THE REGULATION OF NEURONAL EXCITABILITY AND

NOCICEPTION BY TONIC GABAERGIC INHIBITION

by

Robert Paul Bonin

A thesis submitted in conformity with the requirements

for the degree of Doctor of Philosophy

Department of Physiology

University of Toronto

© Copyright by Robert P. Bonin 2011

ii

The Regulation of Neuronal Excitability and Nociception by Tonic

GABAergic Inhibition

Robert Paul Bonin

Doctor of Philosophy

Department of Physiology

University of Toronto

2011

Abstract

The mammalian central nervous system maintains a delicate balance between neuronal excitation

and inhibition. Conventional synaptic inhibition is mediated through the transient activity of

postsynaptic γ-aminobutyric acid (GABA) at type A GABA (GABAA) receptors. A subset of

GABAA receptors is also located outside of inhibitory synapses. These extrasynaptic receptors

generate a tonic inhibitory conductance in response to low concentrations of extracellular

GABA. Tonic inhibition broadly suppresses neuronal activity and regulates many vital processes

such as sleep, consciousness and memory formation.

This thesis examines the physiological effects of tonic inhibition at the cellular level and

in the behaving animal. This thesis also explores whether gabapentin, a commonly used sedative,

anxiolytic, and analgesic drug, enhances tonic GABAergic inhibition. I hypothesize that: (1)

tonic GABAA receptor activity reduces the intrinsic excitability of neurons; (2) the activity of

tonically active GABAA receptors in spinal pain pathways attenuates nociception; and (3) tonic

inhibition can be upregulated by gabapentin.

iii

The results show that a tonic inhibitory current generated by α5 subunit-containing

GABAA (α5GABAA) receptors reduces the excitability of hippocampal pyramidal neurons

excitability by increasing the rheobase, but does not change the gain of action potential firing. A

similar shunting inhibition is present in spinal cord lamina II neurons that is generated by δ

subunit-containing GABAA receptors. The activity of these receptors in spinal nociceptive

pathways reduces acute thermal nociception and may constrain central sensitization in a

behavioural model of persistent pain. Finally, gabapentin increases a tonic inhibitory current in

cultured hippocampal neurons independent from changes in the expression of α5GABAA

receptors or in the concentration of GABAA receptor ligands.

The results of this thesis demonstrate that tonically active GABAA receptors play an

important role in the regulation of neuronal activity and nociception, and that tonic inhibition

represents a novel target of therapeutic drugs.

iv

Acknowledgments

I would like to thank my supervisor, Dr. Beverly Orser for her guidance and mentorship over the

years that we have worked together. Her continued support, training and friendship have made

me a better scientist and a better person. I also thank my supervisory committee Drs. John

MacDonald and Melanie Woodin, who have provided invaluable guidance, and my examination

committee members, Drs. Delia Belelli, Colin McCartney, John Dostrovsky, and Richard

Horner.

I also thank all the members of the Orser lab whom I have been fortunate enough to work

with over the years. Your input and effort made this work easier, and your friendship made this

fun. I thank Agnieszka Zurek, who was first my student and became a great friend and mentor in

both science and life. I will let you know when the time machine is working; Ella Czerwinska,

for keeping the lab running and providing a constant supply of neuronal cultures that star so

prominently in this thesis, and for her jokes and cheeriness; Loren Martin, who I have always

looked up to. Paul Whissell; for his pervasive sanguinity; and Dianshi Wang, our quiet leader. I

also thank Dave Eng, Irene Lecker, William To, and Sinziana Avramescu, who all made the lab

feel like home. Finally, I reserve a special thank you to my family, who provided unwavering

emotional support. I thank my parents, John and Teresa, my brother, Steve, my sister, Amy, and

especially my wife, Anna, for their love, encouragement, and understanding.

List of Contributions

Several investigators contributed to the work presented in this thesis. In Chapter 4, Loren J.

Martin completed the electrophysiological experiments in hippocampal slices. In Chapter 5,

David G. Eng and Paul D. Whissell assisted with the behavioural experiments and Charalampos

Labrakakis assisted with the electrophysiological experiments in spinal slices. In Chapter 6,

Victor Y. Cheng completed a subset of the electrophysiological experiments and Mary W. Chiu

completed the western blot analysis of GABAA receptor expression. Finally, Gail Rauw

conducted the tissue analysis of amino acid and neuroactive steroid levels presented in Chapter 7.

This work was completed with the financial support provided to me by the Natural Sciences and

Engineering Research Council of Canada, the University of Toronto Department of Physiology,

and the University of Toronto Neuroscience Program.

v

Table of Contents

Acknowledgments .......................................................................................................................... iv

Table of Contents ............................................................................................................................ v

List of Figures ................................................................................................................................ xi

List of Tables ............................................................................................................................... xiii

List of Abbreviations ................................................................................................................... xiv

Chapter 1. Introduction ................................................................................................................... 1

1 Overview .................................................................................................................................... 1

1.1 GABA in the central nervous system .................................................................................. 2

1.2 GABAA receptors ................................................................................................................ 2

1.2.1 Tonically active GABAA receptors ......................................................................... 6

1.2.2 GABAA receptor subtypes generating a tonic current ............................................ 7

1.2.3 Sources of GABA responsible for tonic inhibition ............................................... 13

1.3 Physiological function of tonic inhibition ......................................................................... 14

1.3.1 Regulation of neuronal excitability ....................................................................... 14

1.3.2 Regulation of synaptic plasticity ........................................................................... 19

1.3.3 Behavioural roles of tonic inhibition .................................................................... 20

1.3.4 Interactions between α5GABAA and δGABAA receptors .................................... 22

1.4 GABAA receptors as targets for analgesic drugs .............................................................. 24

1.4.1 Regulation of nociception by GABAA receptors .................................................. 24

1.4.2 Regulation of central sensitization by GABAA receptors ..................................... 25

1.4.3 GABAA receptors expression in pain processing pathways ................................. 27

1.5 Gabapentin and GABAA receptor activity ........................................................................ 29

1.5.1 α2δ Ca2+

channel subunit ....................................................................................... 30

vi

1.5.2 Gabapentin and tonic GABAA receptor activity ................................................... 31

Chapter 2. Hypotheses and Aims .................................................................................................. 32

2 Overview .................................................................................................................................. 32

2.1 General Hypotheses .......................................................................................................... 33

2.2 Specific Aims .................................................................................................................... 33

Chapter 3. Methods ....................................................................................................................... 35

3 Overview .................................................................................................................................. 35

3.1 Animal models .................................................................................................................. 35

3.2 Tissue preparation and analysis ........................................................................................ 36

3.2.1 Cultured neurons ................................................................................................... 36

3.2.2 Acute spinal slices ................................................................................................. 37

3.3 Electrophysiological methods ........................................................................................... 37

3.3.1 Electrophysiological recording equipment ........................................................... 39

3.3.2 Drug delivery in vitro ............................................................................................ 40

3.3.3 Intracellular recording solutions ........................................................................... 41

3.3.4 Measurement of tonic current ............................................................................... 42

3.3.5 Measurement of neuronal excitability .................................................................. 42

3.4 Behavioural assays ............................................................................................................ 43

3.4.1 Hot plate assay ...................................................................................................... 43

3.4.2 Tail-flick assay ...................................................................................................... 43

3.4.3 Formalin injection test .......................................................................................... 43

3.4.4 Drug administration .............................................................................................. 44

Chapter 4. α5GABAA receptors regulate the intrinsic excitability of mouse hippocampal

pyramidal neurons .................................................................................................................... 45

4 Overview .................................................................................................................................. 45

4.1 Introduction ....................................................................................................................... 45

vii

4.2 Methods ............................................................................................................................. 48

4.2.1 Hippocampal cell culture and slice preparation. ................................................... 48

4.2.2 Electrophysiology ................................................................................................. 49

4.2.3 Data Analysis. ....................................................................................................... 51

4.2.4 Statistics. ............................................................................................................... 52

4.3 Results ............................................................................................................................... 52

4.3.1 Activation of α5GABAA receptors alters membrane potential ............................. 53

4.3.2 GABA hyperpolarizes WT but not Gabra5−/− neurons ...................................... 57

4.3.3 α5GABAA receptors regulate excitability of pyramidal neurons ......................... 61

4.3.4 Excitability of CA1 pyramidal neurons in hippocampal slices is similar to that

of cultured neurons ............................................................................................... 62

4.3.5 WT and Gabra5−/− neurons differ in sensitivity to GABAA receptor blockade . 65

4.3.6 Low concentrations of GABA further distinguish Gabra5−/− and WT neurons .................................................................................................................. 68

4.3.7 Spontaneous IPSCs in WT and Gabra5−/− neurons ........................................... 68

4.3.8 α5GABAA receptors do not influence the gain of the input-output relationship .. 69

4.4 Discussion ......................................................................................................................... 75

