean B P (1991) Ca2+ channels in rat central and peripheral neurons: high-threshold current resistant to dihydropyridine blockers and omega- conotoxin. Neuron 6: 269-80.

Rizzo M A, Kocsis J D, Waxman S G (1994) Slow sodium conductances of dorsal root ganglion neurons: intraneuronal homogeneity and intemeuronal heterogeneity. J Neurophysiol 72: 2796-815.

Ro L S, Jacobs J M (1993) The role of the saphenous nerve in experimental sciatic nerve mononeuropathy produced by loose ligatures: a behavioural study. Pain 52: 359-69.

Roca G, Aguilar J L, Gomar C, Mazo V, Costa J, Vidal F (1996) Nimodipine fails to enhance the analgesic effect of slow release morphine in the early phases of cancer pain treatment. Pain 68: 239-43.

Rock D M, Kelly K M, Macdonald R L (1993) Gabapentin actions on ligand- and voltage-gated responses in cultured rodent neurons. Epilepsy Res 16: 89-98.

Rowan S, Todd A J, Spike R C (1993) Evidence that neuropeptide Y is present in GABAergic neurons in the superficial dorsal horn o f the rat spinal cord. Neuroscience 53: 537-45.

Rowbotham M, Harden N, Stacey B, Bernstein P, Magnus-Miller L (1998) Gabapentin for the treatment of postherpetic neuralgia: a randomized controlled trial. Jama 2^0: 1837-42.

228

Rowbotham M C, Reisner-Keller L A, Fields H L (1991) Both intravenous lidocaine and morphine reduce the pain of postherpetic neuralgia. Neurology 41: 1024-8.

Roytta M, Wei H, Pertovaara A (1999) Spinal nerve ligation-induced neuropathy in the rat: sensory disorders and correlation between histology of the peripheral nerves. Pain%Q\ 161-70.

Ryu P D, Randic M (1990) Low- and high-voltage-activated calcium currents in rat spinal dorsal horn neurons. J Neurophysiol 63: 273-85.

Saegusa H, Kurihara T, Zong S, Kazuno A, Matsuda Y, Nonaka T, Han W, Toriyama H, Tanabe T (2001) Suppression of inflammatory and neuropathic pain symptoms in mice lacking the N-type Ca2+ channel. EMBO J 20: 2349-56.

Saegusa H, Kurihara T, Zong S, Minowa O, Kazuno A, Han W, Matsuda Y, Yamanaka H, Osanai M, Noda T, Tanabe T (2000) Altered pain responses in mice lacking alpha IE subunit of the voltage-dependent Ca2+ channel. Proc Natl Acad Sci U S A 97: 6132-7.

Sang C N (2000) NMDA-receptor antagonists in neuropathic pain: experimental methods to clinical trials. J Pain Symptom Manage 19: S21-5.

Santicioli P, Del Bianco E, Tramontana M, Geppetti P, Maggi C A (1992) Release of calcitonin gene-related peptide like-immunoreactivity induced by electrical field stimulation from rat spinal afferents is mediated by conotoxin-sensitive calcium channels. Neurosci Lett 136: 161-4.

Santillan R, Hurle M A, Armijo J A, de los Mozos R, Florez J (1998) Nimodipine- enhanced opiate analgesia in cancer patients requiring morphine dose escalation: a double-blind, placebo-controlled study. Pain 76: 17-26.

Satoh O, Omote K (1996) Roles o f monoaminergic, glycinergic and GABAergic inhibitory systems in the spinal cord in rats with peripheral mononeuropathy. Brain Res 728: 27-36.

Satzinger G (1994) A ntiepileptics from gamma-aminobutyric acid. Arzneimittelforschung AA: 261-6.

Sawynok J (1998) Adenosine receptor activation and nociception. Eur J Pharmacol 347: 1-11.

Scadding J W (1981) Development of ongoing activity, mechanosensitivity, and adrenaline sensitivity in severed peripheral nerve axons. Exp Neurol 73: 345-64.

Schofield P R (1989) The GABAA receptor: molecular biology reveals a complex picture. Trends Pharmacol Sci 10: 476-8.

Schouenborg J (1984) Functional and topographical properties o f field potentials evoked in rat dorsal horn by cutaneous C-fibre stimulation. J Physiol 356: 169-92.

229

Schroeder J E, Fischbach P S, McCleskey E W (1990) T-type calcium channels: heterogeneous expression in rat sensory neurons and selective modulation by phorbol esters. J Neurosci 10: 947-51.

Schumacher T B, Beck H, Steinhauser C, Schramm J, Eiger C E (1998) Effects of phenytoin, carbamazepine, and gabapentin on calcium channels in hippocampal granule cells from patients with temporal lobe epilepsy. Epilepsia 39: 355-63.

Scroggs R S, Fox A P (1992a) Calcium current variation between acutely isolated adult rat dorsal root ganglion neurons o f different size. J Physiol 445: 639-58.

Scroggs R S, Fox A P (1992b) Multiple Ca2+ currents elicited by action potential waveforms in acutely isolated adult rat dorsal root ganglion neurons. J Neurosci 12: 1789-801.

Seltzer Z, Cohn S, Ginzburg R, Beilin B (1991) Modulation of neuropathic pain behavior in rats by spinal disinhibition and NMD A receptor blockade of injury discharge. Pain 45: 69-75.

Seltzer Z, Dubner R, Shir Y (1990) A novel behavioral model of neuropathic pain disorders produced in rats by partial sciatic nerve injury. Pain 43: 205-18.

Sheen K, Chung J M (1993) Signs o f neuropathic pain depend on signals from injured nerve fibers in a rat model. Brain Res 610: 62-8.

Shehab S A, Atkinson M E (1986) Vasoactive intestinal polypeptide (VIP) increases in the spinal cord after peripheral axotomy of the sciatic nerve originate from primary afferent neurons. Brain Res 372: 37-44.

Sherman S E, Loomis C W (1994) Morphine insensitive allodynia is produced by intrathecal strychnine in the lightly anesthetized rat. Pain 56: 17-29.

Shimoyama M, Shimoyama N, Hori Y (2000) Gabapentin affects glutamatergic excitatory neurotransmission in the rat dorsal horn. Pain 85: 405-14.

Shimoyama M, Shimoyama N, Inturrisi C E, Elliott K J (1997a) Gabapentin enhances the antinociceptive effects of spinal morphine in the rat tail-flick test. Pain 72: 375-82.

Shimoyama N, Shimoyama M, Davis A M, Inturrisi C E, Elliott K J (1997b) Spinal gabapentin is antinociceptive in the rat formalin test. Neurosci Lett 222: 65-7.

Shir Y, Seltzer Z (1990) A-fibers mediate mechanical hyperesthesia and allodynia and C-fibers mediate thermal hyperalgesia in a new model of causalgiform pain disorders in rats. Neurosci Lett 115: 62-7.

Shir Y, Seltzer Z (1991) Effects of sympathectomy in a model of causalgiform pain produced by partial sciatic nerve injury in rats. Pain 45: 309-20.

230

Shistik E, Ivanina T, Puri T, Hosey M, Dascal N (1995) Ca2+ current enhancement by alpha 2/delta and beta subunits in Xenopus oocytes: contribution o f changes in channel gating and alpha 1 protein level. J Physiol 489: 55-62.

Silos-Santiago I, Molliver D C, Ozaki S, Smeyne R J, Fagan A M, Barbacid M, Snider W D (1995) Non-TrkA-expressing small DRG neurons are lost in TrkA deficient mice. J Neurosci 15: 5929-42.

Simmons D R, Spike R C, Todd A J (1995) Galanin is contained in GABAergic neurons in the rat spinal dorsal horn. Neurosci Lett 187: 119-22.

Sindrup S H, Jensen T S (1999) Efficacy o f pharmacological treatments o f neuropathic pain: an update and effect related to mechanism of drug action. Pain 83: 389-400.

Sindrup S H, Jensen T S (2000) Pharmacologic treatment of pain in polyneuropathy. Neurology 55: 9\5-20.

Singer D, Biel M, Lotan I, Flockerzi V, Hofmann F, Dascal N (1991) The roles of the subunits in the function of the calcium channel. Science 253: 1553-7.

Singh L, Field M J, Ferris P, Hunter J C, Oles R J, Williams R G, Woodruff G N (1996) The antiepileptic agent gabapentin (Neurontin) possesses anxiolytic-like and antinociceptive actions that are reversed by D-serine. Psychopharmacology (Berl) 127: 1-9.

Sivilotti L, Woolf C J (1994) The contribution of GABAA and glycine receptors to central sensitization: disinhibition and touch-evoked allodynia in the spinal cord. J Neurophysiol 72: 169-79.

Sluka K A (1997) Blockade of calcium channels can prevent the onset of secondary hyperalgesia and allodynia induced by intradermal injection of capsaicin in rats. Pain 71: 157-64.

Sluka K A (1998) Blockade o f N- and P/Q-type calcium channels reduces the secondary heat hyperalgesia induced by acute inflammation. J Pharmacol Exp Ther 287: 232-7.

Smith S J, Augustine G J (1988) Calcium ions, active zones and synaptic transmitter release. Trends Neurosci 11: 458-64.

Sohn J H, Lee B H, Park S H, Ryu J W, Kim B O, Park Y G (2000) Microinjection of opiates into the periaqueductal gray matter attenuates neuropathic pain symptoms in rats. Neuroreport 11: 1413-6.

Sorkin L S, Carlton S M (1997). Spinal Anatomy and Pharmacology of Afferent Processing. In Anesthesia: Biologic Foundations, ed. Yaksh, T.L., Lynch III, C., Zapol, W.M., Maze, M., Biebuyck, J.F. & Saidman, L.J. pp. 577-609. Philadelphia: Lippincott-Raven Publishers.

231

Sotgiu M L, Biella G, Castagna A, Lacerenza M, Marchettini P (1994a) Different time-courses o f i.v. lidocaine effect on ganglionic and spinal units in neuropathic rats. Neuroreport 5: 873-6.

Sotgiu M L, Biella G, Riva L (1994b) A study of early ongoing activity in dorsal horn units following sciatic nerve constriction. Neuroreport 5: 2609-12.

Sotgiu M L, Lacerenza M, Marchettini P (1992) Effect o f systemic lidocaine on dorsal horn neuron hyperactivity following chronic peripheral nerve injury in rats. Somatosens Mot Res 9: 227-33.

Spedding M, Paoletti R (1992) Classification o f calcium channels and the sites o f action of drugs modifying channel function. Pharmacol Rev 44: 363-76.

Stanfa L C, Misra C, Dickenson A H (1996) Amplification o f spinal nociceptive transmission depends on the generation of nitric oxide in normal and carageenan rats. Brain Res 737: 92-8.

Stanfa L C, Singh L, Williams R G, Dickenson A H (1997) Gabapentin, ineffective in normal rats, markedly reduces C-fibre evoked responses after inflammation. Neuroreport 8: 587-90.

Stea A, Tomlinson W J, Soong T W, Bourinet E, Dubel S J, Vincent S R, Snutch T P (1994) Localization and functional properties o f a rat brain alpha lA calcium channel reflect similarities to neuronal Q- and P-type channels. Proc Natl Acad S c iU S A 9 \\ 10576-80.

Steel J H, Terenghi G, Chung J M, Na H S, Carlton S M, Polak J M (1994) Increased nitric oxide synthase immunoreactivity in rat dorsal root ganglia in a neuropathic pain model. Neurosci Lett 169: 81-4.

Stefani A, Spadoni F, Bemardi G (1998) Gabapentin inhibits calcium currents in isolated rat brain neurons. Neuropharmacology 37: 83-91.

Stefani A, Spadoni F, Giacomini P, Lavaroni F, Bemardi G (2001) The effects of gabapentin on different ligand- and voltage-gated currents in isolated cortical neurons. Epilepsy Res 43: 239-48.

Stevens C W, Kajander K C, Bennett G J, Seybold V S (1991) Bilateral and differential changes in spinal mu, delta and kappa opioid binding in rats with a painful, unilateral neuropathy. Pain 46: 315-26.

Stucky C L, Lewin G R (1999) Isolectin B(4)-positive and -negative nociceptors are functionally distinct. J Neurosci 19: 6497-505.

Study R E, Krai M G (1996) Spontaneous action potential activity in isolated dorsal root ganglion neurons from rats with a painful neuropathy. Pain 65: 235-42.

Suman-Chauhan N, Webdale L, Hill D R, Woodmff G N (1993) Characterisation of [3H]gabapentin binding to a novel site in rat brain: homogenate binding studies. Eur J Pharmacol 244: 293-301.

232

Suszkiw J B, O'Leary M E, Murawsky M M, Wang T (1986) Presynaptic calcium channels in rat cortical synaptosomes: fast-kinetics o f phasic calcium influx, channel inactivation, and relationship to nitrendipine receptors. J Neurosci 6: 1349-57.

Suzuki R, Chapman V, Dickenson A H (1999) The effectiveness o f spinal and systemic morphine on rat dorsal horn neuronal responses in the spinal nerve ligation model of neuropathic pain. Pain 80: 215-28.

Suzuki R, Gale A, Dickenson A H (2000a) Altered effects o f and A \ adenosinereceptor agonist on the evoked responses o f spinal dorsal horn neurones in a rat model o f mononeuropathy. J Pain 1: 99-110.

Suzuki R, Kontinen V K, Matthews E, Williams E, Dickenson A H (2000b) Enlargement of the receptive field size to low intensity mechanical stimulation in the rat spinal nerve ligation model o f neuropathy. Exp Neurol 163: 408-13.

Suzuki R, Matthews E A, Dickenson A H (2001) Comparison of the effects of MK- 801, ketamine and memantine on responses o f spinal dorsal horn neurones in a rat model o f mononeuropathy. Pain 91: 101-9.

Swartz K J (1993) Modulation of Ca2+ channels by protein kinase C in rat central and peripheral neurons: disruption o f G protein-mediated inhibition. Neuron 11: 305- 20 .

Swartz K J, Merritt A, Bean B P, Lovinger D M (1993) Protein kinase C modulates glutamate receptor inhibition o f Ca2+ channels and synaptic transmission. Nature 361: 165-8.

Swerdlow M, Cundill J G (1981) Anticonvulsant drugs used in the treatment of lancinating pain. A comparison. Anaesthesia 36: 1129-32.

Tabo E, Jinks S L, Eisele J H, Jr., Carstens E (1999) Behavioral manifestations of neuropathic pain and mechanical allodynia, and changes in spinal dorsal horn neurons, following L4-L6 dorsal root constriction in rats. Pain 80: 503-20.

Takahashi K, Akaike N (1991) Calcium antagonist effects on low-threshold (T-type) calcium current in rat isolated hippocampal CAl pyramidal neurons. J Pharmacol Exp Ther 256: 169-75.

Takahashi T, Momiyama A (1993) Different types o f calcium channels mediate central synaptic transmission. Nature 366: 156-8.

Takaishi K, Eisele J H, Jr., Carstens E (1996) Behavioral and electrophysiological assessment o f hyperalgesia and changes in dorsal horn responses following partial sciatic nerve ligation in rats. Pain 66: 297-306.

Talley E M, Cribbs L L, Lee J H, Daud A, Perez-Reyes E, Bayliss D A (1999) Differential distribution o f three members o f a gene family encoding low voltage- activated (T-type) calcium channels. J Neurosci 19: 1895-911.

233

Tarasenko A N, Kostyuk P G, Eremin A V, Isaev D S (1997) Two types o f low- voltage-activated Ca2+ channels in neurones of rat laterodorsal thalamic nucleus. J Physiol 499: 77-86.

Taylor B K (2001) Pathophysiologic mechanisms o f neuropathic pain. Curr Pain Headache Rep 5\ 151-61.

Taylor C P (1995). Gabapentin: Mechanisms of action. In Antiepileptic Drugs, ed. Levy, R.H., Mattson, R.H. & Meldrum, B.S. pp. 829-841. New York: Raven Press.

Taylor C P, Gee N S, Su T Z, Kocsis J D, Welty D F, Brown J P, Dooley D J, Boden P, Singh L (1998) A summary o f mechanistic hypotheses o f gabapentin pharmacology. Epilepsy Res 29: 233-49.

Taylor C P, Vartanian M G, Yuen P W, Bigge C, Suman-Chauhan N, Hill D R(1993) Potent and stereo specific anticonvulsant activity of 3-isobutyl G ABA relates to in vitro binding at a novel site labeled by tritiated gabapentin. Epilepsy Res 14: 11- 5.

Terrian D M, Dorman R V, Gannon R L (1990) Characterization o f the presynaptic calcium channels involved in glutamate exocytosis from rat hippocampal mossy fiber synaptosomes. Neurosci Lett 119: 211-4.

Todd A J, McGill M M, Shehab S A (2000) Neurokinin 1 receptor expression by neurons in laminae 1, 111 and IV of the rat spinal dorsal horn that project to the brainstem. Eur J Neurosci 12: 689-700.

Todd A J, Spike R C, Russell G, Johnston H M (1992) Immunohistochemical evidence that Met-enkephalin and GABA coexist in some neurones in rat dorsal horn. Brain Res 584: 149-56.

Todd A J, Sullivan A C (1990) Light microscope study of the coexistence of GABA- like and glycine-like immunoreactivities in the spinal cord o f the rat. J Comp Neurol 296: 496-505.

Todd A J, Watt C, Spike R C, Sieghart W (1996) Colocalization o f GABA, glycine, and their receptors at synapses in the rat spinal cord. J Neurosci 16: 974-82.

