discovery.ucl.ac.uk · 2020-07-07 · abstract neuropathic pain can involve exaggerated pain...
TRANSCRIPT
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: [email protected] (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: [email protected] (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