4.4.1 Adaptive changes in Gabra5−/− neurons ............................................................ 75

4.4.2 Synaptic inhibition and α5GABAA receptors ....................................................... 78

4.4.3 Tonic inhibition and excitability ........................................................................... 79

Chapter 5. The regulation of acute nociception and central sensitization by δ subunit-

containing GABAA receptors ................................................................................................... 82

5 Overview .................................................................................................................................. 82

5.1 Introduction ....................................................................................................................... 82

5.2 Materials and Methods ...................................................................................................... 84

5.2.1 Animal Subjects .................................................................................................... 84

5.2.2 Drugs and Chemicals ............................................................................................ 85

viii

5.2.3 Tissue Preparation ................................................................................................. 85

5.2.4 Electrophysiology ................................................................................................. 85

5.2.5 Analysis of Electrophysiology Data ..................................................................... 86

5.2.6 Tail Flick Assay .................................................................................................... 87

5.2.7 Hot Plate Assay ..................................................................................................... 87

5.2.8 Motor Coordination Assay .................................................................................... 88

5.2.9 Formalin Nociception Test ................................................................................... 88

5.3 Results ............................................................................................................................... 89

5.3.1 δGABAA receptors generate a tonic inhibitory current in cultured spinal neurons .................................................................................................................. 89

5.3.2 Tonic inhibition regulates neuronal excitability ................................................... 90

5.3.3 δGABAA receptors underlie a tonic inhibitory conductance in spinal lamina II

neurons from adult mice ....................................................................................... 97

5.3.4 Increasing spinal δGABAA receptor activity inhibits acute nociception ............ 100

5.3.5 A low dose of THIP does not impair motor function ......................................... 103

5.3.6 δGABAA receptors constrain sensitization in the formalin assay ....................... 103

5.4 Discussion ....................................................................................................................... 106

Chapter 6. Gabapentin increases a tonic inhibitory current in cultured hippocampal neurons .. 111

6 Overview ................................................................................................................................ 111

6.1 Introduction ..................................................................................................................... 111

6.2 Materials and Methods .................................................................................................... 116

6.2.1 Cell culture and electrophysiological techniques ............................................... 116

6.2.2 Measurement of tonic and synaptic inhibitory conductance ............................... 117

6.2.3 Measurement of GABA-evoked current ............................................................. 119

6.2.4 Membrane preparation and receptor solubilization ............................................ 119

6.2.5 Western blot testing of subunit-containing GABAA receptors ........................... 120

6.2.6 Materials ............................................................................................................. 121

ix

6.2.7 Statistical Analyses ............................................................................................. 121

6.3 Results ............................................................................................................................. 121

6.3.1 Brief gabapentin exposure does not change GABAA current ............................. 121

6.3.2 Prolonged gabapentin exposure increases tonic current ..................................... 122

6.3.3 Synergistic interactions between gabapentin and vigabatrin .............................. 131

6.3.4 Gabapentin does not increase GABAA receptor expression or sensitivity to GABA ................................................................................................................. 132

6.4 Discussion ....................................................................................................................... 138

Chapter 7. Gabapentin does not increase the production of GABAA receptor ligands .............. 142

7 Overview ................................................................................................................................ 142

7.1 Introduction ..................................................................................................................... 142

7.2 Methods ........................................................................................................................... 144

7.2.1 Animal subjects ................................................................................................... 144

7.2.2 Neuronal cultures ................................................................................................ 144

7.2.3 Analysis of amino acids ...................................................................................... 145

7.2.4 Analysis of neuroactive steroids ......................................................................... 146

7.3 Results ............................................................................................................................. 147

7.3.1 Gabapentin does not increase GABAA ligand concentration in vitro ................. 147

7.3.2 Free amino acid concentrations in the brain are not altered by gabapentin in vivo ...................................................................................................................... 152

7.3.3 Gabapentin does not increase tissues concentrations of neuroactive steroids. ... 152

7.4 Discussion ....................................................................................................................... 155

Chapter 8. Discussion ................................................................................................................. 157

8 Overview ................................................................................................................................ 157

8.1 Tonic inhibition of neuronal excitability ........................................................................ 157

8.1.1 Computational effects of tonic inhibition on hippocampal function. ................. 157

8.1.2 Neuroactive steroids as endogenous regulators of neuronal excitability ............ 160

x

8.1.3 Dual regulation of neuronal excitability by α5GABAA and δGABAA receptors 162

8.2 The role of tonic inhibition in nociception ..................................................................... 164

8.2.1 Sensory pathways modulated by δGABAA receptors ......................................... 164

8.2.2 Tonic inhibition of dorsal horn neurons by α5GABAA receptors....................... 165

8.2.3 Neuroactive steroids and nociception ................................................................. 166

8.3 Future directions ............................................................................................................. 167

Appendix GABAA receptors as targets of anesthetics ................................................................ 170

9 Overview ................................................................................................................................ 170

9.1.1 Evolution of anesthetic theory ............................................................................ 171

9.1.2 GABAA Receptors are major anesthetic targets ................................................. 172

9.1.3 Anesthetic binding pocket on GABAA receptors ................................................ 174

9.1.4 Multimodal anesthesia ........................................................................................ 175

9.1.5 Volatile anesthetics ............................................................................................. 186

10 References .............................................................................................................................. 189

xi

List of Figures

Figure 1.1. GABAA receptor structure and function. ...................................................................... 5

Figure 1.2. Subtractive and divisive inhibition of neuronal activity. ............................................ 18

Figure 4.1. Increased tonic GABAergic current in cultured WT neurons. ................................... 56

Figure 4.2. GABA evoked current is hyperpolarizing in cultured WT and Gabra5−/− neurons. 59

Figure 4.3. Picrotoxin occluded the difference in excitability in cultured WT and Gabra5−/−

neurons. ......................................................................................................................................... 64

Figure 4.4. GABAA receptor antagonists modify the excitability of cultured WT but not

Gabra5−/− neurons. ..................................................................................................................... 67

Figure 4.5. Deletion of α5GABAA receptors does not modify spontaneous IPSCs in cultured

pyramidal neurons. ........................................................................................................................ 72

Figure 4.6. α5GABAA receptors do not regulate the frequency of action potentials in cultured

neurons. ......................................................................................................................................... 74

Figure 5.1. The amplitude and sensitivity of the tonic current to neuroactive steroid is reduced in

Gabrd−/− spinal neurons grown in culture. ................................................................................. 93

Figure 5.2. THIP reduced the firing of action potentials in WT but not Gabrd−/− spinal neurons.

....................................................................................................................................................... 96

Figure 5.3. Reduced tonic current amplitude and sensitivity to neuroactive steroids in lamina II

neurons from adult WT and Gabrd−/− mice. .............................................................................. 99

Figure 5.4. δGABAA receptors in the spinal cord regulate nociception. .................................... 102

Figure 5.5. δGABAA receptors modulate responses in the formalin assay. ............................... 105

Figure 6.1. Gabapentin and tonic inhibition ............................................................................... 115

xii

Figure 6.2. Representative traces of whole-cell current activated by GABA. ............................ 124

Figure 6.3. Prolonged gabapentin treatment increases tonic current. ......................................... 127

Figure 6.4. Gabapentin does not modify GABAB receptor activity. ........................................... 130

Figure 6.5. Synergistic interactions between gabapentin and vigabatrin. ................................... 135

Figure 6.6. Gabapentin does not change GABAA receptor expression or sensitivity to GABA. 137

Figure 7.1. Gabapentin treatment did not increase media concentrations of inhibitory ligands

when applied to cultured neurons. .............................................................................................. 149

Figure 7.2. Gabapentin did not increase tissue levels of amino acids in brains of mice. ........... 151

Figure 7.3. Gabapentin does not increase tissue levels of neuroactive steroids in mice. ........... 154

Figure 9.1. Synaptic and extrasynaptic activation of GABAA receptors .................................... 177

Figure 9.2. Amnestic actions of etomidate, as mediated by α5 subunit-containing -aminobutyric

acid subtype A receptors. ............................................................................................................ 180

Figure 9.3. Etomidate hypnosis and immobilization in β3 (Asn265Met) and β2(Asn265Ser) mice.