Todorovic S M, Jevtovic-Todorovic V, Meyenburg A, Mennerick S, Perez-Reyes B, Romano C, Olney J W, Zorumski C F (2001) Redox modulation o f T-type calcium channels in rat peripheral nociceptors. Neuron 31: 75-85.

Todorovic S M, Lingle C J (1998) Pharmacological properties o f T-type Ca2+ current in adult rat sensory neurons: effects of anticonvulsant and anesthetic agents. J Neurophysiol 79: 240-52.

Todorovic S M, Perez-Reyes E, Lingle C J (2000) Anticonvulsants but not general anesthetics have differential blocking effects on different T-type current variants. Mol Pharmacol 58: 98-108.

234

Tomlinson W J, Stea A, Bourinet E, Chamet P, Nargeot J, Snutch T P (1993) Functional properties o f a neuronal class C L-type calcium channel. Neuropharmacology 32: 111 7-26.

Tong Y G, Wang H F, Ju G, Grant G, Hokfelt T, Zhang X (1999) Increased uptake and transport o f cholera toxin B-subunit in dorsal root ganglion neurons after peripheral axotomy: possible implications for sensory sprouting. J Com;? Neurol 404: 143-58.

Torebjork E, Wahren L, Wallin G, Hallin R, Koltzenburg M (1995) Noradrenaline- evoked pain in neuralgia. Pain 63: 11-20.

Toselli M, Taglietti V (1992) Kinetic and pharmacological properties of high- and low-threshold calcium channels in primary cultures o f rat hippocampal neurons. Pflugers Archiv - European Journal o f Physiology 421: 59-66.

Tracey D J, Walker J S (1995) Pain due to nerve damage: are inflammatory mediators involved? Inflamm Res 44: 407-11.

Trafton J A, Abbadie C, Marek K, Basbaum A I (2000) Postsynaptic signaling via the [mu]-opioid receptor: responses of dorsal horn neurons to exogenous opioids and noxious stimulation. J Neurosci 20: 8578-84.

Tsien R W, Lipscombe D, Madison D V, Bley K R, Fox A P (1988) Multiple types o f neuronal calcium channels and their selective modulation. Trends Neurosci 11 : 431-8.

Turner T J, Adams M E, Dunlap K (1992) Calcium channels coupled to glutamate release identified by omega-Aga-IVA. Science 258: 310-3.

Turner T J, Adams M E, Dunlap K (1993) Multiple Ca2+ channel types coexist to regulate synaptosomal neurotransmitter release. Proc Natl Acad Sci USA 90: 9518- 22 .

Vallbo A B, Olausson H, Wessberg J (1999) Unmyelinated afferents constitute a second system coding tactile stimuli of the human hairy skin. J Neurophysiol 81: 2753-63.

Vanegas H, Schaible H (2000) Effects o f antagonists to high-threshold calcium channels upon spinal mechanisms of pain, hyperalgesia and allodynia. Pain 85: 9-18.

Verge V M, Wiesenfeld-Hallin Z, Hokfelt T (1993) Cholecystokinin in mammalian primary sensory neurons and spinal cord: in situ hybridization studies in rat and monkey. Eur J Neurosci 5: 240-50.

Verge V M K, Wiesenfeld-Hallin Z, Hokfelt T (1992) Cholecystokinin in mammalian primary sensory neurons and apinal cord: in situ hydridization studies on rat and monkey spinal ganglia. Eur J Neurosci 5: 1394-1400.

Villar M J, Cortes R, Theodorsson E, Wiesenfeld-Hallin Z, Schalling M, Fahrenkrug J, Emson P C, Hokfelt T (1989) Neuropeptide expression in rat dorsal root ganglion

235

cells and spinal cord after peripheral nerve injury with special reference to galanin. Neuroscience 33: 587-604.

von Bormann B, Boldt J, Sturm G, Kling D, Weidler B, Lohmann E, Hempelmann G (1985) Calcium antagonists in anesthesia. Additive analgesia using nimodipine in heart surgery. Anaesthesist 34: 429-34.

Wakisaka S, Kajander K C, Bennett G J (1992) Effects o f peripheral nerve injuries and tissue inflammation on the levels of neuropeptide Y-like immunoreactivity in rat primary afferent neurons. Brain Res 598: 349-52.

Walker D, De Waard M (1998) Subunit interaction sites in voltage-dependent Ca2+ channels: role in channel function. Trends Neurosci 21: 148-54.

Walker M W, Ewald D A, Pemey T M, Miller R J (1988) Neuropeptide Y modulates neurotransmitter release and Ca2+ currents in rat sensory neurons. J Neurosci 8: 2438-46.

Wall P D, Devor M (1983) Sensory afferent impulses originate from dorsal root ganglia as well as from the periphery in normal and nerve injured rats. Pain 17: 321- 39.

Wallace M S (2001) Pharmacologic treatment o f neuropathic pain. Curr Pain Headache Rep 5: 138-50.

Wamil A W, McLean M J (1994) Limitation by gabapentin of high frequency action potential firing by mouse central neurons in cell culture. Epilepsy Res 17: 1-11.

Wang S D, Goldberger M E, Murray M (1991) Plasticity o f spinal systems after unilateral lumbosacral dorsal rhizotomy in the adult mi. J Comp Neurol 304: 555-68.

Wang Y X, Gao D, Pettus M, Phillips C, Bowersox S S (2000a) Interactions of intrathecally administered ziconotide, a selective blocker of neuronal N-type voltage- sensitive calcium channels, with morphine on nociception in rats. Pain 84: 271-81.

Wang Y X, Pettus M, Gao D, Phillips C, Scott Bowersox S (2000b) Effects of intrathecal administration of ziconotide, a selective neuronal N-type calcium channel blocker, on mechanical allodynia and heat hyperalgesia in a rat model of postoperative pain. Pain 84: 151-8.

Waxman S G (2001) Transcriptional channelopathies: an emerging class of disorders. Nat Rev Neurosci 2: 652-9.

Wegert S, Ossipov M H, Nichols M L, Bian D, Vanderah T W, Malan T P, Jr., Porreca F (1997) Differential activities of intrathecal MK-801 or morphine to alter responses to thermal and mechanical stimuli in normal or nerve-injured rats. Pain 71 : 57-64.

Wei Z Y, Karim F, Roerig S C (1996) Spinal morphine/clonidine antinociceptive synergism: involvement o f G proteins and N-type voltage-dependent calcium channels. J Pharmacol Exp Ther 278: 1392-407.

236

Welch S P, Olson K G (1991) Opiate tolerance-induced modulation o f free intracellular calcium in synaptosomes. Life Sci 48: 1853-61.

Welty D F, Schielke G P, Vartanian M G, Taylor C P (1993) Gabapentin anticonvulsant action in rats: disequilibrium with peak drug concentrations in plasma and brain microdialysate. Epilepsy Res 16: 175-81.

Westenbroek R E, Hoskins L, Catterall W A (1998) Localization of Ca2+ channel subtypes on rat spinal motor neurons, intemeurons, and nerve terminals. J Neurosci 18: 6319-30.

Westlund K N, Carlton S M, Zhang D, Willis W D (1990) Direct catecholaminergic innervation of primate spinothalamic tract neurons. J Comp Neurol 299: 178-86.

Wheeler D B, Randall A, Tsien R W (1994) Roles of N-type and Q-type Ca2+ channels in supporting hippocampal synaptic transmission. Science 264: 107-11.

White D M, Cousins M J (1998) Effect of subcutaneous administration of calcium channel blockers on nerve injury-induced hyperalgesia. Brain Res 801: 50-8.

White G, Lovinger D M, Weight F F (1989) Transient low-threshold Ca2+ current triggers burst firing through an afterdepolarizing potential in an adult mammalian neuron. Proc Natl Acad Sci USA 86: 6802-6.

Wick M J, Minnerath S R, Lin X, Elde R, Law P Y, Loh H H (1994) Isolation of a novel cDNA encoding a putative membrane receptor with high homology to the cloned mu, delta, and kappa opioid receptors. Brain Res Mol Brain Res 27: 37-44.

Wiesenfeld-Hallin Z, de Arauja Lucas G, Alster P, Xu X J, Hokfelt T (1999) Cholecystokinin/opioid interactions. Brain Res 848: 78-89.

Wiesenfeld-Hallin Z, Xu X J, Langel U, Bedecs K, Hokfelt T, Bartfai T (1992) Galanin-mediated control of pain: enhanced role after nerve injury. Proc Natl Acad Sci U S A 89: 3334-7.

Wilcox G L, Alhaider A A (1990). Nociceptive and antinociceptive action o f serotonin agonists administered intrathecally. In Serotonin and Pain. ed. Besson, J.- M. pp. 205-219. Amsterdam: Elsevier Science.

Wilcox G L, Seybold V (1997). Pharmacology o f Spinal Afferent Processing. In Anesthesia: Biologic Foundations, ed. Yaksh, T.L., Lynch III, C., Zapol, W.M., Maze, M., Biebuyck, J.F. & Saidman, L.J. pp. 557-576. Philadelphia: Lippincott- Raven Publishers.

Williams M E, Feldman D H, McCue A F, Brenner R, Velicelebi G, Ellis S B, Harpold M M (1992) Structure and functional expression o f alpha 1, alpha 2, and beta subunits o f a novel human neuronal calcium channel subtype. Neuron 8: 71-84.

237

Wimalawansa S J (1996) Calcitonin gene-related peptide and its receptors: molecular genetics, physiology, pathophysiology, and therapeutic potentials. Endocr Rev 17: 533-85.

Wiser O, Trus M, Hernandez A, Renstrom E, Barg S, Rorsman P, Atlas D (1999) The voltage sensitive Lc-type Ca2+ channel is functionally coupled to the exocytotic machinery. Proc Natl Acad Sci U S A 96: 248-53.

Wiser O, Trus M, Tobi D, Halevi S, Giladi E, Atlas D (1996) The alpha 2/delta subunit of voltage sensitive Ca2+ channels is a single transmembrane extracellular protein which is involved in regulated secretion. FEES Lett 379: 15-20.

Witcher D R, De Waard M, Sakamoto J, Franzini-Armstrong C, Pragnell M, Kahl S D, Campbell K P (1993) Subunit identification and reconstitution of the N-type Ca2+ channel complex purified from brain. Science 261: 486-9.

W oolf C J (1983) Evidence for a central component o f post-injury pain hypersensitivity. Nature 306: 686-8.

W oolf C J (1994). The dorsal horn: state-dependent sensory processing and generation of pain. In Textbook o f pain. ed. Wall, P.D. & Melzack, R. pp. 101-112. London: Churchill Livingstone.

Woolf C J, Bennett G J, Doherty M, Dubner R, Kidd B, Koltzenburg M, Lipton R, Loeser J D, Payne R, Torebjork E (1998) Towards a mechanism-based classification of pain? Pain 77: 227-9.

W oolf C J, Mannion R J (1999) Neuropathic pain: aetiology, symptoms, mechanisms, and management. Lancet 353: 1959-64.

W oolf C J, Reynolds M L, Molander C, O'Brien C, Lindsay R M, Benowitz L I (1990) The growth-associated protein GAP-43 appears in dorsal root ganglion cells and in the dorsal horn of the rat spinal cord following peripheral nerve injury. Neuroscience 34: 465-78.

W oolf C J, Shortland P, Coggeshall R E (1992) Peripheral nerve injury triggers central sprouting of myelinated afferents. Nature 355: 75-8.

Woolf C J, Shortland P, Reynolds M, Ridings J, Doubell T, Coggeshall R E (1995) Reorganization of central terminals of myelinated primary afferents in the rat dorsal horn following peripheral axotomy. J Com/? Neurol 360: 121-34.

Wu L G, Saggau P (1994) Pharmacological identification of two types of presynaptic voltage-dependent calcium charmels at CA3-CAl synapses o f the hippocampus. J Neurosci 14: 5613-22.

Xiao W H, Bermett G J (1995) Synthetic omega-conopeptides applied to the site of nerve injury suppress neuropathic pains in rats. J Pharmacol Exp Ther 274: 666-72.

Xie Y, Zhang J, Petersen M, LaMotte R H (1995) Functional changes in dorsal root ganglion cells after chronic nerve constriction in the rat. J Neurophysiol 73: 1811-20.

238

Xu X J, Hao J X, Aldskogius H, Seiger A, Wiesenfeld-Hallin Z (1992) Chronic pain- related syndrome in rats after ischemic spinal cord lesion: a possible animal model for pain in patients with spinal cord injury. Pain 48: 279-90.

Xu X J, Hokfelt T, Bartfai T, Wiesenfeld-Hallin Z (2000) Galanin and spinal nociceptive mechanisms: recent advances and therapeutic implications. Neuropeptides 34: 137-47.

Xu X J, Puke M J, Verge V M, Wiesenfeld-Hallin Z, Hughes J, Hokfelt T (1993) Up-regulation o f cholecystokinin in primary sensory neurons is associated with morphine insensitivity in experimental neuropathic pain in the rat. Neurosci Lett 152: 129-32.

Yaksh T L (1989) Behavioral and autonomic correlates o f the tactile evoked allodynia produced by spinal glycine inhibition: effects o f modulatory receptor systems and excitatory amino acid antagonists. Pain 37: 111-23.

Yaksh T L (1997). Pharmacology and mechanisms of opioid analgesic activity. In Anesthesia, Biologic Foundations, ed. Yaksh, T.L., Lynch, C., Zapol, W., Maze, M., Biebuyck, J. & Saidman, L. pp. 921-34. Philadelphia,New York: Lippincott-Raven.

Yaksh T L, Hammond D L (1982) Peripheral and central substrates involved in the rostrad transmission of nociceptive information. Pain 13: 1-85.

Yaksh T L, Michener S R, Bailey J E, Harty G J, Lucas D L, Nelson D K, Roddy D R, Go V L (1988) Survey o f distribution of substance P, vasoactive intestinal polypeptide, cholecystokinin, neurotensin, Met-enkephalin, bombesin and PHI in the spinal cord of cat, dog, sloth and monkey. Peptides 9: 357-72.

Yamamoto C, Sawada S, Ohno-Shosaku T (1994) Suppression of hippocampal synaptic transmission by the spider toxin omega-agatoxin-IV-A. Brain Res 634: 349- 52.

Yamamoto T, Sakashita Y (1999) Differential effects o f intrathecally administered morphine and its interaction with cholecystokinin-B antagonist on thermal hyperalgesia following two models of experimental mononeuropathy in the rat. Anesthesiology 9Q: 1382-91.

Yamamoto T, Yaksh T L (1991) Spinal pharmacology o f thermal hyperesthesia induced by incomplete ligation of sciatic nerve. I. Opioid and nonopioid receptors. Anesthesiology 1S\ 817-26.

Yamamoto T, Yaksh T L (1992) Spinal pharmacology of thermal hyperesthesia induced by constriction injury o f sciatic nerve. Excitatory amino acid antagonists. Pain 49: 121-8.

Yang J, Tsien R W (1993) Enhancement o f N- and L-type calcium channel currents by protein kinase C in frog sympathetic neurons. Neuron 10: 127-36.

239

Yang S N, Larsson O, Branstrom R, Bertorello A M, Leibiger B, Leibiger I B, Moede T, Kohler M, Meister B, Berggren P O (1999) Syntaxin 1 interacts with the L(D) subtype o f voltage-gated Ca(2+) channels in pancreatic beta cells. Proc Natl Acad Sci U SA 96: 10164-9.

Yokota T, Nishikawa Y (1979) Action of picrotoxin upon trigeminal subnucleus caudalis neurons in the monkey. Brain Res 171: 369-73.

Yokoyama C T, Sheng Z H, Catterall W A (1997) Phosphorylation of the synaptic protein interaction site on N-type calcium channels inhibits interactions with SNARE proteins. J Neurosci 17: 6929-38.

Yoon M H, Yaksh T L (1999) The effect o f intrathecal gabapentin on pain behavior and hemodynamics on the formalin test in the rat. Anesth Analg 89: 434-9.

Yoon Y W, Na H S, Chung J M (1996) Contributions of injured and intact afferents to neuropathic pain in an experimental rat model. Pain 64: 27-36.

Yoon Y W, Sung B, Chung J M (1998) Nitric oxide mediates behavioral signs of neuropathic pain in an experimental rat model. Neuroreport 9: 367-72.

Yung K K (1998) Localization of glutamate receptors in dorsal horn of rat spinal coxà. Neuroreport 9'. 16?)9-AA.

Zamponi G W, Bourinet E, Nelson D, Nargeot J, Snutch T P (1997) Crosstalk between G proteins and protein kinase C mediated by the calcium channel alpha 1 subunit. Nature 385: 442-6.

Zamponi G W, Snutch T P (1998) Modulation o f voltage-dependent calcium channels by G proteins. Curr Opin Neurobiol 8: 351-6.

Zeltser R, Seltzer Z (1994). A practical guide for the use of animal models in the study o f neuropathic pain. In Touch, temperature and pain in health and disease: mechanisms and assessments, ed. Boivie, J., Hansson, P. & Linblom, U. pp. 2295- 338. Seattle.

Zhang J F, Randall A D, Ellinor P T, Home W A, Sather W A, Tanabe T, Schwarz T L, Tsien R W (1993 a) Distinctive pharmacology and kinetics of cloned neuronal Ca2+ channels and their possible counterparts in mammalian CNS neurons. Neuropharmacology ?>2: 1075-88.