..................................................................................................................................................... 184

xiii

List of Tables

Table 4.1. Action potential characteristics .................................................................................... 60

Table 4.2. Properties of spontaneous IPSCs recorded in CA1 pyramidal neurons in hippocampal

slices .............................................................................................................................................. 70

Table 6.1. Spontaneous mIPSCs recorded in pyramidal neurons showed no change in frequency,

amplitude, or time course after treatement with gabapentin or vigabatrin for 36 – 48 hours. .... 128

Table 9.1. The neuronal, physiological and pharmacological roles of putative anesthetic targets in

the central nervous system. ......................................................................................................... 173

xiv

List of Abbreviations

aCSF = Artificial cerebral spinal fluid

AMPA = α-amino-3-hydroxyl-5-methyl-4-isoxazole-propionate

ANOVA = Analysis of variance

APV = (2R)-amino-5-phosphonovaleric acid

D-AP5 = D(-)aminophosphopentanoic acid

ATP = Adenosine triphosphate

5GABAA = α5 subunit-containing γ-aminobutyric acid subtype A (receptor)

BIC = Bicuculline

CA1 = Cornu Ammonis area 1

CA3 = Cornu Ammonis area 3

CNS = Central nervous system

CNQX = 6-cyano-7-nitroquinoxaline-2,3-dione

δGABAA = δ subunit-containing γ-aminobutyric acid subtype A (receptor)

ECF = Extracellular fluid

EGTA = Ethylene glycol-bis(2-aminoethylether)-N,N,N′,N′-tetraacetic acid

GABA = -Aminobutyric acid

GABAA = -Aminobutyric acid subtype A (receptor)

Gabra5–/– = 5 subunit-containing -aminobutyric acid subtype A gene deletion

Gabrd–/– = δ subunit-containing -aminobutyric acid subtype A gene deletion

GTP = Guanosine triphosphate

HCN = Hyperpolarization-activated, cyclic nucleotide-gated channel

HEPES = 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid

ICF = Intracellular fluid

Ihold = Holding current

i.p. = Intraperitoneal

i.t. = Intrathecal

IPSC = Inhibitory postsynaptic current

IPSP = Inhibitory postsynaptic potential

KCC2 = K+ - Cl

- co-transporter

L6 = L-655,708

xv

LTD = Long-term depression

LTP = Long-term potentiation

MAC = Minimum alveolar concentration

mIPSC = Miniature inhibitory postsynaptic current

NMDA = N-methyl-D-aspartic acid

PTX = Picrotoxin

RMS = Root mean square

RT-PCR = Reverse transcription polymerase chain reaction

sIPSC = Spontaneous inhibitory postsynaptic potential

TEA = Tetraethylammonium

THIP = 4,5,6,7-Tetrahydroisoxazolo[4,5c]pyridine-3-ol

THDOC = 3α,21-dihydroxy-5α-pregnan-20-one; Tetrahydrodeoxycorticosterone

3α,5α-THPROG = 5α-pregnan-3α-ol-20-one; allopregnanolone

TTX = Tetrodotoxin

VDCC = Voltage-dependent calcium channel

Vm = Membrane potential

WT = Wild type

1

Chapter 1. Introduction

1 Overview

The function of the mammalian central nervous system (CNS) requires a delicate balance

between neuronal excitation and inhibition. The majority of inhibition in the brain is mediated

through the actions of the neurotransmitter γ-aminobutyric acid (GABA) and as many as one-

third of all synapses in the brain are GABAergic (Bloom and Iversen, 1971). At inhibitory

synapses, GABA binds to GABA subtype A (GABAA) receptors to induce rapid and transient

inhibition of the postsynaptic neuron. A more recently characterized subpopulation of GABAA

receptors exists outside the synapse (Belelli et al., 2009; Semyanov et al., 2004). These

extrasynaptic GABAA receptors are activated by the ambient concentrations of GABA that are

continuously present in the extracellular space and generate a persistent or „tonic‟ inhibitory

current (Glykys and Mody, 2007; Kullmann et al., 2005). The physiological role of tonic

inhibition in the regulation of neuronal activity and behaviour is only beginning to be fully

appreciated.

Tonic inhibition is generated by a biophysically and pharmacologically distinct

subpopulation of GABAA receptors. The unique properties of these receptors are determined by

their subunit composition. Tonic inhibition is typically generated by GABAA receptors

containing either an α5 or δ receptor subunit (Belelli et al., 2009; Semyanov et al., 2004).

Broadly, this thesis explores how tonic inhibition generated by α5 subunit-containing GABAA

(α5GABAA) receptors and δ subunit-containing GABAA (δGABAA) receptors regulates neuronal

activity and mammalian behaviour. This topic is explored using a variety of complementary

experimental models ranging from electrophysiological measurement of the tonic current in

cultured neurons to the study of nociceptive behaviour in mice. This chapter will provide an

introduction to the subjects and concepts that are central to this thesis.

2

1.1 GABA in the central nervous system

GABA is derived from glutamate by the enzyme glutamic acid decarboxylase (GAD). Two

distinct GAD isoforms exist: GAD65 and GAD67 (Martin et al., 1998). GAD65 is localized to

synaptic terminals and is associated with the surface of synaptic vesicles where it mediates both

GABA synthesis and the packaging of GABA into synaptic vesicles (Jin et al., 2003). In

contrast, GAD67 is distributed throughout the cytoplasm and GABA synthesized by this enzyme

can be released via non-vesicular mechanisms (Kaufman et al., 1991).

During the action potential-dependent activation of GABAergic synapses, the

concentration of GABA that is released into the synaptic cleft reaches very high levels that can

saturate the GABA receptors in the postsynaptic membrane (Maconochie et al., 1994). The

concentration of GABA within the synaptic cleft is decreased by passive diffusion of

neurotransmitter away from synapses, reuptake into presynaptic terminals or surrounding

astrocytes, or through GABA metabolism (Cavelier et al., 2005). The reuptake of GABA is the

primary mode of extracellular GABA removal and occurs via the actions of GABA transporters

(GAT)(Conti et al., 2004). Four subtypes of GAT exist numbered GAT1 through GAT4. GAT1

is the major GABA transporter in neurons while GAT4 is the major GABA transporter in

astrocytes (Conti et al., 2004). Extracellular GABA is also metabolized by GABA transaminase

(GABA-T), which converts GABA to succinic semialdehyde (Tao et al., 2006). The removal of

GABA from the synaptic cleft is never complete and it is estimated that the extracellular GABA

concentrations normally averages between 0.1 – 0.4 M (Attwell et al., 1993).

1.2 GABAA receptors

The fast inhibitory effects of GABA are mediated through ionotropic GABA subtype A

(GABAA) receptors and metabotropic GABA subtype B (GABAB) receptors. The focus of this

thesis is the function and physiological effects of GABAA receptors. GABAB receptors will not

be discussed further. Each GABAA receptor is a pentameric complex containing an integral

anion channel that is permeable to Cl- and, to a lesser extent, bicarbonate ions (Macdonald and

3

Olsen, 1994; Moss and Smart, 2001) (Fig 1.1). GABAA receptors are targets of many

neurodepressive drugs including barbiturates, etomidate, propofol, inhaled anesthetics, and

neuroactive steroids, as well as the sedative and hypnotic benzodiazepines (Belelli and Lambert,

2005; Bonin and Orser, 2008; Franks, 2008; Kopp Lugli et al., 2009). A detailed description of

GABAA receptors as targets for anesthetic drugs is provided in the Appendix. The contents of the

Appendix have been published as a review article in the journal Pharmacology, Biochemistry

and Behaviour (Bonin and Orser, 2008).

Binding of GABA to the GABAA receptor causes a conformational change, which leads

to channel opening and the flux of Cl- anions across the cell membrane. In most mature neurons,

an increase in Cl- permeability results in membrane hyperpolarization due to the flux of Cl

- into

the cell. This process is dependent on the maintenance of a low intracellular Cl- concentration. A

low concentration of Cl- in neurons is maintained through the activity of the K

+ - Cl

- co-

transporter (KCC2), which transports K+ and Cl

- out of the cell (Rivera et al., 1999). The opening

of GABAA receptors also increases the membrane conductance, which can result in shunting

inhibition. Excitatory synaptic input depolarizes the neuronal membrane potential in accordance

with Ohm‟s law, V = I × R. GABAA receptor activity reduces membrane resistance, resulting in

a reduction in membrane depolarization by excitatory synaptic input. This current shunt reduces

the ability of excitatory input to evoke an action potential and thus inhibits neuronal activity

(Silver, 2010).

4

5

Figure 1.1. GABAA receptor structure and function.

(Left) The GABAA receptor is a heteropentameric transmembrane protein. When GABA binds,

the Cl--permeable channel opens. The GABAA receptor is the binding site of many anesthetics

and neurodepressive drugs that allosterically enhance GABAA current. (Right) Action potential-

dependent release of GABA into the synaptic cleft transiently activates GABAA receptors in the

postsynaptic membrane. Inhibitory postsynaptic currents are typically of a brief duration due to

GABA diffusion and uptake (Olsen and Sieghart, 2009). Extrasynaptic GABAA receptors are

activated by low concentrations of GABA in the extracellular space arising from the synaptic

spillover or non-synaptic release mechanisms (Kullmann et al., 2005). These receptors have low

rates of desensitization and produce a continuous or „tonic‟ current. The tonic current is revealed

by the application of a GABAA antagonist, which inhibits the current. Many anesthetics potently

enhance the tonic current at clinically-relevant concentrations.

6

The GABAA receptor is a heteromeric structure composed of multiple subunits. At least

19 mammalian genes encode for the various subunits (α1–6, β1–3, γ1–3, δ, ε, θ, π and ρ1-

3)(Olsen and Sieghart, 2008, 2009). Typically, GABAA receptors are composed of α, β, and γ

subunits in a ratio of 2:2:1; however, in some brain regions, the γ subunit may be absent and be

replaced by a δ, θ, π or ε subunit (Farrant and Nusser, 2005; Hevers and Luddens, 1998;

Sieghart, 1995). The pharmacological and biophysical properties of the GABAA receptor are

determined by the subunit composition of the receptor. Most relevant to this thesis, the subunit

composition of the receptor determines whether the receptors are more likely to participate in the

generation of transient postsynaptic inhibition, or a form of persistent, „tonic‟ inhibition

(Semyanov et al., 2004). The γ subunit-containing GABAA receptors are typically clustered in

the postsynaptic membrane of the inhibitory synapses (Nusser et al., 1996). In contrast, GABAA

receptors containing the δ subunit (δGABAA) are expressed predominantly outside of the

synapse and generate a tonic inhibitory current (Nusser et al., 1998). One major exception to this

pattern is the α5β3γ2 receptor, which is located predominantly outside the synapse in the

hippocampus and olfactory bulb (Brunig et al., 2002; Pirker et al., 2000; Sur et al., 1999a).

1.2.1 Tonically active GABAA receptors

GABAA receptors that are clustered at postsynaptic terminals are briefly activated by very high,

near-saturating concentrations of GABA (Maconochie et al., 1994). At the level of the neuronal

networks, this form of synaptic or „phasic‟ inhibition is thought to be critical for high-fidelity

neuronal communication, the precise timing of action potentials, and the synchronization of

neuronal rhythms (Cobb et al., 1995; Pouille and Scanziani, 2001).

Extrasynaptic GABAA receptors are typically considered to have a high affinity for

GABA and generate whole-cell currents that slowly desensitize (Belelli et al., 2009; Semyanov

et al., 2004). These two properties enable extrasynaptic GABAA receptors to generate tonic

inhibitory current. In whole-cell electrophysiological experiments, the tonic current can be

revealed by application of a GABAA receptor antagonist, such as bicuculline (Fig 1.1). The

application of an antagonist prevents the persistent activation of GABAA receptors, as evidenced

by a change in the current required to hold or clamp a neuron at a specific membrane potential.

7

The tonic current can be activated by the ambient concentrations of GABA present in

vitro, as seen in hippocampal slices (Martin et al., 2009; Martin et al., 2010) and thalamic

slices(Belelli et al., 2005), as well as in cultured hippocampal neurons (Bonin et al., 2007).

Alternatively, the tonic current can be enhanced through the application of exogenous GABA in

vitro (Semyanov, 2005) or by compounds that block GABA metabolism (Caraiscos et al.,

2004b). Tonic GABAergic inhibition is present in a cell-type and regionally specific manner, and

is found throughout the mammalian brain (Kullmann et al., 2005), particularly in cerebellar

granule cells (Brickley et al., 1996; Wall and Usowicz, 1997), the hippocampus (Bai et al., 2001;

Nusser and Mody, 2002; Stell and Mody, 2002), thalamocortical relay neurons of the ventral

basal complex (Belelli et al., 2005; Cope et al., 2005; Jia et al., 2005; Porcello et al., 2003), and

the neocortex (Drasbek et al., 2007; Salin and Prince, 1996).

1.2.2 GABAA receptor subtypes generating a tonic current

The evidence for the existence of distinct GABAA receptor populations that generate phasic

versus tonic inhibitory currents was provided by pharmacological experiments. The tonic

GABAA current was first revealed in cerebellar granule cells by the application of the GABAA

receptor blocker, bicuculline (Brickley et al., 1996). In contrast to synaptic currents, the tonic

current in these neurons is insensitive to the benzodiazepine, zolpidem. A benzodiazepine-

insensitive tonic current has been similarly identified in granule cells of the hippocampus dentate

gyrus (Nusser and Mody, 2002; Stell et al., 2003), hippocampal interneurons (Semyanov et al.,

2003), thalamocortical relay neurons of the ventral basal complex (Belelli et al., 2005; Cope et

al., 2005; Jia et al., 2005; Porcello et al., 2003), and neocortex(Drasbek et al., 2007; Salin and

Prince, 1996). The tonic current that is present in these neurons is generated by benzodiazepine-

insensitive GABAA receptors that contain the δ subunit (δGABAA receptors) (Bianchi et al.,

2002; Bianchi and Macdonald, 2002).

A tonic GABAergic current has also been well characterized in pyramidal neurons of the

hippocampal CA1 region. This tonic current is enhanced by benzodiazepines suggesting that

γGABAA reeptors contribute to the current (Caraiscos et al., 2004b). In hippocampal pyramidal

neurons, the GABAA antagonist SR-95531 can be used to discriminate between putatively

8

synaptic and extrasynaptic GABAA receptors (Bai et al., 2001). A low concentration of SR-

95531 (1 μM) blocks only synaptic GABAA activity, as indicated by a selective loss of miniature

inhibitory postsynaptic current (mIPSCs). High concentrations of penicillin similarly block the

synaptic, but not the tonic current (Yeung et al., 2003). These findings provide further support

that the tonic current is generated by a pharmacologically distinct subpopulation of GABAA

receptors. The tonic inhibitory current in these hippocampal pyramidal neurons is largely

generated by GABAA receptors containing α5β3γ2 subunits (Caraiscos et al., 2004b; Ju et al.,

2009). δGABAA receptor activity also contributes to the tonic current in hippocampal pyramidal

neurons, albeit to a lesser extent than α5GABAA receptors under baseline conditions (Glykys et

al., 2008; Scimemi et al., 2005). A tonic current current generated by α5GABAA receptors has

also been identified in medium spiny neurons of the striatum (Ade et al., 2008).

1.2.2.1 δGABAA receptors

The δGABAA receptor subunit is expressed throughout the CNS in a tissue specific

manner. In adult animals, high levels of expression occur in the striatum, granule cell layer of the

cerebellum, ventrobasal thalamus, hippocampal dentate gyrus, and lower layers of the cortex

(Pirker et al., 2000). At a cellular level, δGABAA receptors are found almost exclusively in the

extrasynaptic space (Nusser et al., 1998; Peng et al., 2002; Wei et al., 2003). In general, the δ

subunit forms GABAA receptors in partner with the α4 subunit (Sur et al., 1999b). A notable

exception to this rule occurs in cerebellar granule cells, where the δ subunit forms GABAA

receptors in partner with α6 subunit (Brickley et al., 2001). Deletion of the α6 subunit or of the

α4 subunit reduces the tonic current recorded in the cerebellum and thalamus, respectively

(Brickley et al., 2001; Chandra et al., 2006).

A tonic inhibitory current generated by δGABAA receptors has been detected in all

regions where expression of the δGABAA receptor subunit is high, including the cerebellum

(Brickley et al., 1996), thalamus (Belelli et al., 2005; Cope et al., 2005; Jia et al., 2005; Porcello

et al., 2003), hippocampal dentate gyrus (Nusser and Mody 2002; Stell 2003), and neocortex

(Drasbek et al., 2007; Salin and Prince, 1996). Additionally, a tonic current generated by

9

putatively δGABAA receptors has been reported in lamina II neurons of the spinal cord (Mitchell

et al., 2007; Takahashi et al., 2006).

Recombinant δGABAA receptors exhibit unique biophysical properties. δGABAA

receptors have a high sensitivity to GABA. Replacing the γ subunit with δ reduces the EC50 of

GABA to the sub-millimolar range (Brown et al., 2002). However, δGABAA receptors have

smaller macroscopic current amplitudes than γGABAA receptors, owing to a reduced mean

channel open time and shorter duration of channel open bursts (Fisher and Macdonald, 1997).

Yet, δGABAA receptors have a comparable single channel conductance to γGABAA receptors

(Brickley et al., 1999; Fisher and Macdonald, 1997). Additionally, recombinant δGABAA

receptors have a slower rate of desensitization compared to γGABAA receptors (Bianchi and

Macdonald, 2002; Saxena and Macdonald, 1996). Overall, the kinetic properties of δGABAA

receptors result in a receptor population that generates a low amplitude current that persists in the

continued presence of agonist. δGABAA receptors are therefore ideally suited to respond to low,

ambient concentrations of GABA.