Zhang X, Bao L, Shi T J, Ju G, Elde R, Hokfelt T (1998) Down-regulation of mu- opioid receptors in rat and monkey dorsal root ganglion neurons and spinal cord after peripheral axotomy. Neuroscience 82: 223-40.

Zhang X, Dagerlind A, Elde R P, Caste 1 M N, Broberger C, Wiesenfeld-Hallin Z, Hokfelt T (1993b) Marked increase in cholecystokinin B receptor messenger RNA levels in rat dorsal root ganglia after peripheral axotomy. Neuroscience 57: 227-33.

Zharkovsky A, Totterman A M, Moisio J, Ahtee L (1993) Concurrent nimodipine attenuates the withdrawal signs and the increase of cerebral dihydropyridine binding

240

after chronic morphine treatment in rats. Naunyn Schmiedehergs Arch Pharmacol 347: 483-6.

Zhuo M, Gebhart G F (1997) Biphasic modulation of spinal nociceptive transmission from the medullary raphe nuclei in the rat. J Neurophysiol 78: 746-58.

Zimmermann M (1983) Ethical guidelines for investigations of experimental pain in conscious animals [editorial]. Pain 16: 109-10.

Zurek J R, Nadeson R, Goodchild C S (2001) Spinal and supraspinal components of opioid antinociception in streptozotocin induced diabetic neuropathy in rats. Pain 90: 57-63.

241

PUBLICATIONS

242

Pain 92 (2001) 235-246

P A INwww.elsevier.nl/locate/pain

Effects of spinally delivered N- and P-type voltage-dependent calcium channel antagonists on dorsal horn neuronal responses

in a rat model of neuropathy

Elizabeth A. Matthews*, Anthony H. DickensonDepartment o f Pharmacology, University College London, Gower Street, London, WCIE 6BT, UK

Received 26 June 2000; received in revised form 29 November 2000; accepted 19 December 2000

Abstract

Neuropathic pain, due to peripheral nerve dam age, can include allodynia (perception o f innocuous stimuli as being painful), hyperalgesia (increased sensitivity to noxious stim uli) and spontaneous pain, often accom panied by sensory deficits. Plasticity in transmission and modulatory system s are im plicated in the im derlying m echanism s. The K im and Chung rodent m odel o f neuropathy (Kim and Chung, Pain 50 (1992) 355) em ployed here involves unilateral tight ligation o f tw o (L5 and L6) o f the three (L4, L5, and L 6) spinal nerves o f the sciatic nerve and reproducibly induced m echanical and cold allodynia in the ipsilateral hindpaw over the 14 day post-operative period. In vivo electrophysiological techniques have then been used to record the response o f dorsal horn neurones to innocuous and noxious electrical and natural (m echanical and thermal) stim uli after spinal nerve ligation (SNL ). A ctivation o f voltage-dependent calcium channels (V D C C s) is critical for neurotransmitter release and neuronal excitability, and antagonists can be antinociceptive. Here, for the first tim e, the effect o f N - and P-type V D C C antagonists (co-conotoxin-GVIA and w-agatoxin-IV A , respectively) on the evoked dorsal horn neuronal responses after neuropathy have been investigated. Spinal m -conotoxm -G VIA (0 .1 -3 .2 p.g) produced prolonged inhibitions o f both the electrically- and low - and high-intensity naturally-evoked neuronal responses in SN L and control rats. Spinal co-agatoxin-IVA (0 .1 -3 .2 p.g) also had an inhibitory effect but to a lesser extent. After neuropathy the potency o f w -conotoxin-G V IA w as increased at low er doses in comparison to control. This indicates an altered role for N -type but not P-type V D C C s in sensory transm ission after neuropathy and selective plasticity in these channels after nerve injury. B oth pre- and post-synaptic V D C C s appear, to b e important. © 2001 International A ssociation for the Study o f Pain. Pubhshed by E lsevier S cien ce B .V . AH rights reserved.

Keywords: Voltage-dependent calcium channels; Neuropathic pain; Conotoxins; Agatoxins; Nociception; Dorsal horn

1. Introduction

Neuropathic pain arises from injury- or disease-evoked damage to the peripheral or central nervous system and patients often experience sensory deficits, ongoing pain, and stimulus-evoked pain (allodynia and hyperalgesia). Animal models of neuropathy have been critical in elucidat­ing the complex causal mechanisms involving plasticity in nociceptive transmission and modulating systems. The spinal nerve ligation (SNL) model (Kim and Chung, 1992) involves tight hgation of two (L5 and L6) of the three spinal nerves that form the sciatic nerve. Behavioural consequences include thermal hyperalgesia, and mechanical and cooling allodynia (Kim and Chung, 1992; Chaplan et al., 1997).

* Coiresponding author. Tel.: 4-44-20-7679-3737; fax: 4-44-20-7679- 7298.

E-mail address: e.matthews@ucl.ac.uk (E.A. Matthews).

Sensory neurones express a number of classes of voltage- dependent calcium channels (VDCCs) (L, N, P, Q, R, and T), distinguished by their electrophysiological and pharma­cological profiles. N- and P-type VDCCs, sensitive to block by (u-conotoxin-GVIA and co-agatoxin-IVA, respectively (Olivera et al., 1994), are widely expressed throughout the brain and spinal cord (Kerr et al., 1988; Mintz et al., 1992; Gohil et al., 1994). The N-type channel is concentrated in laminae I and II of the superficial dorsal horn, where noci­ceptive primary afferents synapse (Kerr et al., 1988; Gohil et al., 1994). In vitro studies have demonstrated the require­ment of calcium ion influx through N- and P-type VDCCs for depolarization-coupled neurotransmitter release (Milja­nich and Ramachandran, 1995). Spinal release of primary afferent peptides, calcitonin gene-related peptide (CGRP) and substance P, is (u-conotoxin-sensitive (Holz et al., 1988; Maggi et al., 1990; Santicioli et al., 1992), and gluta­mate release is both co-conotoxin- and cu-agatoxin-sensitive

0304-3959/01/$20.00 © 2001 International Association for the Study of Pain. Phbhshed by Elsevier Science B.V. AH rights reserved. PH: S 0 3 0 4 -3 9 5 9 (0 1 )0 0 2 5 5 -X

243

236 E.A. Matthews, A.H. Dickenson / Pain 92 (2001) 235-246

(Dickie and Davies, 1992; Turner et al., 1992, 1993). Furthermore, calcium influx into neurones can enhance excitability and produce intracellular changes including gene induction.

VDCC antagonists are antinociceptive in models of in f la m m a tio n , based on behaviour and in vivo electrophy­siology, confirming a role for calcium influx into neurones in the spinal processing of nociceptive information. In the fo r m alin test, behavioural antiuQOciception is observed to varying degrees with specific N-, P-, and L-type VDCC blockers (Mahnberg and Yaksh, 1994, 1995; Bowersox et al., 1996), and Diaz and Dickenson (1997) reported phase- dependent effects of N- and P-type blockers on neuronal responses. Blockade of N- and P-type VDCCs (Neugebauer et al., 1996; Nebe et al., 1997, 1998; Sluka, 1997) has been shown to reduce enhanced neuronal responses after knee joint mflammation and intradermal capsaicin (Sluka,1998). Studies in models of neuropathy have been limited; however, calcium chelation (White and Cousins, 1998) and N-type VDCC antagonism have been shown to be effective (Chaplan et al., 1994; Xiao and Bennett, 1995; Bowersox et al., 1996; White and Cousins, 1998). In contrast, L- and P- type antagonists show little effect (Chaplan et al., 1994; White and Cousins, 1998).

Thus far, there are no in vivo electrophysiological studies investigating the role of spinal VDCCs in the processing of neuropathic pain. This study uses the SNL model, confirmed by behavioural testing, to induce a neuropathic state. After the establishment of neuropathy, electrophysiological studies of dorsal horn spinal neurones were made to inves­tigate the effects of spinally dehvered w-conotoxin-GVLA and co-agatoxin-IVA on a wide range of electrically- and naturally-evoked neuronal activity.—- - .

2. Materials and methods

2.1. SNL

Male Sprague-Dawley rats, initially weighing 130-150 g, were used in this study. All experimental procedures were approved by the Home Office and follow the guidelines under the International Association for the Study of Pain (Zimmerman, 1983). Selective tight ligation of spinal nerves L5 and L6 , and a sham procedure were performed as first described by Kim and Chung (1992). For details see Chapman et al. (1998).

2.2. Behavioural testing

For 2 weeks following surgery the rats were housed in groups of four in plastic cages under a 12 h day/night cycle and their general health was monitored. Successful repro­duction of the neuropathic inpdel was confirmed by beha­vioural testing (post-operative (PO) days 2, 3, 5, 7, 9, 12, and 14) by assessing the sensitivity of both the ipsilateral and contralateral hindpaws to normally non-noxious punc­

tate mechanical (von Frey filaments) and cooling (acetone) stimuli. Rats were placed in transparent plastic cubicles on a mesh floor and allowed to acclimatize before any tests were initiated. Foot withdrawals to von Frey filaments 1, 5, and 9 g (trials of ten) and acetone (trials of five) were quantified as described in Chapman et al. (1998) and expressed as difference scores = ipsilateral response — contralateral response.

2.3. Spinal cord electrophysiology

Subsequent to behavioural testing (PO days 14-17), the operated rats were used for electrophysiological studies (Dickenson and SuUivan, 1986). Briefly, anaesthesia was induced with 3% halothane in a mixture of 66% N2O and 33% O2 and a cannula was inserted into the trachea. A laminectomy was performed (vertebrae L1-L3) to expose segments L4 and L5 of the spinal cord and the level of halothane was reduced to 1.8%. Extracellular recordings of single convergent neurones located deep within the dorsal horn (>500 pum), receiving input from the toe region ipsi­lateral to the SNL or sham procedure, were made using a parylene coated tungsten electrode. Neurones selected responded to both noxious (pinch) and non-noxious (touch) stimuli.

2.3.1. Cell characterizationSpontaneous activity exhibited by a neurone was

recorded over 10 min. Action potentials evoked by natural stimuli apphed constantly over 10 s were quantified by the application of both punctate mechanical (voh Frey filaments 9 and 75 g) and thermal (constant water jet at 45°C) stimuli applied to the centre of the neurone’s receptive field. The thermal response to 45°C was determined by subtracting the response to 32°C (a nomnoxious^emperature so as to ascer­tain any mechanical response evoked by the water jet) from the response to 45°C. responses to natural stimuh were normalized by the subtraction of any spontaneous activity measured before the application of each stimulus. The response of the neurone to intracutaneous electrical stimu­lation. was established by insertion of two fine needles into the centre of its peripheral receptive field. A test consisted of a train of 16 stimuli (2 ms wide pulse at 0.5 Hz at three times the threshold required to evoke a C-fibre response), and a post-stimulus histogram was constructed. The thresholds were determined by increasing the electrical stimulus from 0 mA until an action potential was evoked in the corresponding latency band. Electrically-evoked action potentials were separated on a latency basis into A^-fibres (0-20 ms), A8-fibres (20-90 ms), C-fibres (90-300 ms) and post-discharge (300-800 ms). The ‘input’ (non-potentiated response), and the ‘wind-up’ (potentiated response, evident by increased neuronal excitability to repeated stimulation) were calculated as follows: input = (action potentials (90- 800 ms) evoked by first pulse at three times C-fibre threshold) X total number of pulses (16); wind-up = (total

244

E.A. Matthews. A.H. Dickenson / Pain 92 (2001) 235-246 237

a ctio n p o ten tia ls ( 9 0 - 8 0 0 m s) after 16 train stim u lu s at th ree tim es C -n b re th resh o ld ) — input.

n ever a cco m p a n ie d b y the p a in -lik e b eh av iou rs d isp la y e d

b y th e le s io n e d h in d p aw o f S N L rats.

2.3.2. Pharmacological studiesT h e testin g p ro to co l, in itia ted ev ery 15 m in , c o n s is ted o f

an e le c tr ic a l te st fo llo w e d b y th e natural s tim u li, as

d escr ib ed ab o v e . S ta b iliza tio n o f the n eu ron al r e sp o n ses w a s at le a st three co n s isten t pre-drug resp o n se s ( < 1 0 %

varia tion o v er at le a st 1 h) for all m easu res. T h e se v a lu e s w e re then avera g ed to g en era te pre-drug con tro l v a lu es w ith w h ic h to co m p a re the e ffe c t o f drug a d m in istra tion on su b seq u en t e v o k e d resp o n ses. T h e e ffe c t o f tw o d ifferen t

V D C C b lock ers w a s stu d ied on th ese resp o n ses. T h e N - ty p e c a lc iu m ch an n el b lo ck er , co -co n o to x in -G V IA (S ig m a -

A ld r ich C om p an y L td ., P o o le , D o rse t, U K ), and the P -ty p e c a lc iu m ch an n el b lo ck er , co -aga tox in -IV A (P ep tid e In sti­tu te , In c ., S c ie n tific M ark etin g A s s o c ia te s , B arn et, H ertford ­sh ire, U K ) (0 .1 , 0 .4 , 0 .8 , 1 .2, 2 .2 , and 3 .2 |xg), w ere ea ch

d is s o lv e d in sa h n e and a p p lied d irectly on to th e sp in a l cord in 5 0 fjcl v o lu m es . T h e e ffe c ts o f ea ch d o se w ere fo l lo w e d u n til th e r esp o n ses p la teau ed (a m in im u m o f 6 0 m in ), w h en th e n ex t d o se w o u ld be ap p lied cu m u la tiv e ly . T h e resu lts w ere ca lcu la ted as m a x im u m p ercen ta g e ch a n g es from the a v era g ed pre-drug v a lu e for ea ch neu ron e and the o v era ll resu lts for ea ch d o s e w ere e x p r essed as m ea n s ± standard error o f the m ean (S E M ) o f the n o rm a lized data. S ta tistica l a n a ly s is o f m a x im a l drug e ffec ts at each d o se co m p a red to th e avera g ed pre-drug v a lu e w a s determ in ed b y a pa ired r-test on raw data. A n un paired r-test on the n o rm a lized

data w as u sed for th e co m p a r iso n o f drug e ffe c ts b e tw een d ifferen t ex p erim en ta l groups. T h e le v e l o f s ig n ifica n ce w a s tak en as P s 0 .0 5 .

3. Results

3.1. Behavioural studies

D u rin g the PO p er io d th e a n im a ls sh o w ed norm al w e ig h t

ga in and m a in ta in ed g o o d gen era l health . R ats su b jec ted to S N L ex h ib ited abn orm al fo o t p ostu re ip s ila tera l to n erv e in jury w h ereb y to e s w ere h eld to g e th er in a ‘g u a rd in g ’ b e h a ­viou r . T h is d id not o c cu r in e ith er th e contralateral h in d p a w , or in the sh am -op era ted rats. S u c c e ss fu l rep lica tio n o f th e nerve injury m o d e l w a s co n firm ed b y b e h av iou ra l te s tin g w h ic h d em on stra ted th e d e v e lo p m e n t o f m ech a n ic a l and c o o h n g a llo d y n ia o f the in ju red h in d p aw o f S N L rats. E v o k e d a llo d y n ia in r esp o n se to iim o cu o u s m e c h a n ic a l (v o n F rey fila m en ts b e n d in g fo r c e 1 - 9 g ) and c o o lin g (a c e t­o n e ) stim u h w a s d isp la y e d as a brisk w ith d raw al, a c co m p a ­n ie d in so m e c a se s b y sh a k in g and h ck in g o f th e fo o t ip s ila tera l to S N L . T h is w as e v id e n t at PO day 2 , r ea ch e d a m a x im u m at P O d a y s 7 - 1 2 and w a s still m a in ta in ed at P O day 14 (F ig . 1). C o n s isten t w ith d raw al r esp o n se s w ere n ev e r

e x h ib ite d b y th e co n tro l group or b y the contralateral h in d ­p a w o f the e x p er im en ta l grou p , and w h en p resen t w ere

3.2. Spinal cord electrophysiology

3.2.1. Cell characterizationT h e n u m ber o f ipsila tera l dorsal h o m n eu ron es character­

iz e d in e a ch group w ere 23 in S N L rats, 21 in sh am -op era ted rats, and 3 7 in n a iv e rats (T ab le 1). A ll n eu ron es h ad a r ec ep tiv e fie ld o v er th e le ft ipsila tera l h in d p aw . T he le v e l o f o n g o in g sp on tan eou s a c tiv ity w a s fou n d to b e s ig n if i­can tly h ig h er (P < 0 .0 5 ) in n eu ron es in S N L rats com p ared

to sh am -op era ted rats. F urtherm ore, 71% o f n eu ron es ch ar­a c ter ized in S N L rats e x h ib ited sp on tan eou s a c tiv ity at a rate

greater than 0.1 H z in com parison , to o n ly 20% o f character­iz e d n eu ron es in sh am -op erated rats (P = 0 .0 0 3 , F ish er ’s ex a ct test). N o s ig n ifica n t d ifferen ce w as fou n d b e tw een ex p erim en ta l groups in the m ean c e ll depth o f recorded n eu ron es and the m ean neuronal resp o n ses ev o k e d by e le c ­trical and nam ral stim u lation .

3.2.2. Pharmacological studiesT h e e ffec t o f w -c o n o to x in -G V lA and co-agatox in -IV A

( 0 .1 - 3 .2 |xg), app hed d irectly o n to the sp inal cord , on the e le c tr ic a lly - and n a tu ra lly -evok ed dorsal h o m neuronal resp o n ses w as te sted in S N L , sh am -op erated and n a iv e an im als. S in ce the sh am -op eration is the appropriate control

7 -,

6-

5 -

4 -

3 -

2-

1 -

0 -

0 o%1Û

-1 - 1---- 1-----1-----1-----18 10 12 14 16

Post operative D ay

Fig. 1. Development of mechanical and cooling allodynia in the ipsilateral hindpaw over the 2 week period following SNL- Data are expressed as the mean difference score ± SEM (n = 15) in the withdrawal response to punctate mechanical stimuli (von Frey filaments bending forces 1 (■), 5 (A) and 9 g ( • ) ) and cooling stimulus (drop of acteone (♦ )) applied to the plantar surface of the hindpaws (trials of ten for the mechanical and five for the cooling). Difference score = (ipsilateral response frequency) — (contralateral response frequency). No nociceptive behaviour was observed in the sham-operated rats (data not shown).

245

238 E.A. Matthews, A.H. Dickenson / Pain 92 (2001) 235-246

Table 1Comparison of dorsal hom neuronal characteristics in naive, sham-operated and SNL rats at PO days 14—17*

Naive rats Sham-operated rats SNL rats

Cell depth (|xm) 778 ± 31 916 ± 40 883 ± 42Spontaneous activity (Hz) 0.83 ± 0.69 0.34 ± 0.22 6.36 ± 2.24"*C-Fibre threshold (mA) 1.49 ± 0.13 1.21 ± 0.16 1.38 ± 0 .1 3C-Fibre response (APs) 330 ± 22 305 ± 24 321 ± 31A^-Fibre response (APs) 102 ± 6 107 ± 9 116 ± 7Wind-up response (AP) 243 ± 31 263 ± 31 268 ± 4 1Von Frey 9 g (APs) 152 ± 33 137 ± 29 173 ± 43Von Frey 75 g (APs) 678 ± 66 563 ± 76 544 ± 6 945°C heat (APs) 364 ± 6 3 265 ± 49 237 ± 48

* Data are expressed as the mean ± SEM. *Statistical significance (P z 0.05). Recorded neurones had receptive fields over the hindpaw ipsilateral to surgery and similar depths from the spin^ cord surface in naive (n = 37), sham-operated (n = 21) and SNL (n = 23) groups. The frequency of spontaneous neuronal activity was significantly higher in SNL rats compared to sham-operated rats. All groups had similar neuronal responses (action potentials (APs)) evoked by transcutaneous electrical stimulation at three times the threshold required to evoke a C-fibre response, which was delivered via two fine needles inserted into the centre of the cell’s receptive field, and similar C-fibre thresholds. Likewise, the experimental groups had comparable neuronal responses evoked over 10 s by natural stimuli (von Frey filaments bending force 9 and 75 g and 45°C water jet normalized by subtraction of response to 32°C) applied to the centre of the cell’s receptive field.

for SNL, and for clarity, the results obtained from the naive, non-operated group shall not be displayed on the graphs.

3.2.2.1. Effects o f co-conotoxin-GVIA. At the higher doses, m-conotoxin-GVLA produced inhibitions of all the electrically-evoked responses in neurones in all experimental groups (Fig. 2). Interestingly, in the sham- operated group, low doses tended to facilitate the neuronal responses, especially those induced by electrical stimulation. For each dose (0.1, 0.4, 0.8, 1.2, 2.2, and 3.2 fxg) the effects of the toxm were slow in onset with clear effects seen around 40 min, and maximal inhibitions estabhshed at 70-90 min (Fig. 3), which were irreversible. In SNL rats, all doses of co-conotoxin-GVIA (0.1-3.2 |xg) elicited significant effects compared to pre-drug control of the C-fibre, wind-up, input, post-discharge and AB-fibre responses (Fig. 2a-e, respectively, P ^ 0.05, n = 8), with similar mean maximal inhibitions at top dose in the range 86 ± 6% to 96 ± 3%. The A^-fibre response was signihcantly inhibited with 0.4—3.2 p.g co-conotoxin- GVIA, but only reached a maximal 41 ± 10% inhibition (Fig. 2f).

In contrast to SNL animals, sham-operated rats required higher concentrations of the toxin to achieve significant inhibition of the electrically-evoked responses, but still reached similar maximal levels of inhibition at the top dose. co-Conotoxin-GVlA (0.8-3.2 jjug) significantly inhib­ited the post-discharge and AB-fibre responses (Fig. 2d,e, P < 0.05, n = 6), whereas the C-fibre, wind-up and input responses (Fig. 2a-c) were sigitificantly inhibited over the dose range 1.2-3.2 p,g {P ^ 0.05, n = 6). In a similar fash­ion to the sham-operated group, naive animals also required higher concentrations of the toxin to achieve significant inhibition of the electrically-evoked responses. cu-Cono- toxin-GVlA (0.8-3.2 p,g) significantly inhibited the C- fibre, input and AB-fibre responses {P ^ 0.05, « = 1 1 ) ^ d 2.2-3.2 fjLg produced a significant inhibition (JP ^ 0.05,

n = 11) of the post-discharge and wind-up responses. Maxi­mum inhibitions ranged from 69 ± 12% to 89 ± 6%. As in SNL rats, co-conotoxin-GVIA had limited effects on the A^- fibre response in sham (Fig. 2f) and naive rats where the mean maximum inhibitions were 25 ± 7% and 44 ± 6%, respectively.

After neuropathy the inhibitory effects of co-conotoxin- GVIA on the electrically-evoked responses appeared to be more marked at the lower end of the dose range in compar­ison to sham-operated rats. This was statistically significant (P ^ 0.05, « = 8) at 0.1 and 0.4 |xg for the C-fibre, wind-up, input, and post-discharge responses (Fig. 2a-d) and at 0.4 |jLg for the AB-fibre response (Fig. 2e). co-Conotoxin-GVlA (0.1 and 0.4 p,g) had a significantly greater effect (P ^ 0.05) on the C-fibre response (Fig. 2a) in SNL rats (29 ± 5 % and 47 ± 9% inhibitions, respectively, « = 8) compared to sham-operated rats (7 ± 10% and 5 ± 15% facilitations, n = 7), the wind-up response (Fig. 2b) m SNL rats (63 ± 12% and 77 ± .8% inhibitions, respectively, n = 8) compared to sham-operated rats (14 ± 10% andl5 ± 17% inhibitions, n = 7), the.input response (Fig. 2c) in SNL rats (30 ± 5% and 54 ± 9% inhibitions, respectively, « = 8) compared to sham-operated rats (19 ± 15% and 14 ± 21% facilitations, n — 7), and the post-discharge (Fig. 2d) in SNL rats (50 ± 13% and 67 ± 10% inhibitions, respectively, n = 8) compared to sham-operated rats (6 ± 11% and 20 ± 16% inhibitions, n = 6). co-Conotoxin-GVIA (0.4 |xg) had a significantly greater effect (P < 0.05) on the AB-fibre (Fig. 2e) in SNL rats (50 ± 10% inhibition, « = 8) compared to sham-operated rats (21 ± 19% inhibi­tion, « = 5).

w-Conotoxin-GVlA also produced an inhibitory effect on the naturally-evoked neuronal responses (Fig. 4). At 3.2 |xg cü-conotoxin-GVlA, the response to non-noxious mechani­cal stimulation (von Frey 9 g) was inhibited by 95 ± 3% in both SNL (n = 5) and sham-operated rats (n = 5), and by 75 ± 6% in naive rats (« = 10) (Fig. 4a, naive not shown).

246

E.A. Matthews, A.H. Dickenson/Pain 92 (2001) 235-246 239

lOOn

7 5 -

5 5 0 -

2 5 -

-25-10.1 1 10

100-,

7 5 -

5 0 -

0

-25J

100-1lOOn

7 5 -7 5 -

o 5 0 -5 0 -

2 5 -

0 -

-25 J

100-1 100 -,

7 5 - 7 5 -

o 5 0 - o 5 0 -

2 5 - 2 5 -

0 -

0.1 10ü)-conotox in -G V IA (jig)

0.1 1co-conotox in -G V IA (^ g)

10

Fig. 2. Comparison of the effect of spinally applied w-conotoxin-GVIA on the electrically-evoked dorsal hom neuronal responses (see Section 2) recorded from SNL rats ( • ) and sham-operated rats (A) at PO days 14—17. After SNL (n = 8), the inhibitory effects of co-conotoxin-GVIA were more marked at the lower doses in comparison to sham-operated rats (n = 7) and this reached significance for (a) C-fibre, (b) wind-up, (c) input, (d) post-discharge and (e) A5-fibre responses. Significance was not reached for (f) A(3-fibre response. Data are expressed as the maximal mean % inhibition of the pre-drug values ± SEM. A significantly greater inhibitory effect of the toxin at a specific dose in SNL rats compared to sham-operated rats is denoted by *P ^ 0.05.

The response to noxious mechanical stimulation (von Frey 75 g) was inhibited by 83 ± 5% in SNL rats (n = 7), 78 ± 12% in the control group {n = 6), and 69 ± 5% in naive rats (n = 10) (Fig. 4b, naive not shown). The response to noxious thermal stimulation (water jet at 45°C) was inhib­ited by 89 ± 5% in SNL rats in = 6), 93 ± 5% in sharn- operated rats (n = 5) and 79 ± 9% in naive rats (n = 8)

(Fig. 4c, naive not shown). After neuropathy, co-cono- toxin-GVIA had a greater inhibitory effect on the natu­rally-evoked neuronal responses at the lower end of the dose range compared to sham-operated animals. This was found to be statistically significant at 0.1 and 0.4 p,g w- conotoxin-GVIA for the von Frey 9 g evoked responses (Fig. 4a) (55 ± 6% and 69 ± 7% inhibitions, respectively.

247

240 E.A. Matthews, A.H. Dickenson / Pain 92 (2001) 235-246

Pre-drug CD-conotoxin G VIA (cumulative dose)

O.lfig 0.4)ig 0.8ng150-,

100 -

I

i-fibre

■ C-fibre

■von Frey75g0-J4000 100 200 300 500

um e (min)

Fig. 3. Time course of the effect of spinally applied w-conotoxin-GVIA on the evoked response of a typical dorsal hom neurone (see Section 2) recorded from a sham-operated rat. Examples of the effect on the A|3- fibre (A), C-fibre (O) and von Frey 75 g ( • ) measurements are shown with the cumulative dose indicated.

in S N L rats, c om p ared to 6 ± 12% fa c ilita tio n and

17 ± 11% in h ib itio n , r e sp e c tiv e ly , in sh am -op era ted rats; P < 0 .0 0 1 , n = 8 ), and at 0 .4 p.g for the v o n Frey 7 5 g e v o k e d resp o n ses (F ig . 4 b ) (51 ± 6% in h ib it io n in S N L rats co m p a red to 21 ± 12% in h ib ition in sh am -op era ted

rats; P s 0 .0 5 , n — 8).

3.2.2.2. Effects o f co-agatoxin-IVA. cu -A gatox in -IV A in h ib ited the e le c tr ic a lly -ev o k ed n eu ron al r esp o n ses m easu red in a ll ex p erim en ta l grou p s, but to a le sser ex ten t than that a c h ie v e d w ith cu -con otox in -G V IA (F ig . 5). In S N L rats, at the h ig h es t d o se , m ean m a x im a l in h ib it io n s from pre-drug c o n tro ls w ere in the ran ge 3 7 ± 13% to 6 0 ± 10% for w in d -u p , input, p o st-d isc h a rg e and A ô -fib re (F ig . 5 b - e ) , and o n ly 18% for th e C - and A (3-fibre (F ig . 5 a ,f , res p e c tiv e ly ) , at w h ic h s ig n ifica n ce w a s a ch ie v e d {P :S 0 .0 5 , n - 7 ). S im ila r m a x im a l in h ib it io n s w e re o b serv e d in sh am -op era ted rats, and th ere w as no d ifféren ce in the drug e ffe c t at the lo w er d o s e s com p ared to S N L a n im a ls. m -A g a to x in -IY A had n o s ig n ifica n t inh ib itory e ffe c ts on the resp o n ses in n a iv e an im als; h o w ev e r , w ith the lo w d o se s (0 .1 -0 .8 p,g) there w a s a te n d en cy fo r th e e le c tr ic a lly -e v o k e d n euronal r esp o n ses to be fa c ih ta ted and th is w a s fou n d to b e sta tis tic a lly s ig n ifica n t (P < 0 .0 5 , n = 5) fo r th e C -fibre resp o n se .

T h e n a tu ra lly -e v o k e d n eu ron al r esp o n se s w e re a lso in h ib ­ited b y m -a g a to x in -rV A (F ig . 6 ). In S N L rats, at the h ig h est d o se , m ean m a x im a l in h ib it io n s from th e p re-drug con tro l w ere in th e range 5 7 ± 11% to 71 ± 12% , and th is w a s sta tis tica lly s ig n ifica n t fo r v o n F rey 9 and 7 5 g (P s 0 .0 5 , n = 7 ). A g a in , n o d iffere n ce in the ex ten t o f in h ib it io n w a s o b serv ed b e tw e e n ex p erim e n ta l groups.

T h e sp o n ta n eo u s rate o f a c tiv ity , w h en presen t, w a s c lea r ly in h ib ited b y both to x in s . D u e to the lo w pre-drug b a se lin e v a lu e s and var ia b ih ty b e tw een n eu ro n es in th is o n g o in g a c tiv ity , it w a s n o t qu antified .

1 0 0 -,

7 5 -

.2 5 0 -

2 5 -

-25 -1

0.1 1 10100 n

7 5 -

= 2 5 -

- 2 5 -J

0.1 1 10

100 n

7 5 -

c 2 5 -

-25 J

0.1 1 10ü )-co n o to x in -G V IA (p g )

Fig. 4. Comparison of the effect of spinally applied w-conotoxin-GVIA on the naturally-evoked dorsal hom neuronal responses (see Section 2) recorded from SNL rats ( • ) and sham-operated rats (A), (a) Innocuous punctate mechanical stimulus (von Frey 9 g); (b) noxious punctate mechan­ical stimulus (von Frey 75 g); (c) noxious heat stimulus (water jet at 45°C normalized by response to water jet at 32°C). After SNL (n = 8), the inhibitory effects of co-conotoxin-GVIA were more marked at the lower doses in comparison to sham-operated rats (n = 7) and this reached signif­icance for the punctate mechanical evoked responses (a,b). Data are expressed as the maximal mean % inhibition of the pre-drug values ± SEM. A significantly greater inhibitory effect of the toxin at a specific dose in SNL rats compared to sham-operated rats is denoted by

< 0.05 and **P S 0.001.

248

E.A. Matthews, A.H. Dickenson / Pain 92 (2001) 235-246 241

1 0 0 -, 100

7 5 -7 5 -

,2 5 0 -.2 5 0 -

2 5 -

0-

-25 J-25 J ■n10 0.10.1 1 101

100-1100-1

7 5 -7 5 -

o 5 0 -2 5 0 -

2 5 -

-25-1-25-110 0.10.1 1 1 10

1 0 0 -,

7 5 -

.2 5 0 -

2 5 -

-25-1

100.1 1

100 -,

75 -

.2 5 0 -

- 2 5 -

0.1 1 10(D-agatoxin-IV A (^g) Cü-agatoxin-IVA ()U.g)

Fig. 5. Comparison of the effect of spinally applied o)-agatoxin-IV A on the electrically-evoked dorsal hom neuronal responses (see Section 2) recorded from SNL rats ( • ) and sham-operated rats (A) at PO days 14—17. (a) C-Fibre, (b) wind-up, (c) input, (d) post-discharge, (e) A5-fibre, and (f) Ap-hbre responses. No difference in the effects of (u-agatoxin-IVA was observed between SNL rats {n = 6) and sham-operated rats (n = 6). Data are expressed as the maximal mean % inhibition of the pre-drug values ± SEM.

4. Discussion

This is the first electrophysiological study addressing the role of VDCCs in the spinal processing of sensory informa­tion after nerve injury. Spinal ou-conotoxin-GVIA, an N- type blocker, inhibited the electrically- and naturally- (innocuous and noxious) evoked dorsal hom neuronal responses. The effects of low doses were significantly

enhanced after the establishment of neuropathy. In compar­ison, the P-type antagonist cu-agatoxin-IVA inhibited the evoked neuronal responses to a lesser extent and its profile was unaltered after nerve injury.

Although the concentration of the toxins at their sites of action in vivo obviously cannot be determined, the concen­tration apphed to the surface of the spinal cord at a mid­range dose (0.8 pug) was 5 |xM for co-conotoxin-GVIA and 3

249

242 EA. Matthews, A.H. Dickenson / Pain 92 (2001) 235-246

100 n

7 5 -

co

-2 5 J

0.1 10100-, b

7 5 -

- 2 5 -J

0.1 1 10lOOn

75 -

5 0 -o

I 2 5 -

- 2 5 -

0.1 1 10Cü-agatoxin-IVA (jig)

Fig. 6 . Comparison of the effect of spinally applied m-agatoxin-IVA on the naturally-evoked dorsal horn neuronal responses (see Section 2) recorded from SNL rats ( • ) and sham-operated rats (A), (a) Iimocuous punctate mechanical stimulus (von Frey 9 g); (b) noxious punctate mechanical stimulus (von Frey 75 g); (c) noxious heat stimulus (water jet at 45°C normalized by response to water jet at 32°C). No statistically significant diiference in the effects of w-agatoxin-IVA was observed between SNL rats [n = 6) and sham-operated rats (n = 6). Data are expressed as the maximal mean % inhibition of the pre-drug values ± SEM.

|xM for co-agatoxin-IVA. These concentrations are higher than the selective dose ranges for these toxins in vitro (Hofmann et al., 1999), where there is direct application to the membrane. However, the concentration within the spinal cord wül be at least three orders of magnitude less and also diffusion of these large peptides into the tissue will be hmited as demonstrated by the slow establishment of maximal inhibitions (Fig. 3). The different profile of the toxins suggests that they are acting at the appropriate VDCC.