1.2.2.1.1 δGABAA receptor pharmacology

δGABAA receptors have distinct pharmacological properties. One key characteristic of δGABAA

receptor activity is their insensitivity to benzodiazepines (Cope et al., 2005; Nusser and Mody,

2002). Nevertheless, δGABAA receptors are targets for many sedative and hypnotic drugs

(Belelli et al., 2009). Notably, ethanol has been recently shown to exert a neurodepressive effect

by enhancing δGABAA activity (Boehm et al., 2006; Glykys et al., 2007; Wei et al., 2004). Also,

the sedative and analgesic drug 4,5,6,7-tetrahydroisoxazolo[4,5c]pyridine-3-ol (THIP), or

„gaboxadol‟, is a selective agonist at the δGABAA receptor (Brown et al., 2002; Storustovu and

Ebert, 2006). At δGABAA receptors, THIP evokes a larger maximal current than GABA, which

acts as partial agonist at δGABAA receptors. In contrast, THIP activates a current of similar

amplitude to GABA at non-δGABAA receptors. The potency of THIP is also higher at δGABAA

receptors than non-δGABAA receptors (Storustovu and Ebert, 2006). For these reasons, THIP

has been used as a pharmacological tool to probe the in vivo effects of δGABAA receptor activity

(Chandra et al., 2006).

10

Caution must be used when interpreting the in vivo effects of THIP. The GABAA receptor

selectivity of THIP is highly dependent on concentration. Saturating concentrations of THIP can

activate most γGABAA receptors to the same extent as saturating concentrations of GABA

(Storustovu and Ebert, 2006). THIP also has potent effects at α5GABAA receptors and activate

α5GABAA receptors at concentrations of less than 10 μM (Storustovu and Ebert, 2006). Yet, at

low concentrations, THIP evokes a current of considerably higher magnitude at recombinant

δGABAA receptors than at recombinant α5GABAA receptors (Storustovu and Ebert, 2006).

δGABAA receptors are also uniquely sensitive to modulation by neuroactive steroids

(Belelli and Lambert, 2005). The potentiation of the δGABA current by neuroactive steroids is

considerably greater than that of γGABAA receptors (Brown et al., 2002; Wohlfarth et al., 2002).

This is largely due to the low gating efficacy of GABA at δGABAA receptors, which is greatly

increased by neuroactive steroid binding (Bianchi and Macdonald, 2003). The potency with

which steroids enhance GABAA receptor function is determined by the α subunit of the GABAA

receptor (Belelli et al., 2002). The influence of the α subunit on steroid potency is due in part to

the presence of a neuroactive steroid binding site near the α subunit of GABAA receptors (Hosie

et al., 2009; Hosie et al., 2006).

Neuroactive steroids, including the progesterone metabolite, 5α-pregnan-3α-ol-20-one

(3α,5α-THPROG) and the deoxycorticosterone metabolite 5α-pregnan-3α,21-diol-20-one

(THDOC), enhance GABAA receptor function (Belelli and Lambert, 2005). These steroids can

be locally synthesized by enzymes expressed within the central nervous system and do not

require peripheral synthesis and transport to target sites (Agis-Balboa et al., 2006; Mensah-

Nyagan et al., 2008). Local synthesis allows neuroactive steroids to act in a paracrine fashion to

influence a small population of neurons. The enzyme 5α-reductase converts progesterone to 5α-

dihydroprogesterone, which is converted to 3α,5α-THPROG by 3α-hyroxysteroid dehydrogenase

(3α-HSD)(Mellon and Griffin, 2002). Similarly, 5α-reductase converts 11-deoxycorticosterone

to 5α-dihydrodeoxycorticosterone, which is converted by 3α-HSD to THDOC. Thus, the

enzymes 5α-reductase and 3α-HSD are crucial for the synthesis of both steroids. These two

enzymes are broadly expressed throughout the nervous system, including the spinal cord (Patte-

Mensah et al., 2006).

11

Overall, δGABAA receptors are the most functionally sensitive of the GABAA receptors

to neuroactive steroids. Null mutant mice that do not express δGABAA receptors (Gabrd−/−)

have a greatly reduced behavioural response to exogenously administered neuroactive steroids

(Mihalek et al., 1999). Synthetic steroids, such as alphaxalone and ganaxalone, also

preferentially enhance δGABAA receptor function. The recent development of the selective

δGABAA receptor positive allosteric modulator, 4-chloro-N-[2-(2-thienyl)imidazo[1,2-

a]pyridine-3-yl benzamide (DS2), may allow the more precise pharmacological study of

δGABAA receptors in the future (Wafford et al., 2009).

1.2.2.2 α5GABAA receptors

The distribution of α5GABAA receptors is relatively sparse in the mammalian brain. The α5

subunit is incorporated into only 5% of the total GABAA receptors in the brain, but this

proportion increases to 20-30% of GABAA receptors in the hippocampus (Sur et al., 1999a). In

the hippocampus, the α5 subunit is localized to the apical and basal dendrites of pyramidal

neurons in the stratum radiatum and stratum oriens of the CA1 and CA3 regions (Pirker et al.,

2000). Additionally, a high density of α5 subunit-immunoreactivity has been detected in the

olfactory bulb, inner layers of the cerebral cortex, endopiriform nucleus, subiculum and

ventromedial hypothalamic nucleus (Pirker et al., 2000). The function of α5GABAA receptors in

these brain regions has not been extensively studied.

Through a series of experiments using immunocytochemistry and in situ hybridization

(Brunig et al., 2002), immunofluorescence (Christie et al., 2002), and immunogold labeling

(Serwanski et al., 2006), it was determined that α5GABAA receptors are localized primarily to

extrasynaptic regions of pyramidal neurons in the hippocampus and cortex, with a smaller

proportion present within GABAergic synapses. α5GABAA receptors are also expressed at low

densities within the synaptic sites of interneurons (Christie et al., 2002). This mixed expression

pattern contrasts with δGABAA receptors, which are almost exclusively found extrasynaptically

(Nusser et al., 1998; Peng et al., 2002; Wei et al., 2003).

The expression pattern of α5GABAA receptors likely arises from the partnering of the α5

12

subunit with the γ2 subunit in GABAA receptors (Ju et al., 2009; Sur et al., 1998). The γ2 subunit

interacts with a synaptic scaffolding network of proteins including gephyrin, which is critical for

anchoring GABAA receptors at the synapse (Essrich et al., 1998). α5GABAA receptors also

uniquely interact with the cytoskeletal protein, radixin, which binds to F-actin in a

phosphorylation-dependent manner (Ju et al., 2009; Loebrich et al., 2006). The interaction of

α5GABAA receptors with radixin may be responsible for the extrasynaptic localization of these

receptors (Loebrich et al., 2006).

1.2.2.2.1 Synaptic and extrasynaptic α5GABAA receptor activity

Electrophysiological studies show that α5GABAA receptors are expressed at both synaptic and

extrasynaptic locations (Caraiscos et al., 2004b; Zarnowska et al., 2009). However, there is

stronger evidence for the contribution of α5GABAA receptors to tonic current than to synaptic

current. In hippocampal slices, benzodiazepines enhance the amplitude of spontaneous IPSCs

(sIPSCs) and evoked IPSCs, yet the enhancement of synaptic events is reduced in slices from

mutant mice that express a point mutation (H105R) in the α5 subunit that renders the α5GABAA

receptor insensitive to benzodiazepines (Zarnowska et al., 2009). Additionally, the amplitude of

evoked IPSCs was reduced and decay times were slowed in hippocampal slices prepared from

mice that do not express the α5 subunit (Gabra5–/– ) (Collinson et al., 2002). In contrast, the

kinetics and amplitude of sIPSCs are not altered by pharmacologically inhibiting α5GABAA

receptors (Caraiscos et al., 2004b; Glykys et al., 2008). Moreover, low concentrations of the

general anesthetic etomidate enhance a tonic conductance in WT slices, but do not alter the

frequency or peak amplitude of sIPSCs in hippocampal pyramidal neurons (Cheng et al., 2006b;

Martin et al., 2009). It is possible that α5GABAA receptors contribute to a subset of synaptic

events or only contribute to synaptic events under specific network conditions. In agreement with

this, only a subset of IPSCs in hippocampal neurons are mediated by α5GABAA receptors

(Zarnowska et al., 2009). These IPSCs have a slow rise time and may be caused by the activity

of specific interneuron populations.

13

1.2.2.2.2 Pharmacology of α5GABAA receptors

α5GABAA receptors are sensitive to positive allosteric modulation by benzodiazepines.

However, the benzodiazepine zolpidem has negligible affinity for α5GABAA receptors (Ki >

15,000 nM) and is considered to be “α5GABAA receptor sparing” (Pritchett and Seeburg, 1990).

Certain benzodiazepine-like compounds negatively modulate GABAA receptor function through

allosteric effects. These drugs are called „inverse agonists‟ because they reduce the intrinsic

activity of the receptor. Several inverse agonists have been developed with high affinity for the

α5GABAA receptors.

The inverse agonist L-655,708 displays a 50 to 100 fold higher affinity for α5GABAA

receptors over other benzodiazepine-sensitive GABAA receptors (Quirk et al., 1996). In vivo, L-

655,708 has cognitive-enhancing effects but has a relatively short half-life (Atack et al., 2006).