Unilateral tight ligation of L5 and L6 spinal nerves produced reproducible nociceptive syndromes in the lesioned hindpaw. A clear withdrawal reflex with associated aversive behaviours, indicative of the development of mechanical and cooling allodynia, was produced as previously described by Chapman et al. (1998). Thus, all SNL animals used for the electrophysiology and subsequent pharmacology exhibited neuropathic signs; sham-operated animals did not.

A possible explanation of these results is that if there are increases in the probability and/or frequency of opening of pre-existing VDCCs, the likelihood of a given concentration of (u-conotoxin-GVIA blocking a functionally important channel would be increased. Indeed, it is likely that calcium influx via VDCCs on central terminals of peripheral nerves and neurones, and consequently neurotransmitter release and membrane depolarization (Smith and Augustine, 1988), are increased due to enhanced neuronal activity observed peripherally after neuropathy (Wall and Devor, 1983; Sheen and Chung, 1993). Furthermore, despite reduced afferent input via L5 and L6 spinal nerves, the magnitude of neuronal responses recorded was not dimin­ished after SNL. Conversely, increased frequency and occurrence of spontaneous activity was observed. These points have been recently discussed in more detail (Suzuki et al., 2000a) but the neuronal profiles seen after nerve injury are similar to those previously reported in this model (Chapman et al., 1998). Furthermore, in keeping with these ideas, we have recently shown enlarged receptive fields of these neurones (Suzuki et al., 2000b). This suggests that there are compensatory increases in peripheral and/or spinal neuronal activity after neuropathy leading to a greater functional role of VDCCs in the observed neuronal responses.

Another possibility is that rather than an increase in activ­ity of existing VDCCs there is an increase in VDCC expres­sion. In support of this idea, the expression of the N-type VDCC a IB subunit is increased in small dorsal root gang- hon (DRG) cells and in lamina H of the spinal cord after neuropathy (Cizkova et al., 1999). Furthermore, the nxRNA and protein for the auxiliary «28 subunit of VDCCs is also upregulated in the ipsilateral DRG and spinal cord of SNL rats (Luo et al., 1999). In addition to its ability to modulate VDCC kinetics (Hobom et al., 2000), a 2§ can modulate the binding affinity of w-conotoxins to the N-type VDCC (Brust et al., 1993). Although it is generally beheved that the

250

EA. Matthews, AH . Dickenson / Pain 92 (2001) 235-246 243

cù-toxins primarily affect the functioii of the a i subunit (Olivera et al., 1994), evidence indicates that they can also bind-the a ; component of the a 20 subunit (Barhanin et al., 1988). It remains to be determined whether upregula- tion of is responsible for the selective enhancement of co-conotoxin-GVIA’s inhibitory actions we report here.

Peripheral nerve damage leads to the generation of ecto­pic activity at the injury site (Wall and Devor, 1983) or within the DRG (Wall and Devor, 1983; Kajander et al.,1992). This provides an ongoing, continual barrage of neuronal activity via primary afferents into the spinal cord which could lead to central sensitization and hyperexcitabü- ity (Gracely et al., 1992; Sheen and Chung, 1993; Yoon et al., 1996). Evidence indicates that enhanced nociceptive transmission in neuropathic pain states is mediated, at least in part, by the action of excitatory amino acids, parti­cularly at //-methyl-D-aspartate (NMDA) receptor sites (Dubner and Bennett, 1983; Dickenson and Sullivan, 1987; Coderre et al., 1993), and NMDA antagonists have been shown to produce antinociception in patients (Bide et al., 1994; Price et al., 1994). Thus, any increased neuronal depolarization, such as that produced by NMDA receptor events, would increase the likelihood of VDCC activity in spinal neurones since they are voltage-operated ion chan­nels. Indeed events such as these regulate the expression of new channels and receptor subunits.

4.1. Roles o f N-type VDCCs

It is evident here that blockade of N-type VDCCs results in a greater inhibition of electrically- and naturally-evoked dorsal horn neuronal responses after SNL, which is indica­tive of a role for N-type VDCCs in mechanisms ofxentral sensitization of nociceptive dorsal horn neurones after neuropathy. Increased primary ^ e re n t activity after periph­eral nerve damage could result in enhanced exocytosis of excitatory transmitters, which would in turn increase activa­tion of receptor systems such as those for glutamate (NMDA), substance P and CGRP, which produce excitation of spinal cord neurones. Nerve terminals throughout the dorsal horn have been demonstrated to be immunoreactive for N-type VDCCs, often correlating with the presence of substance P (Westenbroek et al., 1998). Since N-type VDCCs are abundant on nerve terminals and crucial for neurotransmitter release (Smith and Augustine, 1988; Miljanich and Ramachandran, 1995) the simplest explana­tion for the effects of co-conotoxin-GVIA is that it blocks pre-synaptic N-type VDCCs and therefore calcium influx and evoked neurotransmitter release. This is supported by the observation that the non-potentiated ‘input’ response recorded in this study, which would relate to the level of synaptic transmission between the central terminals of primary afferents and the neurones of the spinal cord dorsal horn, is significantly inhibited by the spinal application of the toxin. Small DRG cells express N-type calcium currents (Scroggs and Fox, 1992a,b) and the release of CGRP (Santi-

cioli et al., 1992); sensory neuropeptides (Maggi et al., 1990) and substance P in culture (Holz et al., 1988) are all ü)-conopeptide-sensitive. As for the control of glutamate release and subsequent NMDA receptor/channel activation, implicated in central hyperexcitability, the role of N-type VDCCs is debatable. Glutamatergic synaptic transmission between DRG cells and spinal cord neurones is inhibited by blockade of pre-synaptic N-type VDCCs (Gruner and Silva, 1994), and w-conopeptides can cause significant reduction in depolarization-evoked glutamate release (Terrian et al., 1990; Dickie and Davies, 1992). Given the evidence for a role of N-type VDCCs in peptide release it would be expected that a similar situation would exist for glutamate, although other studies have reported that N-type VDCCs have a minimal role in mediating glutamate release (Pocock and NichoUs, 1992; Turner et al., 1992).

It is also likely that w-conotoxin-GVIA may also inhibit N-type VDCCs at post-synaptic sites (Takahashi and Momiyama, 1993) and thereby reduce excitation of sjpinal cord neurones. Nociceptive signals arriving at the spinal cord from damaged peripheral nerves may be subject to amplification via intemeurones that project onto an output neurone, resulting in the phenomena of wiad-up and central sensitization. In support of a possible intemeuronal site of action is the fact that NMDA receptors are not considered to be directly post-synaptic to primary afferents (Davies and Watkins, 1983). Also, the greatest inhibitions seen here produced by the low doses of w-conotoxin-GVIA after SNL were observed on the ‘wind-up’ and post-discharge responses, indicators of neuronal hyperexcitability, and mediated by post-synaptic events. However, in the latter case, blockade of pre-synaptic transmitter release by the VDCC antagonists could indirectly also prevent post-synap­tic hyperexcitability.

4.2. Roles o f P-type VDCCs

In contrast to the actions of w-conotoxin-GVIA, the P- type VDCC antagonist co-agatoxin-IVA had limited inhibi­tory effects and no difference was seen between experimen­tal groups. Interestingly, a significant inhibition of wind-up and post-discharge was seen with higher doses in sham- operated and SNL rats, althoujgh the maximum effect was lower than after the co-conotoxin-GVIA. In naive and sham- operated animals, the electricaUy-evoked responses tended to be facilitated at low doses and similarly in an m vivo electrpphysiological study (Nebe et al., 1997), co-agatoxin- IVA increased neuronal responses in the absence of inflam­mation but mediated inhibition in its presence. In vitro studies have demonstrated that co-agatoxin-IVA blocks excitatory (Luebke et al., 1993; Castillo et al., 1994; Yama­moto et al., 1994) as well as inhibitory (Takahashi and Momiyama, 1993) synaptic transmission, which may explain the facüitatory and inhibitory effects observed here. P-type VDCCs may be involved in the release of glutamate, aspartate, dopamine, serotonin, noradrenaline.

251

244 E.Al Matthews, A.H. Dickenson / Pain 92 (2001) 235-246

GABA and glycine (Turner et al., 1992, 1993; Takahashi and Momiyama, 1993; Kimura et al., 1995; Miljanich and Ramachandran, 1995).

(ü-Agatoxin-IVA given prior to capsaicin into the hind­paw (Sluka, 1997) and cairageenan into the knee joint (Sluka, 1998) prevents the development of hyperalgesia. This indicates that P-type chaimels may be most important in the initiation of a facihtated pain state although we find a Limited role in the maintenance of neuropathy. The P-type channel appears to be hnked selectively to NMDA receptor- mediated events as here post-SNL-administered co- agatoxin-IVA had most marked effects on NMDA recep­tor-mediated wind-up and post-discharge measurements. This is in accordance with results with the formalin response where co-agatoxin-IVA only inhibited neuronal responses in the second phase (Diaz and Dickenson, 1997), for which NMDA receptor activity is critical (Haley et al., 1990).

Evidence to support a post-synaptic role for P-type VDCCs is based on the fact that pOst-synaptic currents in numerous central nervous system neurones are markedly suppressed by co-agatoxin-IVA and reduced to a lesser extent by co-conotoxin-GVIA (Takahashi and Momiyama,1993). Pre-synaptic P-type channels are also present. Inter­estingly, in the superficial dorsal horn, the distribution of P/ Q-type channels complements rather than co-localizes with N-type VDCC-containing neurones. Immunolocalization has demonstrated the presence of P/Q-type VDCCs primar­ily on the nerve terminals of dorsal horn neurones in lami­nae H -VI, but unlike N-type channels, rarely in substance P- containing neurones (Westenbroek et al., 1998). Intiigu- ingly, this may indicate that N-type channels regulate the release of neurotransmitters, including glutamate, from peptide-containing. C-fibres,.--whereas—P--type .-channels could possibly be localized to non-peptide, IB4-positive neurones. Furthermore, both N- and P-type VDCCs are found in the deeper laminae of the dorsal horn, suggesting localization on terminals of large fibres and spinal neurones.

4.3. Implications

From this study, in agreement with studies in inflamma­tion, the spinal N-type VDCC is the predominant isoform involved in the pre- and post-synaptic processing of sensory nociceptive information. This is the first electrophysiologi- cal study investigating the role of N- and P-type VDCCs in the transmission of nociception at the level of the spinal cord after nerve injury and it extends results from beha­vioural studies. Tactile allodynia in the SNL model is blocked by intrathecal N-type blockers, SNX-111, -239 and -159, without effect when administered intravenously or regionally to the nerve. w-Agatoxin-IVA and SNX-230 (synthetic P/Q-type selective conopeptide homologue) were inactive (Chaplan et al., 1994; Bowersox et al., 1996). In another model of nerve injury, SNX-Tll and -124 reduced heat hyperalgesia and mechano-allodynia when delivered directly to the site of injury (Xiao and Bennett, 1995).

Mechanical hyperalgesia in rats with partial nerve ligation was reduced by subcutaneous injection of SNX-111 into the receptive field, whereas SNX-230 had no effect. Neither compound had any effect in control animals (White and Cousins, 1998). We found a broad spectrum of activity of the N-type channel blocker encompassing the single modal­ities used in these behavioural studies. It has been suggested that N-type channels are located predominantly on pre- synaptic terminals with a smaller proportion found post- synaptically. P-type channels appear to have functional roles both pre- and post-synapticaUy. If this holds for sensory pathways m the dorsal horn, then the present results would suggest that the shift in the dose-response curve for the N-type VDCC but not the P-type VDCC is indicative of considerable plasticity in the N-type VDCCs pre-synapti- caUy with less pronounced changes at post-synaptic sites. It may then be that the peripheral neuropathy selectively impacts upon the terminals of the injured and remaining non-injured peripheral neurones. The increased activity of spinal N-type VDCCs would be expected to lead to increased transmitter release promoting spinal mechanisms of hyperexcitability that contribute to the ensuing neuro­pathic syndromes. ;

N-type VDCCs have been demonstrated to be analgesic in both animal models of acute and chronic pain, and in humans. The doses utilized here are comparable to those used in behavioural studies where no obvious side-effects were noted (Malmherg and Yaksh, 1994,1995). In a clinical study, SNX-111 delivered intrathecally to chronic pain patients produced some pain relief (Brose et al., 1997). However, pharmacological targeting of these channels with toxin antagonists as a potential therapy for the treat­ment of chronic pain is hindered by the adverse systemic side-effects (Penn and Paice, 2000) and the inconvenient spinal route of adrriinistration.

It is noteworthy that the anti-epüeptic drug gabapentin (GBP), with analgesic efficacy in patients with postherpetic neuralgia (Rowbotham et al., 1998) and diabetic polyneuro­pathy (Backonja et al., 1998), may exert its effects via bind­ing to VDCCs. Despite being a GABA analogue it has been demonstrated to bind to the auxiliary subunit of VDCCs (Taylor et al., 1998) where it is assumed to act as an antago­nist. GBP may highhght the importance of VDCCs as targets in pain control. It is difficult to envisage how inter­actions of GBP with ubiquitous «28 subunits can result in specific therapeutic effects. However, DRG cells have now been shown to express at least two distinct forms of the a 2Ô subunit that differ from those in spinal cord and brain (Luo, 2000). This may allow tissue-specific effects of agents that act on VDCCs. ’

Whatever the caSe, the present results, together with a number of previous behavioural reports, emphasize the key roles of VDCCs in sensory events within the spinal cord. Here, we show that N-type VDCCs are the predomi­nant functional type in terms of a wide range of sensory modalities and exhibit plasticity following nerve injury.

252

E.A. Matthews, A.H. Dickenson / Pain 92 (2001) 235-246 245

The increased ability of (o-conotoxin-GVIA to inhibit evoked neuronal responses to a wide range of stimuli suggests that N-type VDCCs are a key component of increased central excitability that follows nerve injury.

Acknowledgements

E.A.M. is supported by the Wellcome Trust Four Year Neuroscience PhD Programme.

References

Backonja M, Beydoun A, Edwards KR, Schwartz SL, Fonseca V, Hes M, LaMoreaux L, Garofalo E. Gabapentin for the symptomatic treatment of painful neuropathy in patients with diabetes meUitus: a randomized controlled trial. J Am Med Assoc 1998;280:1831-1836.

Barhanin J, Schmid A, Lazdunski M. Properties of structure and interaction of the receptor for omega-conotoxin, a polypeptide active on Ca " channels. Biochem Biophys Res Commun 1988:150:1051-1062.

Bowersox SS, Gadbois T, Singh T, Pettus M, Wang YX, Luther RR. Selec­tive N-type neuronal voltage-sensitive calcium channel blocker, SNX- 111, produces spinal antinociception in rat models of acute, persistent and neuropathic pain. J Pharmacol Exp Ther 1996;279:1243-1249.

Brose WG, Gutlove DP, Luther RR, Bowersox SS, McGuire D. Use of intrathecal SNX-111, a novel, N-type, voltage-sensitive, calcium chan­nel blocker, in the management o f intractable brachial plexus avulsion pain. Clin J Pain 1997;13:256-259.

Brust PF, Simerson S, McCue AF, Deal CR, Schoonmaker S, Williams ME, Velicelebi G, Johnson EC, Harpold MM, Ellis SB. Human neuronal voltage-dependent calcium channels: studies on subunit structure and role in chamiel assembly. Neuropharmacology 1993;32:1089-1102.

Castillo PE, Weisskopf MG, Nicoll RA. The role of Ca^ channels in hippocampal mossy fiber synaptic transmission and long-term potentia­tion. Neuron 1994;12:261-269.

Chaplan SR, Pogrel JW, Yaksh TL. Role of voltage-dependent calcium channel subtypes in'experirhental tactile allodynia. J Pharmacol Exp Ther 1994;269:1117-1123.

Chaplan SR, Malmberg AB, Yaksh TL. Efficacy of spinal NMDA receptor antagonism in formalin hyperalgesia and nerve injury evoked allodynia in the rat. J Pharmacol Exp Ther 1997;280:829-838.

Chapman V, Suzuki R Dickenson AH. Electrophysiological characteriza­tion. of spinal neuronal response properties in anaesthetized rats after hgation of spinal nerves L5-L6. J Physiol 1998;507:881-894.

Cizkova D, Marsala M, Stauderman K, Yaksh TL. Calcium channel a lB subunit in spinal cord/DRG of normal and nerve-injured rats. Proc. 9th World Congress on Pain, Vienna: lASP Press, 1999, p. 134.

Coderre TJ, Katz J, Vaccarino AL, Melzack R. Contribution of central neuroplasticity to pathological pain: review of clinical and experimental evidence. Pain 1993;52:259-285.

Davies J, Watkins JC. Role o f excitatory amino acid receptors m mono- and polysynaptic excitation in the cat spinal cord. Exp Brain Res 1983;49:280-290.

Diaz A, Dickenson AH. Blockade of spinal N- and P-type, but not L-type, calcium channels inhibits the excitability o f rat dorsal horn neurones produced by subcutaneous formalin inflammation. Pain 1997;69:93- 100 .

Dickenson AH, Sullivan AF. Electrophysiological smdies on the effects of intrathecal morphine on nociceptive neurones in the rat dorsal horn. Pain 1986;24:211-222.

Dickenson AH, Sullivan AF. Evidence for a role of the NMDA receptor in the frequency dependent potentiation of deep rat dorsal hom nocicep­tive neurones following C fibre stimulation. Neuropharmacology 1987;26:1235-1238.

Dickie BG, Davies JA. Calcium channel blocking agents and potassium- stimulated release of glutamate firom cerebellar slices. Eur J Pharmacol 1992;229:97-99.

Dubner R, Bennett GJ. Spinal and trigeminal mechanisms of nociception. Aimu Rev Neurosci 1983;6:381-418.

Eide PK, Jorum E, Stubhaug A, Bremnes J, Breivik H. Relief of post­herpetic neuralgia with the N-methyl-D-aspartic acid receptor antago­nist ketamine: a double-blind, cross-over comparison with morphine and placebo. Pain 1994;58:347-354.

Gohil K, Bell JR, Ramachandran J, Miljanich GP. Neuroanatomical distri­bution of receptors for a novel voltage-sensitive calcium-channel antagoitist, SNX-230 (omega-conopeptide MVnC). Brain Res 1994;653:258-266.

Gracely RH, Lynch SA, Bennett GJ. Painful neuropathy: altered central processing maintained dynamically by peripheral input (published erra­tum appears in Pain 1993;52(2):251-253). Pain 1992;51:175-194.

Gruner W, Silva LR. Omega-conotoxin sensitivity and presynaptic inhibi­tion of glutamatergic sensory neurotransmission in vitro. J Neurosci 1994;14:2800-2808.

Haley JE, Sullivan AF, Dickenson AH. Evidence for spinal N-methyl-D- aspartate receptor involvement in prolonged chemical nociception in the rat. Brain Res 1990;518:218-226.

Hobom M, Dai S, Marais E, Lacinova L, Hofinann F, KJugbauer N. Neuro­nal distribution and functional characterization of the calcium channel « 28-2 subunit. Eur J Neurosci 2000;12:1217-1226.

Hofmann F, Lacinova L, Klugbauer N. Voltage-dependent calcium chan­nels: firom structure to function. Rev Physiol Biochem Pharmacol 1999;139:33-87.

Holz GG, Dunlap K, Kream RM. Characterization of the electrically evoked release of substance P firom dorsal root ganglion neurons: meth­ods and dihydropyridine sensitivity. J Neurosci 1988;8:463-471.

Kajander KC, Wakisaka S, Beimett GJ. Spontaneous discharge originates in the dorsal root ganglion at the onset of a painful peripheral neuropathy in the rat. Neurosci Lett 1992;138:225-228.

Kerr LM, Filloux F, Olivera BM, Jackson H, Wamsley JK. Autoradio­graphic localization of calcium chaimels with [ 1251]omega-conotoxin in rat brain. Eur J Pharmacol 1988;146:181-183.

Kim SH, Chung JM. An experimental model for peripheral neuropathy produced by segmental spinal nerve ligation in the rat. Pain 1992;50i355-363.

Kimura M, Yamanishi Y, Hanada T, Kagaya T, Kuwada M. Watanabe T, Katayama K, Nishizawa Y. Involvement of P-type calcium charmels in high potassium-elicited release of neurotransmitters from rat brain slices. Neuroscience 1995;66:609-615.

Luebke JI, Dunlap K, Turner TJ. Multiple calcium channel types control glutamatergic synaptic transmission in the hippocampus. Neuron 1993;11:895-902.

Luo ZD. Rat dorsal root ganglia express distinctive forms of the 012 calcium chaimel subunit. NeuroReport 2000;11:3449-3452.

Luo ZD, Higuera ES, Stauderman KA, Williams ME, Chaplan SR. Regula­tion o f 028 calcimn channel subunit in dorsal root ganglia and spinal cord o f rats with tactile allodynia. Proc. 9th World Congress on Pain, Vieima: lASP Press, 1999, p. 7.

Maggi CA, Tramontana M, Cecconi R, Santicioli P. Neurochemical evidence for the involvement of N-type calcium channels in transmitter secretion from peripheral endings of sensory nerves in guinea pigs. Neurosci Lett 1990;114:203-206.

Malmberg AB, Yaksh TL. Voltage-sensitive calcium channels in spinal nociceptive processing: blockade o f N- and P-type channels inhibits formalin-induced nociception. J Neurosci 1994;14:4882-4890.

Malmberg AB, Yaksh TL. Effect of continuous intrathecal infusion of omega-conopeptides, N-type calcium-chaimel blockers, on behavior and antinociception in the formalin and hot-plate tests in rats. Pain 1995;60:83-90.

Miljanich GP, Ramachandran J. Antagonists of neiuonal calcium channels: structure, function, and therapeutic implications. Annu Rev Pharmacol Toxicol 1995;35:707-734.

253

246 EA. Matthews, A.H. Dickenson / Pain 92 (2001) 235-246

Mintz IM, Adams ME, Bean BP. P-type calcium channels in rat central and peripheral neurons. Neuron 1992;9:85-95.

Nebe J, Vanegas H, Neugebauer V, Schaible HG. Omega-agatoxin IVA, a P-type calcium chaimel antagonist, reduces nociceptive processing in spinal cord neurons with input from the inflamed but not fiom the normal knee joint - an electrophysiological study in the rat in vivo. Eur J Neurosci 1997;9:2193-2201. ■

Nebe J, Vanegas H, Schaible HG. Spinal application o f omega-conotoxin G VIA, an N-type calcium channel antagonist, attenuates enhancement of dorsal spinal neuronal respoûses caused by intra-articular injection of mustard oü in the rat. Exp Brain Res 1998;120:61-69.

Neugebauer V, Vanegas H, Nebe J, Rumenapp P, Schaible HG. Effects of N- and L-type calcium channel antagonists on the responses of noci­ceptive spinal cord neurons to mechanical stimulation of the normal and the inflamed knee joint J Neurophysiol 1996;76:3740-3749.

Olivera BM, Miljanich GP, Ramachandran J, Adams ME. Calcium channel diversity and neurotransmitter release: the omega-conotoxins and omega-agatoxins. Annu Rev Biochem 1994;63:823-867.

Penn RD, Paice JA. Adverse effects associated with the intrathecal admin­istration of zinconotide. Pain 2000;85:291-296.

Pocock JM, NichoUs DG. A toxin (Aga-GI) from the venom of the spider Agelenopsis aperta inhibits the mammalian presynaptic Ca '*' channel coupled to glutamate exocytosis. Eur J Pharmacol 1992;226:343-350.

Price DD, Mao J, Frenk H, Mayer DJ. The N-methyl-D-aspartate receptor antagonist dextromethorphan selectively reduces temporal summation of second pain in man. Pain 1994;59:165-174.

Rowbotham M, Harden N, Stacey B, Bernstein P, Magnus-MUler L. Gaba­pentin for the treatment of postherpetic neuralgia: a randomized controUed trial. J Am Med Assoc 1998;280:1837-1842.

Santicioli P, Del Bianco E, Tramontana M, Geppetti P, Maggi CA. Release of calcitonin gene-related peptide like immunoreactivity induced by electrical fleld stimulation from rat spinal afferents is mediated by conotoxin-sensitive calcium channels. Neurosci Lett 1992;136:161- 164.

Scroggs RS, Fox AP. Calcium current variation between acutely isolated adult rat dorsal root ganglion neurons of different size. J Physiol 1992a;445:639-658.

Scroggs RS, Fox AP. Multiple Ca " currents elicited by action potential waveforms in acutely isolated adult rat dorsal root gangUon neurons. J Neurosci 1992b;12:1789-1801. - . .

Sheen K, Chung JM. Signs of neuropathic pain depend on signals from injured nerve fibers in a rat model. Brain Res 1993;610:62-68.

Sluka KA. Blockade of calcium channels can prevent the onset of second­ary hyperalgesia and allodynia induced by intradeimal injection of capsaicin in rats. Pain 1997;71:157-164.

Sluka KA. Blockade of N- and P/Q-type calcium channels reduces the secondary heat hyperalgesia induced by acute inflammation. J Pharma­col Exp Ther 1998;287:232-237.

Smith SJ, Augustine GJ. Calcium ions, active zones and synaptic transmit­ter release. Trends Neurosci 1988;11:458-464.

Suzuki R, Gale A, Dickenson AH. Altered effects o f and A, adenosine receptor agonist on the evoked responses of spinal dorsal hom neurones in a rat model of mononeuropathy. J Pain 2000a; 1:99-110.

Suzuki R, Kontinen VK, Matthews E, Williams E, Dickenson AH. Enlar­gement o f the receptive field size to low intensity mechanical stimula­tion in the rat spinal nerve ligation model of neuropathy. Exp Neurol 2000b;163:408-413.

Takahashi T, Momiyama A. Different types of calcium channels mediate central synaptic transmission. Nature 1993;366:156-158.

Taylor CP, Gee NS, Su TZ, Kocsis JD, Welty DF, Brown JP, Dooley DJ, Boden P, Singh L. A summary of mechanistic hypotheses of gabapentin pharmacology. Epilepsy Res 1998;29:233-249.

Terrian DM, Dorman RV, Gannon RL. Characterization o f the presynaptic calcium channels involved in glutamate exocytosis from rat hippocam­pal mossy fiber synaptosomes. Neurosci Lett 1990;119:211-214.

Turner TJ, Adams ME, Dunlap K. Calcium channels coupled to glutamate release identified by omega-Aga-IVA. Science 1992;258:310-313.

Turner TJ, Adams ME, Dunlap K. Multiple Ca '*' channel types coexist to regulate synaptosomal neurotransmitter release. Proc Natl Acad Sci USA 1993;90:9518-9522.

Wall PD, Devor M. Sensory afferent impulses originate from dorsal root ganglia as well as from the periphery in normal and nerve injured rats. Pain 1983;17:321-339.

Westenbroek RE, Hoskins L, Catterall WA. Localization of Ca " channel subtypes on rat spinal motor neurons, intemeurons, and ne;ve terminals.J Neurosci 1998;18:6319-6330.

White DM, Cousins MJ. Effect of subcutaneous administration of calcium channel blockers on nerve injury-induced hyperalgesia. Brain Res 1998;801:50-58.

Xiao WH, Bennett GJ. Synthetic omega-conopeptides applied to the site of nerve injury suppress neuropathic pains in rats. J Pharmacol Exp Ther 1995;274:666-672.

Yamamoto C, Sawada S, Ohno-Shosaku T. Suppression of hippocampal synaptic transmission by the spider toxin omega-agatoxin-IV-A. Brain Res 1994;634:349-352.

Yoon YW, Na HS, Chung JM. Contributions of injured and intact afferents to neuropathic pain in an experimental rat model. Pain 1996;64:27-36.

Zimmerman M. Ethical guidelines for investigations of experimental pain in conscious animals. Pain 1983;16:109-110.

254

ejpELSEVIER European Journal of Pharmacology 415 (2001) 141—149

www.elsevler.nl/locate/ejphar

Effects of ethosuximide, a T-type channel blocker, on dorsal hom neuronal responses in rats

Elizabeth A. Matthews *, Anthony H. DickensonDepartment o f Pharmacology, University College London, Gower Street, London WCIE 6BT, UK

Received 26 October 2000; received in revised form 24 January 2001; accepted 30 January 2001

Abstract

Plasticity in transmission and modulatory systems are implicated in mechanisms of neuropathic pain. Studies demonstrate the importance of high voltage-activated Ca '*' channels in pain transmission, but the role of low voltage-activated, T-type Ca '*' channels in nociception has not been investigated. The Kim and Chung rodent model of neuropathy [Pain 50 (1992) 355] was used to induce mechanical and cold allodynia in the ipsilateral hindpaw. In vivo electrophysiological techniques were used to record the response of dorsal hom neurones to innocuous and noxious electrical and natural (mechanical and thermal) stimuli after spinal nerve ligation. Spinal ethosuximide (5-1055 |xg) exerted dose-related inhibitions of both the electrically and low- and high-intensity mechanical and thermal evoked neuronal responses and its profile remained unaltered after neuropathy. Measures of spinal cord hyperexcitabihty were most susceptible to ethosuximide. This study, for the first time, indicates a possible role for low voltage-activated Ca '*’ channels in sensory transmission. © 2001 Elsevier Science B.V. AH rights reserved.

Keywords: Ca '*' channels, voltage-dependent; Neuropathic pain; T-type channel; Analgesia; Nociception; Dorsal hom

1. Introduction

Neuropathic pain, arising from injury- or disease-evoked damage to the peripheral or central nervous system, often responds poorly to traditional analgesics. Patients often experience sensory deficits, persistent and stimulus-evoked pain (allodynia and hyperalgesia). Peripheral nerve damage provides abnormal input into the central nervous system and then leads to dorsal hom hyperexcitabihty. Under the influence of both excitatory and inhibitory neurotrans­mitter systems, the dorsal hom is a site of peripheral input modulation before projection to higher brain centres, thus it controls the stimulus-response relationship. Reduction of this excitabihty is a possible key to neuropathic pain management.- Animal models of neuropathy have been critical in elucidating its complex causal mechanisms, involving plas­

* Coiresponding author. Tel.; 4-44-20-7679-3737; fax: 444-20-7679- 7298.

E-mail address: e.matthews@ucl.ac.uk (E.A. Matthews).

ticity in nociceptive transmission and modulating systems. The rat spinal nerve hgation model (Kim and Chung, 1992) involves tight hgation of two (L5 and L6) of the three spinal nerves that form the sciatic nerve. Behavioural consequences include thermal hyperalgesia, and mechani­cal and coohng allodynia (Kim and Chung, 1992; Chaplan et al., 1997).

High voltage-activated Ca " channels (L-, N-, P /Q - and R-types), consisting of a pore-forming a 1 subunit and modulatory accessory subiinits, |3, a j-S and 7 (Walker and De Waard, 1998), are widely expressed throughout the brain and spinal cord (Kerr et al., 1988; Mintz et al., 1992; Gohil et al., 1994). They are activated by relatively strong membrane depolarisation and permit Ca '*' influx in re­sponse to action potentials. Consequential secondary ac­tions include neurotransmitter release; thus these channels estabhsh a major link between neuronal excitabihty and synaptic transmission. For these reasons high voltage- activated Ca "*" channels have been the focus of both acute and persistent pain transmission studies. Animal models have demonstrated the antinociceptive abhities of antago­nists specific for L-, N- and P/Q -type Ca "*" channels, highhghting the differential role each subtype plays in

0014-2999/01 /$ - see front matter ©2001 Elsevier Science B.V. All rights reserved. PE: 8 0 0 1 4 -2 9 9 9 ( 0 1 )0 0 8 1 2 -3

255

142 EA. Matthews, A.H. Dickenson/European Journal o f Pharmacology 415 (2001) 141—149

nociception, often dependent on the nature of the pain state (Vanegas and Schaible, 2000).

In addition to high voltage-activated Ca '*' channels, kineticaHy distinct low voltage-activated Ca '*’ channels, or T-type channels, also exist both in neuronal and non-neu­ronal cells. They activate at voltages near the resting membrane potential, inactivate rapidly, deactivate slowly and have a small single channel conductance (Huguenard, 1996). These unique gating properties prohibit T-type channels alone to support neurotransmission, however they permit their involvement in low-amplitude oscillations, neuronal bursting, synaptic signal boosting, Ca " entry promotion and lowering threshold for high-threshold spike generation. This Ca '*’ current appears to play an important physiological role in near-threshold phenomena and regu­lation of neuronal excitabihty.

Until recently, the low voltage-activated current was considered a single entity. However, a lG , a lH and a l l subunits have now been cloned, showing 30% homology to high voltage-activated charmel forming a l subunits (Cribbs et al., 1998; Perez-Reyes, 1998; Lee et al., 1999) and hallmark native T-type Ca "*" charmel properties when expressed heterologously. In situ hybridisation studies on the rat brain have shown that these charmels have unique distributions, including the dorsal hom of the spinal cord and sensory gangha (TaUey et al., 1999). This is comph- mented by reported T-type currents in primary sensory neurones (Carbone and Lux, 1984; Kostyuk et al., 1992; Scroggs and Fox, 1992; Todorovic and Lingle, 1998) and some superficial rat dorsal hom neurones (Ryu and Randic, 1990), an important site for the processing and integration of sensory information, including pain. Unlike the high voltage-activated Ca " chmmels, thejmyqlyernent of T-type channels in pain-related central sensitisation has been hin­dered by a scarcity of specific pharmacological agents.

Neuropathic pain and epilepsy both share neuronal hy- perexcitabUity as a common underlying mechartism. There are established antiepüeptic dmgs that target the genera­tion of neuronal hyperexcitabihty in the brain and some of these have been proven effective in the treatment of vari­ous forms of neuropathic pain (Swerdlow and Cundhl, 1981; McQuay et al., 1995). The succinimde derivative ethosuximide, or 2-ethyl-2-methylsuccirurnide, is an anti­convulsant (Macdonald and McLean, 1986) effective in the treatment of absence epilepsy (Coulter et al., l989b); a condition characterised by spike-wave rhythm likely gener­ated by T-type Ca "*" current. Ethosuximide has been demonstrated to be a relatively specific T-type charmel antagonist in thalamic (Coulter et al., 1989a) and dorsal root ganghon neurones (Kostyuk et al., 1992). This study uses the spinal nerve hgation model, confirmed by be­havioural testing, to induce a neuropathic state, subsequent to which electrophysiological studies of dorsal hom spinal neurones were made to investigate the effects of spinaUy dehvered ethosuximide on a wide range of electrical and natural-evoked neuronal activity.