Several other inverse agonists displaying α5GABAA receptor selectivity have been developed

since the characterization of L-655,708 (Chambers et al., 2003; Chambers et al., 2004; Dawson

et al., 2006). L-655,708 has been used to demonstrate that α5GABAA receptors do not contribute

to synaptic GABAergic events in hippocampal pyramidal neurons (Caraiscos et al., 2004b;

Glykys et al., 2008).

1.2.3 Sources of GABA responsible for tonic inhibition

The low concentrations of extracellular GABA that activate the tonic inhibitory conductance

may arise through several mechanisms. One key source of extracellular GABA in the

hippocampus is spill-over from the synaptic cleft (Glykys and Mody, 2007). There is a

correlation between the frequency of IPSCs and the amplitude of the tonic current. A similar

spill-over mechanism contributes to the tonic current in cerebellar granule cells (Hamann et al.,

2002) and lamina II neurons of the spinal cord (Ataka and Gu, 2006). Yet, a tonic current can

often be detected in the presence of compounds that inhibit action potential-dependent release of

GABA and in the absence of exogenously applied GABA (Bai et al., 2001; Prenosil et al., 2006).

This suggests that synaptic spill-over is unlikely to be the sole source of GABA responsible for

the tonic inhibitory current. Non-vesicular mechanisms may also contribute to the extrasynaptic

14

GABA pool. GABA can be released from astrocytes (Kozlov et al., 2006; Liu et al., 2000), and

through the reversal of GABA transporters (Gaspary et al., 1998).

1.3 Physiological function of tonic inhibition

1.3.1 Regulation of neuronal excitability

Tonic inhibition regulates neuronal activity through two mechanisms: hyperpolarization of the

membrane and shunting inhibition. Membrane hyperpolarization occurs when the anion Cl-

enters the cell through open GABAA receptors to create an outward current. Because action

potentials are generated when neurons depolarize, membrane hyperpolarization reduces the

probability that an action potential will be generated. For Cl- flux to create an outward current,

there must be a low intracellular concentration of Cl-, which is maintained through the activity of

KCC2 (Rivera et al., 1999). In mature neurons the reversal or equilibrium potential of Cl- in

neurons is close to the resting membrane potential. Accordingly, there is often a low driving

force for Cl- influx which results in a low amplitude anion current when GABAA channels open.

The inward Cl- flux will increase as the membrane potential depolarizes. In this manner, GABAA

receptor activity can „clamp‟ the neuronal membrane potential near the Cl- reversal potential.

The intracellular concentration of Cl- can become elevated in many situations. The

expression of KCC2 is developmentally regulated, and a low expression of KCC2 in immature

neurons results in a high intracellular Cl- concentration (Rivera et al., 1999). The depolarization

of immature neurons through GABAA receptor activity may contribute to synapse formation and

neuronal migration (Ben-Ari, 2002). KCC2 activity is also downregulated by brain-derived

neurotrophic factor (BDNF) (Coull et al., 2005; Rivera et al., 2004), neuroinflammation (Coull et

al., 2003) and coincident synaptic activity (Woodin et al., 2003). These factors also result in an

increased intracellular Cl- concentration.

GABAA receptor activity also regulates neuronal excitability through shunting inhibition.

The opening of GABAA receptors increases the conductance of the membrane and reduces the

membrane time constant. Excitatory synaptic input depolarizes the neuronal membrane potential

15

in accordance with Ohm‟s law, V = IR. A reduction in membrane resistance through GABAA

receptor activity results in less membrane depolarization by synaptic input. This current shunt

reduces the ability of excitatory input to evoke an action potential and thus inhibits neuronal

activity.

Both phasic and tonic GABAA receptor activity can produce shunting inhibition (Bonin et

al., 2007; Chance et al., 2002; Mitchell and Silver, 2003; Prescott and De Koninck, 2003), but it

is arguable that tonic GABAA receptor activity is a more effective source of shunting inhibition

than phasic receptor activity. Tonic inhibition is by definition a „persistent‟ conductance. In

contrast, high frequency and temporally overlapping phasic input would be required for synaptic

GABAA receptor activity to shunt current over long periods of time. Indeed, quantitative

measurements have shown that, over time, the current generated by tonic GABAA receptor

activity is much larger than the inhibitory current generated by synaptic GABAA receptor activity

(Ataka and Gu, 2006; Bai et al., 2001; Hamann et al., 2002).

Shunting inhibition can regulate neuronal activity by changing either the minimum

amount of current required to generate an action potential, also known as the rheobase, or the

frequency at which a neuron generates action potentials for a given stimulus intensity, also called

the gain of action potential firing (Fig 1.2)(Silver, 2010). The effect of shunting inhibition on the

rheobase or gain depends on the nature of the synaptic input. This relationship was studied by

Mitchell and Silver (2003) using dynamic clamp recordings in cerebellar granule cells (Mitchell

and Silver, 2003). In this study, neurons were excited by low-noise stimuli, such as a flat

depolarizing current step, shunting inhibition increased only the rheobase but not the gain. An

increase in rheoase was seen as a rightward shift in the input-output curve, which is considered a

„subtractive‟ change (Fig 1.2). Yet, when the noise of the stimulus was increased by exciting

neurons with fluctuating synaptic excitation, shunting inhibition not only increased the rheobase

but also reduced the gain. A reduction in the gain of action potential firing was seen as a

reduction in the input-output curve, which is considered a „divisive‟ change (Fig 1.2). Shunting

inhibition can therefore induce both divisive and subtractive changes in the neuronal input-output

relationship depending on the nature of synaptic input received by the neuron. However, in

neurons with low intrinsic activity, such as seen in hippocampal pyramidal neurons (Czurko et

16

al., 1999), the increase in rheobase will likely dominate effects of shunting inhibition on network

function, since changes in gain will be less drastic at low action potential frequencies.

The activity of α5GABAA receptors may also regulate the computational properties of

neurons. α5GABAA receptor expression is upregulated in cortical neurons of mice lacking the

hyperpolarization-activated cyclic nucleotide-gated channel subtype 1 (HCN1) (Chen et al.,

2010). This finding raises the interesting possibility that α5GABAA receptors have similar

functions as HCN1 channels, including the regulation of synaptic integration and computation

(Wahl-Schott and Biel, 2009).

17

18

Figure 1.2. Subtractive and divisive inhibition of neuronal activity.

Shunting inhibition can modulate the input-output relationship of neurons through subtractive or

divisive changes in the frequency of action potential firing. Subtractive changes occur when

there is increase in the rheobase, which is the minimum amount of stimulation required to

generate action potentials. This is seen as a rightward shift in the input-output relationship.

Divisive changes occur when there is reduction in the gain, which is the frequency at which a

neuron generates action potentials for a given stimulus intensity. This is seen as a reduction in

the slope of the input-output relationship. Generally, shunting inhibition such as a tonic

GABAergic inhibition is associated with subtractive changes in neuronal activity.

19

1.3.2 Regulation of synaptic plasticity

Tonic inhibition can regulate the development of plastic changes in synaptic strength. One form

of plastic change that occurs widely throughout the CNS is the activity-dependent increase or

decrease in the strength of synaptic connections between neurons, called long-term potentiation

(LTP) and long-term depression (LTD), respectively (Bliss and Collingridge, 1993). There are

several signaling pathways and mechanisms through which LTP or LTD can be induced; the

specific mechanism that underlies the plastic changes is dependent on the neuron type and the

nature of neurotransmission involved (Malenka and Bear, 2004). A complete review of these

mechanisms is beyond the scope of this chapter but an introduction to one common mechanism

is provided here.

During LTP, there is an increase in the strength of existing synaptic connections and/or

an increase in the number of synapses (Malinow and Malenka, 2002). An increase in synaptic

strength can occur within minutes of LTP induction. This increase in strength arises from the

phosphorylation of synaptic AMPA receptors and from the insertion of additional AMPA

receptors into synapses, which leads to the an increase in AMPA receptor-generated current

(Malenka and Bear, 2004). An increase in synapse number can occur hours after the induction of

LTP and is dependent on protein synthesis.

In contrast with LTP, LTD involves a reduction in the strength of existing synapses,

and/or a reduction in the number of synapses (Beattie et al., 2000). This process essentially

mirrors the changes in LTP, and results from both the removal of AMPA receptors from the

synapse as well as the dephosphorylation of AMPA receptors (Malenka and Bear, 2004). While

LTP is often considered to be a cellular mechanism of memory formation, LTD is not necessarily

the erasure of memory and is critical for the expression of learning in some situations (Kemp and

Manahan-Vaughan, 2004).