2. Materials and methods

2.1. Spinal nerve ligation

Male Sprague-Dawley rats, initiahy weighing 130-150 g, were used in this study. All experimental procedures were approved by the Home Office and foUow the guide­lines under the Intemational Association for the study of Pain (Zimmermaim, 1983). Selective tight hgation of spinal nerves L5 and L6, and a sham procedure were performed as first described by Kim and Chung (1992). For details, see Chapman et al. (1998).

2.2. Behavioural testing

For 2 weeks foUowing surgery, the rats were housed in groups of 4, in plastic cages under a 12/12 h day/night cycle and their general health monitored. Successful repro­duction of the neuropathic model was confirmed by be­havioural testing (post-operative days 2, 3, 5, 7, 9, 12 and 14) assessing the sensitivity of both the ipsilateral and contralateral hindpaws to normaUy non-noxious punctate mechanical (von Frey filaments) and coohng (acetone) stimuli. Rats were placed in transparent plastic cubicles on a mesh floor and aUowed to acclimatise before initiating any tests. Foot withdrawals to von Frey filaments 1, 5 and 9 g (trials of 10) and acetone (trials of 5) were quantified as described in Chapman et al. (1998) and expressed as Difference Scores = Ipsilateral response — contralateral re­sponse.

2.3. Spinal cord electrophysiology

Subsequent to behavioural testing (post-operative days 14-17), the operated rats were used for electrophysiologi­cal studies (Dickenson and SuUivan, 1986). Briefly, anaes­thesia was induced with 3% halothane in a mixture of 66% NjO and 33% Oj and a cannula inserted into the trachea. A laminectomy was performed (vertebrae L1-L3) to ex­pose segments L4-L5 of the spinal cord and the level of halothane was reduced to 1.8%. ExtraceUular recordings of single convergent neurones, located deep within the dorsal hom ( > 500 |xm), receiving input fi-om the toe region ipsilateral to the spinal nerve hgation or sham procedure, were made using a parylene coated tungsten electrode. Neurones selected responded to both noxious (pinch) and non-noxious (touch) stimuh.

2.3.1. Cell characterisationSpontaneous activity exhibited by a neurone was

recorded over 10 min. Action potentials evoked by natural stimuh apphed constantly over 10 s were quantified by the apphcation of both punctate mechanical (von Frey fila-

256

EA. Matthews, A.H. Dickenson/European Journal of Pharmacology 415 (2001) 141-149 143

meats 9 and 75 g) and thermal (constant water jet at 45°C) stimuli applied to the centre of the neurone’s receptive field. The thermal response to 45°C was determined by subtracting the response to 32°C (a non-noxious tempera­ture so as to ascertain any mechanical response evoked by the water jet) from the response to 45°C. All responses to natural stimuli were normahsed by the subtraction of any spontaneous activity measured before the apphcation of each stimulus. Response of the neurone to transcutaneous electrical stimulation was estabhshed by insertion of two fine needles into the centre of its peripheral receptive field. A test consisted of a train of 16 stimuh (2 ms wide pulse at 0.5 Hz at three times the threshold required to evoke a C-fibre response), and a post-stimulus histogram was con­structed. The thresholds were determined by increasing the electrical stimulus from 0 mA until an action potential was evoked in the corresponding latency band. Electrically evoked action potentials were separated on a latency basis into A 3-fibres (0-20 ms), A 5-fibres (20-90 ms), C-fibres (90-300 ms) and post-discharge (300-800 ms). The ‘in­put’ is the number of action potentials (90-800 ms) evoked by the first stimulus of the train. ‘Excess spikes’ is a measure of ‘wind-up’, which is increased neuronal ex­citabihty to repeated constant stimulation. Excess spikes was calculated as the total action potentials (90-800 ms) after 16 stimulus train minus the input X 16. Wind-up graphs for individual neurones show how the combined number of evoked C-fibre and post-discharge action poten­tials (i.e., 90-800 ms) increases with each repeated electri­cal stimulation.

2.3.2. Pharmacological studiesThe testing protocol, initiated every 10 min, consisted

of an electrical test fohowed by the natural stimuh, as described above. Stabhisation of the neuronal responses was confirmed with at least three consistent pre-drug responses (< 10% variation), for all measures. These val­ues were then averaged to generate pre-drug control values with which to compare the effect of ethosuximide (Sigma- Aldrich, Poole, Dorset, UK) administration on subsequent evoked responses. Ethosuximide was dissolved in saline and apphed directly onto the spinal cord in 50 p.1 volumes. Each dose (5, 55, 555, 1055 p.g) was followed until maximum effects were exerted (a minimum of 50 min), when the next dose would be apphed cumulatively. The results were calculated as maximum percentage inhibition from the averaged pre-drug value for each neurone and the overall results for each dose were expressed as means ± standard error of mean (S.E.M.) of the normahsed data. Statistical analysis of maximal drug effects at each dose compared to the averaged pre-drug value was determined by paired t-test on raw data. An unpaired r-test on the normahsed data was used for the comparison of drug effects between different experimental groups. The level of significance was taken as P < 0.05.

3. Results

3.2. Behavioural studies

During the post-operative period the animals showed normal weight gain and maintained good general health. Rats subjected to spinal nerve hgation exhibited abnormal foot posture ipsilateral to nerve injury whereby toes were held together in a ‘guarding’ behaviour. This did not occur in either the contralateral hindpaw, or in the sham-operated rats. Successful rephcation of the nerve injury model was confirmed by behavioural testing which demonstrated the development of mechanical and coohng ahodynia of the injured hindpaw of sphial nerve hgated rats. Evoked allo­dynia, in response to innocuous mechanical (von Frey filaments bending force 1-9 g) and coohng (acetone) stimuh, was displayed as a brisk withdrawal, accompanied in some cases by shaking and hcking of the foot ipsilateral to spinal nerve hgation. This was evident at post-operative day 2, reached maximum at days 7-12 and still maintained at day 14 (Fig. 1). Consistent withdrawal responses were never exhibited by the control group or by the contralateral hindpaw of the experimental group, and when present were never accomparhed by the pain-like behaviours displayed by the lesioned hindpaw of spinal nerve hgated rats.

3.2. Spinal cord electrophysiology

3.2.1. Cell characterisationThe numbers of ipsilateral dorsal hom neurones charac­

terised in each group were 11 in spinal nerve hgated rats, 6

7

6

a

Q 21

0

Post operative Day

von Frey Ig

von Frey 5g

von Frey 9g

A cetone

Fig. 1. Development of mechanical and cooling allodynia in the ipsilat­eral hindpaw over the 2-week period following spinal nerve ligation. Data is presented as the mean difference score ± S.E.M. (n = 11) in the with­drawal response to punctate mechanical stimuli (von Frey filaments) and coohng stimulus (drop of acetone) apphed to the plantar surface of the hindpaws (trials of 10 for the mechanical and 5 for the coohng). Difference score = (ipsilateral response frequency) — (contralateral re­sponse frequency). No nociceptive behaviour was observed in the sham- operated rats (data not shown).

257

144 EA. Matthews, A.H. Dickenson/European Journal of Pharmacology 415 (2001) 141-149

in sham-operated rats and 8 in naïve rats. All neurones had a receptive field over the left ipsilateral hindpaw. No significant differences were found between experimental groups in the mean values of recorded neurone depth, responses evoked by electrical and natural stimulation, and level of ongoing spontaneous activity. It is worth noting that 45% of neurones characterised in spinal nerve Ligated rats exhibited spontaneous activity at a rate greater than 0.1 Hz in comparison to only 17% of characterised neu­rones in sham-operated rats and 25% of naive.

3.2.2. Pharm acological studiesThe effect of ethosuximide (5-1055 |ag), applied di­

rectly onto the spinal cord, on the electrically and naturally evoked dorsal hom neuronal responses was tested in spinal nerve ligated, sham-operated and naïve animals. Since the sham operation is the appropriate control for spinal nerve ligation, and for clarity, the results obtained from the naïve, non-operated, group shall not be displayed on the graphs. However, no difference in the effects of ethosux­imide was observed between sham and naïve groups.

Ethosuximide produced a dose-related inhibition of the electrically and naturally evoked responses in neurones in all experimental groups (Figs. 2, 3 and 4) and there was no difference in the extent of its effects between groups. Clear effects were seen around 40 min, with maximal inhibitions established at around 60 min. For sham and spinal nerve ligated groups, all doses of ethosuximide (5-1055 |xg) elicited statistically significant inhibitions of the electri­cally evoked responses, compared to pre-drug control (Fig. 2; F < 0.05; n = 6-11). For naïve animals, statistically significant inhibitions were ehcited by ethosuximide: at all doses (5-1055 |xg) for the A (3-fibre, A 6-fibre, post-dis- charge and excess spikes measurements; at 55-1055 |xg ethosuximide for the C-fibre response; and at 555 and 1055 |xg for the input response ( F < 0.05; n = 7). In all three groups, the A (3-fibre response was least affected with mean maximal inhibitions at top dose ranging from 17 ± 6% to 26 ± 8% (Fig. 2F). In aU three groups, the C-fibre and A 6-fibre responses reached similar mean maximal inhibitions at top dose ranging from 33 + 13% to 47 ± 17% (Fig. 2D and E, respectively). Greater inhibitory effects

(A )Input100 100 -,

60 - 60 -

40 40 -

20 - 20 -

01 10 100 1000 10000

(B) Post-discharge100 -

60 -

40 -

20 -

100 —I-------11000 10000

0 J

(C) Excess spikes

i

100 1000 10000

100-

80 -

60 -a■3 40 -

20 -

0 -

(D) c-fibre

10 100

1 0 0 -

80 -

60 -

40 -

20 -

0 -

1000 10000

(E) A5-fibre (F) AP-fibre100-1

80 - -Sham

■Spinal Nerve Ligated

60 -

4 0 -

20 -

01000 10000100 1000 10000 1 10 100

Ethosuximide (gg) Ethosuximide (gg) Ethosuximide (gg)

Fig. 2. Effect of spinally applied ethosuximide on the electrically evoked dorsal hom neuronal responses (see Materials and methods) recorded from spinal nerve ligated in = 6 -11) and sham-operated (n = 6) rats at post-operative days 14-17. Data is expressed as maximal mean % inhibition of the pre-drug values + S.E.M.

258

EA. Matthews, A.H. Dickenson / European Journal o f Pharmacology 415 (2001) 141-149 145

100-

80 -

= 60 -

a,o

20 -

0 J

A) Spinal Nerve Ligated

1 I I I I 8 10 12 14 162 4 6

Stimulus Number

60 -

4 0-

20 H

0 J

B) Sham

- i ------- 1-----r2 4 6

“I 1----- 1------1------18 10 12 14 16

-O— Pre-drug Control

- k — Ethosuximide 5 |4g

-O— Ethosuximide 55 p.g

-V— Ethosuximide 5555 |i.g

-* — Ethosuximide 1055 p,g

Stimulus Number

Fig. 3. Examples of the inhibitory effect of spinally applied ethosuximide on individual neurones exhibiting wind-up recorded from (A) spinal nerve ligated and (B) sham-operated rats.

over the entire ethosuximide concentration range were observed on the input, post-discharge and excess spikes measurements (Fig. 2A, B and C, respectively). At 1055 fjtg, ethosuximide maximally inhibited the input and excess spikes to within the range 48 ± 8% to 61 ± 14%. The greatest effect was observed on post-discharge, which was maximally inhibited ih spinal nerve ligated, sham and naïve rats ranging from 58 ± 10% to 75 ± 13%.

Fig. 3A and B shows examples of the effects of etho­suximide on the wind-up of an individual neurone recorded from a spinal nerve hgated and sham rat, respectively. It is clear that ethosuximide produces a dose-dependent inhibi­tion from the control response of both the wind-up (evi­dent by a flattening of_the cmve) andjo a lesser extent, the input, with httle difference seen between experimental animal groups.

Ethosuximide also produced an inhibitory effect on the naturally evoked neuronal responses (Fig. 4). In spinal nerve ligated animals this was significant (P < 0.05; n = 6- 11) for ah concentrations of ethosuximide employed in this study on the response to non-noxious mechanical stimulation (von Frey 9g, Fig. 4A), noxious mechanical stimulation (von Frey 75g, Fig. 4B) and noxious thermal stimulation (water jet at 45°C, Fig. 4C). In sham rats the von Frey 9g and heat responses were significantly inhib­ited at all concentrations of ethosuximide, and the von Frey 75g response was inhibited by 555 and 1055 |xg (P ^ 0.05; n = 5-6). In naïve animals responses to both von Frey hairs were significantly inhibited by aU concen­trations of ethosuximide, and inhibition of the heat re­sponse reached significance at 555 and 1055 p.g (P < 0.05; n = 5-7). No differences in the effects of the drug were

(A) von Frey 9glo o ­

se -

140 - T20 - I

0 -1

100

100 -

80

60

4 0 -

2 0 -

0 J

(B) von Frey 75g

Sham

1000 10000

Spinal Nerve Ligated

10

(C) Heat100-

80 -

60 -

40 -T

20 - Ii

0 -100 1000 10000

—I------ 1------- 1100 lOOO 10000

Ethosuximide (pg) Ethosuximide (pg)Ethosuximide (pg)

Fig. 4. Effect of spinally applied ethosuximide on the naturally evoked dorsal hom neuronal responses (see Materials and methods) recorded from spinal nerve ligated (n = 6 -11) and sham-operated rats (n = 4 -6). (A) Innocuous punctate mechanical stimulus (von Frey 9g), (B) noxious pimctate mechanical stimulus (von Frey 75g) and (C) noxious heat stimulus (water jet at 45°C normalised by response to water jet at 32°C). Data is expressed as maximal mean % inhibition of the pre-dmg values ± S.E.M.

259

146 EA. Matthews, A H . Dickenson / European Journal o f Pharmacology 415 (2001) 141-149

apparent between experimental groups and the mean maxi­mal inhibitions established at top dose were in the range 61 ± 11% to 77 ± 10%.

4. Discussion

Unilateral tight ligation of L5 and L6 spinal nerves produced reproducible nociceptive syndromes in the le­sioned hindpaw. A clear withdrawal reflex with associated aversive behaviours, indicative of the development of me­chanical and cooling allodynia, was produced as previ­ously described by Chapman et al. (1998). Thus, all spinal nerve hgated animals used for the electrophysiology and subsequent pharmacology exhibited neuropathic signs; sham-operated did not. This is the first electrophysiologi­cal study addressing the role of low voltage-activated T-type Ca " channels in the spinal processing of sensory information after nerve injury. Spinal ethosuximide, a rela­tively specific T-type channel antagonist, mediated signifi­cant inhibition of the electrical and natural (innocuous and noxious) evoked rat dorsal hom neuronal responses, sug­gesting some role for T-type Ca " channels in sensory transmission.

Extensive behavioural and electrophysiological nocicep­tive studies have demonstrated an important role for high voltage-activated Ca '*’ channels in the processing of pain (Vanegas and Schaible, 2000). The contribution of N-, P /Q - or L-type channels appears to alter depending on the nature of the pain (acute or chronic, inflammatory or neuropathic in origin), and this has been estabhshed by the use of specific channel antagonists. A predominant noci­ceptive- role for N-type^channelS:Jias been^estabhshed, which is enhanced after neuropathy (Chaplan et al., 1994; Xiao and Bennett, 1995; Bowersox et al., 1996; Brose et al., 1997; White and Cousins, 1998; Matthews and Dicken­son, 2000). Upon substantial membrane depolarisation N- and P-type voltage-activated Ca '*' channels mediate the release' of excitatory neurotransmitters, such as glutamate, substance P and calcitonin gene-related peptide, critical for wind-up and central sensitisation, in the presence of con­stant afferent input (Dickenson, 1994).

This study highhghts the role of Ca '*’ influx via T-type channels in the nociception pathway, within which etho­suximide may be exerting its effects at a number of sites. The biophysical characteristics of the T-type channel have led to its implication mainly in the regulation of cell excitabihty. The broad action of ethosuximide, with no marked selectivity for a particular modahty or evoked response, is in keeping with an action on postsynapticaHy located Ca^^ channels. Since T-type channel activation occurs close to resting potential, they allow Ca '*' influx when cells are at rest (Magee and Johnston, 1995) or in response to subthreshold synaptic inputs (Markram and' Sakmann, 1994; Magee and Johnston, 1995). Thus, these channels enhance neuronal excitability and contribute to

the generation of subthreshold membrane potential oscilla­tions that lead to bursts of sodium dependent action poten­tials (Huguenard, 1996). Whilst unable to mediate synaptic transmission alone, T-type channels do serve to boost synaptic inputs and lower threshold for high-threshold spike generation. Their block would result in an overall reduction in the underlying level of neuronal excitability, rendering the achievement of threshold levels of mem­brane depolarisation less likely. PostsynapticaHy, NMDA receptor activation would be reduced as would the conse­quential development of central sensitisation.

In this study, ethosuximide exerted its greatest in­hibitory effects upon the input, post-discharge and excess spikes. Each of these electrical measures can be related to a specific part of the nociceptive pathway. The non-poten­tiated input response can be related to the level of synaptic transmission between the central terminals of primary af­ferents and the neurones of the spinal cord dorsal hom. Thus, blockade of T-type channels at this location is likely causing a reduction in the exocytosis of excitatory neuro­transmitters by prohibiting the depolarisation required to activate high voltage-activated Ca^^ channels. Alterna­tively, ethosuxhnide could be exerting its effects directly on neurones located early in polysynaptic pathways. Even more susceptible to the actions of ethosuximide were the postsynaptic NMDA receptor-mediated post-discharge and excess spike measurements. These are indicative of central sensitisation and neuronal hyperexcitabihty, and since T- type Ca " channels are heavily linked to the level of neuronal excitabihty it follows that they would have a greater functional role here.

What is surprising is that we saw no difference in the effects-of ethosuximide after the estabhshment of neuropa­thy, especially since neuronal hyperexcitabihty is a key underlying factor. Interestingly it has been shown that although unilateral cortical ablation causes a 68% increase in T-current measured from isolated rat thalamic relay neurones, a-methyl-u-phenylsuccimide (another related T- type charmel antagonist) was more effective in reducing T-current in normal rats compared to axotorrtised animals (Chung et al., 1993). The authors suggest an injury in­duced alteration in the pharmacological properties the T- type charmels either by de novo synthesis and /o r modifi­cation. In this study, it may be that there is indeed no increase in the functional role of T-type Ca^^ channels after nerve injury. Alternatively, the specificity and/or potency of ethosuximide may be such that subtle differ­ences in T-type channel function were not highlighted.

Peripheral nerve injury results in reduced afferent input via L5 and L6 spinal nerves, yet as observed here, the magnitude of neuronal responses recorded was not dimin­ished in comparison to sham and normal rats. Conversely, increased frequency and occurrence of spontaneous activ­ity was observed. This suggests that perhaps compensatory increases in peripheral and /o r spinal neuronal activity are in play after neuropathy. Ectopic C-fibre activity originat­

260

E.A. Matthews, A.H. Dickenson/European Journal o f Pharmacology 415 (2001) 141-149 147

ing within the dorsal root ganglion, the nerve injury site, or residual intact afferents provides an ongoing, continual barrage of neuronal activity via primary afferents into the spinal cord (Wall and Devor, 1983; Kajander et al., 1992). Evidence suggests that this is one possible source for the initiation and maintenance of central sensitisation and therefore the positive sensory symptoms observed after nerve damage (Devor and Seltzer, 1999). Interestingly, spinal nerve hgation has been shown to increase the prevalence of subthreshold membrane potential oscillations in dorsal root ganghon neurones which augments ectopic discharge (Liu et al., 2000). It does not seem unreasonable to postulate a role for T-type Ca^^ channels in this under­lying oschlatory behaviour, .which may imply increased activity via T-type channels in the remaining afferents tantamount to a restoration of excitabihty to that in the non-injured simation.

The existence of a neuronal Ca """ current ehcited just above the resting potential was first estabhshed in primary sensory neurones (Carbone and Lux, 1984; Bossu et al., 1985; Fedulova et al., 1985; Nowycky et al., 1985), and low voltage-activated current has since been observed in a wide variety of ceU types (Huguenard, 1996). The pres­ence of a relatively large T-type current in some superfi- ciahy located rat spinal dorsal hom neurones is of much interest because this region is involved in processing and integration of sensory information, including pain (Ryu and Randic, 1990). T-type channels are pharmacologically and physiologicaUy heterogeneous (Akaike, 1991; Hugue­nard, 1996; Tarasenko et al., 1997), which may reflect differential expression of the three known subtypes (a lG , H and I). The regional and cehular distribution of gene expression for the different T-type Ca " channel family members in the rat central and peripheral nervous systems has recently been determined using in situ hybridisation (Tahey et al., 1999). Ah three transcripts were detected in sensory areas and in the dorsal hom of the spinal cord where in particular a lH was mainly restricted to the outermost laminae I and n. In the dorsal root ganglion high levels of a lH and moderate levels of a l l mRNA were found restricted to small and medium sized neurones, whereas the extremely large were not labelled. This corre­lates with substantial T-type current observed in medium- diameter dorsal root ganglion neurones isolated from adult rats that is absent in larger dorsal root ganglion cells (Scroggs and Fox, 1992). Since dorsal root ganghon cell body diameter is correlated to axon conduction velocity and sensory modahty (Yaksh and Hammond, 1982), this evidence is indicative of T-type current specificaUy lo- cahsed to smaUer A 8- and C-type sensory neurones that convey thermal and nociceptive information and not to larger A P-type neurones that subserve tactile and proprio­ceptive pathways. In the present study the extent of inhibi- tion observed with the highest dose of ethosuximide was A 8-fibre > C-fibre > Ap-fibre, which fits well with these studies. However, the A 8- and C-fibre responses were not

markedly inhibited over the Ap-fibre response, as one might expect if smaher diameter sensory neurones exhib­ited substantial low voltage-activated Ca " current Also no difference in the extent of inhibition was observed for the evoked responses to innocuous and noxious mechani­cal and thermal stimuh. This may again be explained by the existence of the three à 1 subunits, each with a unique distribution, encoding for T-type Ca "* channels. Data suggests that diversity exists between T-currents of differ­ent ceh types both in terms of kinetics and pharmacologi­cal sensitivity (Huguenard, 1996). For example, the current clinical use of ethosuximide to treat epilepsy is via block of T-current in thalamic neurones (Coulter et al., 1989a,b). However, T-current in GH3 cells is relatively resistant to block by ethosuximide (Herrington and Lingle, 1992) and in dorsal root ganghon neurones ethosuximide is over an order of magnitude less effective (Coulter et al., 1989a,b). Furthermore, the blockade is complete in dorsal root gan­ghon but only p ^ a l (40%) in thalamic neurones, a lG is the predominant subtype found in thalamic relay neurones, and therefore may be more sensitive to the effects of ethosuximide in comparison to the a lH T-type charmel which is more abundant than a IG in the outer lamina of the spinal cord and dorsal root gangha (Tahey et al.,1999).

Two other hcensed anticonvulsants, carbamazepine and gabapentin, have been investigated using the same experi­mental protocol as the present study (Chapman et al., 1998). Carbamazepine, a sodium channel blocker, and gabapentin, thought to act via Ca " charmels (see Matthews and Dickenson, 2000) were found to have similar efficacy and range of effectiveness as ethosuximide. Although both gabapentin and ethosuximide were equahy effective in sham and neuropathic animals, carbamazepine was only effective in the latter group and this could possibly be a consequence of differential regulation of Ca '*’ and sodium charmels fohowing nerve injury. To our knowledge the present smdy is the first to demonstrate a possible role of T-type Ca '*’ charmels in the spinal processing of sensory information related to pain. Given the parallels between epilepsy and pain, the hkehhood of common causal mech­anisms and the ability of antiepileptic drugs to be effective in neuropathic pain states, the results indicate that ethosux­imide may merit both behavioural testing in animals and human studies.

Acknowledgements

E.M. is supported by the Wellcome Trust 4-year Neuro­science PhD programme.

References

Akaike, N., 1991. T-type caldum channel in mammalian CNS neurones. Comp. Biochem. Physiol. A. Physiol. 98, 31-40.

261

148 EA. Matthews, A.H. Dickenson/European Journal o f Pharmacology 415 (2001) 141-149

Bossu, J.L., Feltz, A., Thomaim, JM ., 1985. Depolaiization elicits two distinct calcium currents in vertebrate sensory neurones. Pflugers Arch.— Eur. J. Physiol. 403, 360-368.

Bowersox, S.S., Gadbois, T., Singh, T., Pettus, M., Wang, Y.X., Luther, R.R., 1996. Selective N-type neuronal voltage-sensitive calcium chan­nel blocker, SNX-111, produces spinal antinociception in rat models of acute, persistent and neuropathic pain. J. Pharmacol. Exp. Ther. 279, 1243-1249.

Brose, W.G., Gutlove, D.P., Luther, R.R., Bowersox, S.S., McGuire, D.,1997. Use of intrathecal SNX-111, a novel, N-type, voltage-sensitive, calcium channel blocker, in the management of intractable brachial plexus avulsion pain. Clin. J. Pain 13, 256-259.

C^bone, E., Lux, H D ., 1984. A low voltage-activated, fully inactivating Ca channel in vertebrate sensory neurones. Nature 310, 501-502.

Chaplan, S.R., Pogrel, J.W., Yaksh, TD., 1994. Role of voltage-depen­dent calcium channel subtypes in experimental tactile allodynia. J. Pharmacol. Exp. Ther. 269, 1117-1123.

Chaplan, S R., Malmberg, A.B., Yaksh, TD., 1997. Efficacy of spinal . NMDA receptor antagonism in formalin hyperalgesia and nerve in­

jury evoked allodynia in the rat. J. Pharmacol. Exp. Ther. 280, 829-838.

Chapman, V., Suzuki, R., Chamarette, H L., Rygh, L.J., Dickenson, A.H.,1998. Effects o f systemic carbamazepine and gabapentin on spinal neuronal responses in spinal nerve ligated rats. Pain 75, 261-272.

Chung, JM., Huguenard, J.R., Prince, D.A., 1993. Trmsient enhance­ment of low-threshold calcium current in thalamic relay neurons after corticectomy. J. Neurophysiol. 70, 20-27.

Coulter, D.A., Huguenard, J.R., Prince, D.A., 1989a. Characterization of ethosuximide reduction of Ibw-threshold calcium current in thalamic neurons. Arm. Neurol. 25, 582-593.

Coulter, D.A., Huguenard, J.R., Prince, D.A., 1989b. Specific petit mal anticonvulsants reduce calcium currents in thalamic neurons. Neu-

. rosci. Lett. 98, 74-78.Cribbs, L.L., Lee, J.H., Yang, J., Satin, J., Zhang, Y., Daud, A., Barclay,

J., Williamson, M.P., Fox, M., Rees, M., Perez-Reyes, E., 1998. Cloning and characterization of alphalH fi'om human heart, a member of the T-type Ca " charmel gene family. Circ. Res. 83, 103-109.

Devor, M., Seltzer, Z., 1999. Pathophysiology of damaged nerves in relation to chronic pain. In: Wall, P.D., Melzack, R. (Eds.), Textbook of Paim'Churchin L i'^p ïoÇ 'L ond5n rp pri29-164“

Dickenson, A.H., 1994. NMDA receptor antagonists as analgesics. In: Fields, H.L., Liebeskind, J.C. (Eds.), Pharmacological Approaches to the Treatment of Chronic Pain: New Concepts and Critical Issues. lASP, Seattle, pp. 173-187.

Dickenson, A.H., Sullivan, A.F., 1986. Electrophysiological studies on the effects of intrathecal morphine on nociceptive neurones in the rat

' dorsal hom. Pain 24, 211-222.Fedulova, S.A., Kostyuk, P.G., Veselovsky, N.S., 1985. Two types of

. calcium charmels in the somatic, membrane of new-born rat dorsal root ganglion neurones. J. Physiol. 359, 431-446.

Gohil, K., Bell, J R., Ramachandran, J., Miljanich, G.P., 1994. Neu­roanatomical distribution of receptors for a novel voltage-sensitive calcium-chaimel antagonist, SNX-230 (omega-conopeptide MVUC). Brain Res. 653, 258-266.

Herrington, J., Lingle, C.J., 1992. Kinetic and pharmacological properties of low voltage-activated Ca "*" current in rat clonal (GH3) pituitary cells. J. Neurophysiol. 68, 213-232.

'-Huguenard, J.R., 1996. Low-thresholdxalcium currents in central nervous system neurons. Annu. Rev. PnySiol. 58, 329-348.

Kajander, KC., Wakisaka, S., Beimett, G.J., 1992. Spontaneous dis­charge originates in the dorsal root ganglion at the onset of a painful peripheral neuropathy in the rat. Neurosci. Lett. 138, 225-228.

Kerr, L.M., Filloux, F., Olivera, BM ., Jackson, H., Wamsley, J.K , 1988. Autoradiographic localization of calcium channels with [125I]omega- conotoxin in rat brain. Eur. J. Pharmacol. 146, 181-183.

Kim, SJI., Chung, J.M., 1992. An experimental model for peripheral neuropathy produced by segmental spinal nerve ligation in the rat. Pain 50, 355-363.

Kostyuk, P.G., Molokanova, E.A., Pronchuk, N.F., Savchenko, AJ4., Verkhratsky, AJ4., 1992. Different action of ethosuximide on low- and high-threshold calcium currents in rat sensory nemrons. Neuro­science 51, 755-758.

Lee, J.H., Daud, A.N., Cribbs, L.L., Lacerda, A.E., Pereverzev, A., Klockner, U., Schneider, T., Perez-Reyes, E., 1999. Cloning and expression of a novel member of the low voltage-activated T-type calcium charmel family. J. Neurosci. 19, 1912-1921.

Liu, C.-N., Michaelis, M., Amir, R., Devor, M., 2000. Spinal nerve injury enhances subthreshold membrane potential oscillations in DRG neu­rons: relation to neuropathic pain. J. Neurophysiol. 84, 205-215.

Macdonald, R.L., McLean, M J., 1986. Anticonvulsant drugs: mecha­nisms of action. Adv. Neurol. 44, 713-736.

Magee, J.C., Johnston, D., 1995. Synaptic activation of voltage-gated channels in the dendrites of hippocampal pyramidal neurons [see comments]. Science 268, 301-304.

Markram, H., Sakmann, B., 1994. Calcium transients in dendrites of neocordcal neurons evoked by single subtbreshold excitatory post­synaptic potentials via low voltage-activated calcium charmels. Proc. Natl. Acad. Sci. U. S. A 91, 5207-5211.

Matthews, E.A., Dickenson, A.H., 2001. Effects of spinally delivered N- and P-type voltage-dependent calcium charmel antagonists on dorsal hom neuronal responses in a rat model of neuropathy. Pain, in press.

McQuay, H., Carroll, D., Jadad, A.R., Wiffen, P., Moore, A., 1995. Anticonvulsant drugs for management o f pain: a systematic review [see comments]. Br. Med. J. 311, 1047-1052.

Mintz, I.M., Adams, M.E., Bean, B.P., 1992. P-type calcium charmels in rat central and peripheral neurons. Neinon 9, 85-95.

Nowycky, M.C., Fox, A.P., Tsien, R.W., 1985. Three types o f neuronal calcium charmel with different calcium agonist sensitivity. Nature 316, 440-443.

Perez-Reyes, E., 1998. Molecular characterization o f a novel family of low voltage-activated, T-type, calcimn channels. J. Bioenerg. Biomembr. 30, 313-318.

Ryu, P.D., Randic, M., 1990. Low and high voltage-activated calcium currents in rat spinal dorsal hom neurons. J. Neurophysiol. 63,

“'273-285. ■ ■Scroggs, R.S., Fox, A.P., 1992. Calcium current variation between acutely

isolated adult rat dorsal root ganglion neurons o f different size. J. Physiol. 445, 639-658.

Swerdlow, M., Cundill, J.G., 1981. Anticonvrrlsant drags used in the treatment of lancinating pain. A comparison. Anaesthesia 36, 1129- 1132.

Talley, E.M., Cribbs, L.L., Lee, J.H., Daud, A., Perez-Reyes, E., Bayliss, D A ., 1999. Differential distribution of three members o f a gene family encoding low voltage-activated (T-type) . calcium channels. J. Neurosci. 19, 1895-1911.

Tarasenko, A.N., Kostyuk, P.G., Eremin, A.V., Isaev, D.S., 1997. Two types of low voltage-activated Ca "*" channels in neurones o f rat laterodorsal thalamic nucleus. J. Physiol. 499, 77-86.

Todorovic, S.M., Lingle, C.J., 1998. Pharmacological properties of T-type Ca '*' current in adult rat sensory neurons: effects of anticonvulsant and anesthetic agents. J. Neurophysiol. 79, 240-252.

Vanegas, H., Schaible, H., 2000. Effects of antagonists to high-threshold calcimn charmels upon spinal mechanisms of pain, hyperalgesia and allodynia. Pain 85, 9 -18 .

Walker, D., D e Waard, M., 1998. Subunit interaction sites in voltage-de­pendent Ce?* charmels: role in charmel function. Trends Neurosci. 21, 148-154.

WaU, P.D., Devor, M., 1983. Sensory afferent impulses originate from dorsal root ganglia as weE as firom the periphery in normal and nerve injured rats. Pain 17, 321-339.

262

EA. Matthews, A.H. Dickenson / European Journal o f Pharmacology 415 (2001) 141-149 149

White, D.M., Cousins, M J., 1998. Effect of subcutaneous administration Yaksh, TJL., Hammond, D.L., 1982. Peripheral and central substrates of calcium channel blockers on nerve injury-induced hyperalgesia. involved in the rostrad transmission of nociceptive information. PainBrain Res. 801, 50-58. 13, 1-85.

Xiao, W.H., Bennett, G.J., 1995. Synthetic omega-conopeptides applied Zimmermann, M., 1983. Ethical guidelines for investigations of experi-to the site of nerve injury suppress nemropathic pains in rats. J. mental pain in conscious animals [editorial]. Pain 16, 109—110.Pharmacol. Exp. Ther. 274, 666-672.

263