Tonic inhibition generated by α5GABAA receptors can regulate the development of LTP

at CA1 pyramidal neuron synapses (Cheng et al., 2006b; Martin et al., 2010). CA1 pyramidal

neurons receive input from the CA3 via Schaeffer collaterals. Because CA1 pyramidal neurons

have a prominent tonic inhibitory current that is generated by α5GABAA receptors (Caraiscos et

al., 2004b), it is plausible that α5GABAA receptor activity can regulate synaptic plasticity in

20

these neurons. Yet, in hippocampal slices from WT and Gabra5−/− mice, there was no

difference between the induction or maintenance of LTP at CA1-Schaffer collateral synapses

when LTP is induced by high intensity stimulation (Cheng et al., 2006b; Collinson et al., 2002;

Martin et al., 2010). However, blocking the activity of α5GABAA receptors with L-655,708

enhanced LTP in these neurons (Atack et al., 2006). Additionally, enhancing the activity of

α5GABAA receptors with etomidate inhibited LTP (Cheng et al., 2006b; Martin et al., 2009).

Tonic inhibition generated by δGABAA receptors can also inhibit CA1 LTP, but this

phenomenon was not seen in adult animals when δGABAA expression declines (Shen et al.,

2010). The regulation of plasticity by α5GABAA receptors depends on the parameters of

stimulation. Stimulation of Schaeffer collaterals at a frequency of 10 Hz induced LTD in

hippocampal slices from WT mice but produced LTP in slices from Gabra5 −/− mice or when

α5GABAA receptor activity was blocked by L-655,708 (Martin et al., 2010).

1.3.3 Behavioural roles of tonic inhibition

One means of determining the behavioural role of tonic inhibition is to enhance the activity of

the underlying GABAA receptors and quantify the behavioural effect. Many anesthetic, sedative,

and hypnotic drugs potently enhance the activity of tonically active GABAA receptors (Bonin

and Orser, 2008; Franks, 2006; Kopp Lugli et al., 2009). It is also possible to determine the

physiological effects of tonic inhibition by reducing the activity of the select GABAA receptors

either through subtype-selective GABAA receptor antagonists or inverse agonists, or through

genetic manipulations.

Deletion of the gene for the α5GABAA receptor subunit improves the memory

performance of mice in the Morris water maze (Collinson et al., 2002). Similarly, inhibiting the

activity of α5GABAA receptors with subunit-selective inverse agonists improves the

performance of both rodents and humans in memory tasks (Atack et al., 2006; Dawson et al.,

2006; Martin et al., 2010; Nutt et al., 2007). These data strongly suggest that α5GABAA

receptors play a role in learning and memory, which might be predicted from the relatively high

expression of α5GABAA receptors in the hippocampus (Sur et al., 1999a).

21

These results suggest that the physiological role of the tonic current generated by

α5GABAA receptors is to reduce memory formation. The benefit of inhibiting memory formation

is unclear. Alternatively, α5GABAA receptors may play a subtle role in the regulation of learning

and memory that has not yet been determined. For example, tonic inhibition may enhance

memory discrimination by reducing neuronal firing rates to increase the sparseness of neuronal

activity (Hamann et al., 2002). Sparse neuronal encoding of information increases the capacity of

a network to encode semantic memory by increasing the number of distinct network states that

are possible (Olshausen and Field, 2004). Tonic inhibition may therefore act to increase memory

capacity.

The deletion of the δGABAA receptor subunit has surprisingly little behavioural

consequence, with the exception of possible abnormalities in maternal behavior (Maguire et al.,

2009; Mihalek et al., 1999). The loss of tonic inhibition generated by δGABAA receptors may be

compensated for by an increase in other forms of inhibition. The deletion of δGABAA receptors

caused a modest increase in the expression of γ2 subunit expression in the forebrain (Peng et al.,

2002). Also, deletion of the α6 subunit reduced the expression of δGABAA receptors in

cerebellar granule cells (Brickley et al., 2001). The loss of δGABAA receptors in these neurons

was functionally compensated for by an increase in persistent neuronal inhibition through

TASK-1 „leak‟ potassium channels that reduced neuronal excitability (Brickley et al., 2001).

However, a compensatory change that replaces the tonic inhibition in neurons from Gabrd−/−

has not yet been demonstrated.

The activity of δGABAA receptors in vivo is strongly regulated by neuroactive steroids.

This is well demonstrated in a pair of studies by (Maguire and Mody, 2008; Maguire et al.,2009)

that explored how maternal behaviour is regulated by neuroactive steroid production and

δGABAA receptor activity. During pregnancy, there was a large increase in the production of

progesterone, which is metabolized to the neuroactive steroid, 3α,5α-THPROG (Maguire and

Mody, 2008). However, a consistent level of neuronal inhibition was maintained through a

reduction in the expression of γGABAA receptors and δGABAA receptors. The increase in the

levels of neuroactive steroids was associated with a decrease in the expression of both δGABAA

receptors and γGABAA receptors in the hippocampus (Maguire and Mody, 2008). The reduced

GABAA receptor expression in the presence of increased neuroactive steroid production resulted

22

in a consistent level of neuronal inhibition (Maguire et al., 2009). The levels of neuroactive

steroids dropped dramatically in the post-partum period, and this decline was accompanied by a

rapid increase in the expression of δGABAA receptors. In Gabrd−/− mice and mice with a

heterozygous deletion of the δGABAA receptors (Gabrd+/−), there was an overall deficit in

neuronal inhibition in the post-partum period caused by the absent or reduced expression of

δGABAA receptors. These mice exhibited abnormal maternal post-partum behavior, resulting in

pup neglect and death (Maguire and Mody, 2008). In Gabrd+/− mice both neuronal inhibition

and maternal behavior was restored by the administration of THIP (Maguire and Mody, 2008).

Overall, these behavioural findings, along with the normal phenotype of Gabrd−/− mice suggest

that behavioural regulation by δGABAA receptors may strongly depend on changes in the

production of neuroactive steroids.

Beyond the regulation of maternal behavior, δGABAA receptor activity has been shown

to have a role in anxiety, sedation, and sleep regulation (Belelli et al., 2005; Jia et al., 2005;

Orser, 2006; Shen et al., 2007). The synthetic neuroactive steroid alphaxalone and the

endogenous steroid pregnanolone disrupted the righting reflex for shorter durations in Gabrd−/−

mice (Mihalek et al., 1999). However, midazolam, etomidate, propofol, ketamine and halothane

all produced comparable loss-of-righting reflex in Gabrd−/− mice compared to WT mice

(Mihalek et al., 1999).

δGABAA receptors of the thalamus may mediate the sedative effects of anesthetics, THIP

and ethanol (Belelli et al., 2009). The thalamus plays a key role in the regulation of sleep

(Steriade, 2005) and δGABAA receptors are highly expressed in thalamocortical neurons of the

ventrobasal (VB) thalamus (Belelli et al., 2005; Cope et al., 2005; Jia et al., 2005). Indeed, the

sedative effects of THIP were reduced in both Gabra4−/− and Gabrd−/− mice (Chandra et al.,

2006; Herd et al., 2009).

1.3.4 Interactions between α5GABAA and δGABAA receptors

The tonic inhibition generated by α5GABAA receptors and δGABAA receptors may have

overlapping or distinct physiological effects. The activity of both receptor subtypes can shunt

23

excitatory current to reduce neuronal excitability (Bonin et al., 2007; Brickley et al., 1996;

Hamann et al., 2002; Pavlov et al., 2009), and both receptors can inhibit LTP in the hippocampus

(Martin et al., 2010; Shen et al., 2010). Both GABAA receptor subtypes are expressed

concurrently in some cells such as pyramidal neurons and interneurons in the CA1 and CA3

regions of the hippocampus (Glykys et al., 2008; Scimemi et al., 2005) and in lamina II neurons

of the spinal cord (Takahashi et al., 2006).

α5GABAA receptors and δGABAA receptors differ in their biophysical and

pharmacological properties. These differences suggest that these two GABAA receptor subtypes

may serve distinct physiological roles. GABAA receptors containing a δ subunit have a higher

affinity for GABA than the γ subunit-containing α5GABAA receptors and would therefore be

more responsive to low concentrations of ambient GABA (Brown et al., 2002; Kang et al.,

2002). However, the activity of extrasynaptic α5GABAA receptors increases in response to

heightened neuronal network activity because synaptic spill-over increases the concentration of

extrasynaptic GABA (Ataka and Gu, 2006; Glykys and Mody, 2007; Scimemi et al., 2005). The

co-expression of both receptor subtypes can thus extend the range of GABA concentrations over

which a tonic current can be generated: δGABAA receptors predominantly generate the tonic

current when GABA concentrations are very low, but the contribution of α5GABAA receptor

activity to the tonic current increases as GABA concentrations increase.

Changes in the relative activity of δGABAA receptors and α5GABAA receptors will

affect the pharmacological properties of the tonic current. Neuroactive steroids will more

strongly enhance a tonic current that is predominantly generated by δGABAA receptors than a

tonic current that is primarily generated by α5GABAA receptors. Whether the relative

contribution of α5GABAA and δGABAA receptors to the tonic current have functional

consequences for neuronal excitability and network activity remains to be determined.

24

1.4 GABAA receptors as targets for analgesic drugs

1.4.1 Regulation of nociception by GABAA receptors

It has long been recognized that GABA is a key neurotransmitter in the regulation of nociception

(Enna and McCarson, 2006; Jasmin et al., 2004). GABA is highly concentrated in the dorsal

horn of the spinal cord, a region of the spinal cord that is critically involved in the processing and

transmission of noxious sensation (Miyata and Otsuka, 1975). GABAA receptors are also highly

expressed in the dorsal horn (Waldvogel et al., 1990). The administration of the GABAA receptor

agonists or positive allosteric modulators can reduce nocifensive behaviour in animal models of

acute, inflammatory, and neuropathic pain (Hammond and Drower, 1984), (Hwang and Yaksh,

1997; Knabl et al., 2009; Levy and Proudfit, 1977; Malan et al., 2002; Zambotti et al., 1991;

Zorn and Enna, 1987). Finally, the responses of dorsal horn neurons to noxious stimulation in

vitro are regulated by GABAA receptor activity (Dickenson et al., 1985; Green and Dickenson,

1997).

Given the abundant evidence that GABAA receptors modulate nociception, it is surprising

that the GABAA receptor is not a common therapeutic target for analgesia. The high incidence of

undesirable effects arising from the enhancement of GABAA receptor activity limits the clinical

use of these drugs in the treatment of pain (Brotz et al., 2010; van Tulder et al., 2003). Moreover,

benzodiazepine use was associated with worse pain control in some forms of pain such as

persistent back pain (Brotz et al., 2010). Benzodiazepines inhibit nociception in animal models,

but the degree of analgesia is strongly influenced by the pain model studied, the drug used, and

the route of drug administration (Enna and McCarson, 2006; Zeilhofer et al., 2009). For

example, the intrathecal administration of the diazepam and midazolam was more effective at

inhibiting acute thermal nociception than the systemic administration of these benzodiazepines

(Nishiyama, 2006; Zambotti et al., 1991). The intrathecal administration of midazolam inhibited

acute thermal nociception (Nishiyama, 2006), but lacked analgesic effect in the formalin assay

(Dirig and Yaksh, 1995). Additionally, the intrathecal administration of diazepam did not inhibit

acute mechanical nociception (Knabl et al., 2008).

25

Targeting specific GABAA receptors with subunit selective benzodiazepines can inhibit

nociception and minimize adverse effects such as sedation (Knabl et al., 2008; Knabl et al.,

2009). The selective targeting of benzodiazepine-sensitive GABAA receptors in the spinal cord

that contain the α2 or α3 subunit prevented hyperalgesia that develops in inflammatory models of

pain (Knabl et al., 2008; Knabl et al., 2009). However, selectively enhancing α2 and α3 GABAA

receptor activity did not reduce nociception in the absence of hyperalgesia (Knabl et al., 2008;

Knabl et al., 2009). It is therefore possible to selectively modulate some forms of nociception via

the subtype-selective upregulation of GABAA receptor activity. Yet, the contribution of many

GABAA receptor subpopulations to the regulation of pathological pain is still unclear.

A loss of neuronal inhibition is a key feature in several pain states (Coull et al., 2005;

Coull et al., 2003). Injury to the sciatic nerve causes the release of BDNF from activated

microglia neurons in the spinal cord lamina I pain pathway (Coull et al., 2005). BDNF inhibits

the activity of KCC2, resulting in the accumulation of intracellular Cl- and the collapse of the Cl

-

driving force (Coull et al., 2003). With the loss of GABAergic inhibition, there is an increase in

the output of lamina I neurons (Keller et al., 2007). Given that GABAA receptor activity is much

less inhibitory under these conditions, it is interesting that increasing the activity of α2 and

α3GABAA receptors with benzodiazepines still inhibits nociception (Knabl et al., 2009).

Additionally, nerve injury also reduces the levels of GABA and GAD65 in the spinal cord, and

this loss may further contribute to a reduction in the level of neuronal inhibition in the dorsal

horn (Moore et al., 2002).

1.4.2 Regulation of central sensitization by GABAA receptors

Neuronal inhibition generated by GABAA receptors may also critically regulate the development

of central sensitization and the development of chronic pain. Central sensitization refers to an

enhancement in the function or activity of neurons and circuits in nociceptive pathways

(Latremoliere and Woolf, 2009). This process parallels the process of LTP seen in CA1

pyramidal neurons (Ji et al., 2003). Pain hypersensitivity can be induced through the intense

26

activation of C-fiber afferents, either through the repetitive presentation of noxious electrical or

thermal stimulation (Wall and Woolf, 1984; Woolf, 1983), or through the peripheral application

of noxious chemicals such as formalin (McNamara et al., 2007), mustard oil (Jordt et al., 2004)

or capsaicin (LaMotte et al., 1991). Notably, LTP can be induced in lamina I neurons that project

to the brain by using the same stimuli that induce behavioural hypersensitivity (Sandkuhler,

2007). Similar to LTP processes at CA1-Schaeffer collateral synapses, the activation of NMDA

receptors was required for the development of both behavioural sensitization and dorsal horn

LTP (Coderre and Melzack, 1992) (Ikeda et al., 2003). Additionally, spinal LTP involves the

phosphorylation and insertion of AMPA receptors to synapses, resulting in an increase in

synaptic strength (Woolf and Salter, 2000).

GABAA receptor activity regulates the development of central sensitization in many

experimental models of pain. Blocking spinal GABAA receptors immediately after nerve injury

enhances the development of hyperalgesia (Yamamoto and Yaksh, 1993). Conversely, the

intrathecal administration of GABA permanently reverses the development of thermal

hyperalgesia and mechanical allodynia after neuropathic injury (Eaton et al., 1999). Blocking the

activity of spinal GABAA receptors during low intensity stimulation of sciatic nerve afferents

results in the development of LTP rather than LTD of dorsal horn field potentials (Garcia-Nicas

et al., 2006; Miletic and Miletic, 2001). Finally, the spinal application of bicuculline enhances

the activity of dorsal horn neurons during the sensitization process that occurred following

injection of formalin into the hind paw of rats (Green and Dickenson, 1997). These data suggest

that GABAA receptor activity constrains the development of sensitization that occurs following

nerve injury or intense activation of nociceptive C-fiber afferents.

Basal GABAergic activity in the spinal cord may constrain nociception even in the

absence of injury or the induction of sensitization. The intrathecal administration of bicuculline

produces mechanical allodynia in the absence of nerve injury (Sivilotti and Woolf, 1994).

Additionally, the expression of the neuronal transcription factor Fos in the dorsal horn, indicative

of plastic changes that indicate sensitization, is induced following blockade of basal GABAA

receptor activity (Cronin et al., 2004). These data suggest that neuronal inhibition generated by

the basal activity of GABAA receptors tonically suppresses sensitization processes in the spinal

cord. It is unclear whether this inhibition is generated by extrasynaptic GABAA receptors in the

27

dorsal horn. Reducing the expression of δGABAA receptors through deletion of either the δ- or

α4GABAA receptor subunit is not associated with any change in basal responsiveness to acute

electrical or thermal stimuli, respectively (Chandra et al., 2006; Mihalek et al., 1999). These

results suggest that the basal activity of δGABAA receptors does not regulate nociception.

1.4.3 GABAA receptors expression in pain processing pathways

GABAA receptors are broadly expressed throughout pain processing pathways. This is perhaps

unsurprising since GABAA receptors are a critical source of neuronal inhibition throughout the

CNS. Most relevant to this thesis, the expression of GABAA receptors has been well

characterized in the dorsal horn of the spinal cord (Waldvogel et al., 1990).

1.4.3.1 Presynaptic GABAA receptors in primary afferent terminals

GABAA receptors are expressed on the axonal terminals of primary afferents (Kullmann et al.,

2005). There is a high concentration of Cl- within the axonal terminals of primary afferents

(Price et al., 2005). The intracellular Cl- level is maintained through the activity of Na

+-K

+-Cl

--

cotransporter (NKCC1), which pumps Cl- into the cell (Price et al., 2005). Because of this high

intracellular concentration of Cl-, the activation of presynaptic GABAA receptors leads to

depolarization of the terminal, which reduces the release of neurotransmitters and the

transmission of sensory information (Cervero et al., 2003). This is thought to happen because

GABAA receptor-generated depolarization inactivates voltage-dependent Ca2+

channels

(VDCCs), preventing the activation of VDCCs in response to membrane depolarization caused

by action potentials (Cervero et al., 2003). Without Ca2+

, vesicular fusion cannot occur. In the

pain gating model, the activation of to