 

. 3.:
.1 .
L535...

.3.

.337...
.23....
w. s \s

Hurt?

5.14:9:

. .:l I .‘Ic‘zv’o .tov‘ds 13 II. ﬁll-Into).-
‘Ilv .
3.

mg "a”... ‘ i. .eéﬁa

It?

5

«

 

.ni
e 3:
Cat}. 2.

 

 

 

 

 

.p 2:.

It‘iuar1
7.. xIa:
.3

1.2.;
13.3,
«a... gift

. 1'9
:31
29.2..»
$2..
n. ..
( i
e

\

«.3.»

In:

a? h
11

.y n

r .r:.
V .1.)
.44!

,
i

53!...”
vi 71....

. 3‘ L
.4: :
. x. -
a bun.
, . . .
a: «so; 14
9.4.th1:

n .3
Sonic. .
2.. .. I31
, 34L

:0! E

.,,9£4;u ..
4.4 .3 :4!
Its...

 

 

)5. churn”
Wu... .3. 1 J\

 

I ‘..o.

9.00%

 

This is to certify that the
dissertation entitled

COMPARATIVE EFFECTS OF MEHG ON GABAA RECEPTOR
FUNCTION IN CELLS IN CULTURE

presented by

CHRISTINA J. HERDEN

has been accepted towards fulfillment
of the requirements for the

PhD. degree in Neuroscience

 

 

 

Major Professor’s Signature
5/30/06

Date

 

MSU is an Afﬁrmative Action/Equal Opportunity Institution

 

LIBRARY
Michigan State
University

 

 

 

-V-.-.-.-._.

 

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/ClRC/DateDuelp65-p, 15

 

 

 

—

 

COMPARATIVE EFFECTS OF MEHG ON GABAA RECEPTOR FUNCTION IN
CELLS IN CULTURE

By

Christina J. Herden

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Neuroscience Program

2006

ABSTRACT

COMPARATIVE EFFECTS OF MEHG ON GABAA RECEPTOR FUNCTION IN
CELLS IN CULTURE

By

Christina Jean Herden

MeHg is an environmental contaminant that still poses a signiﬁcant toxicological
threat. MeHg is highly neurotoxic as it readily crosses the blood brain barrier due to its
lipophilic nature. In the brain, the cerebellum is one of the primary targets for MeHg.
Interestingly, GABAergic synaptic transmission in the cerebellum is especially sensitive
to the effects of MeHg as compared with excitatory synaptic transmission. The effects of
MeHg on GABAA receptor-mediated responses are poorly understood. To characterize
the effects of MeHg on IGABA in cells in culture, whole-cell IGABA were recorded in the
presence of MeHg (0.1, l, and 10 M) in cerebral cortical cells in primary culture. IGABA
were gradually and irreversibly suppressed by MeHg in a time- and concentration-
dependent manner in cortical cells. Kinetically, MeHg prolonged [GABA slow decay rate
in cortical cells in a concentration-independent manner. To determine whether or not
MeHg interacts with the channel pore, IGABA were recorded at a range of voltages. The
effects of MeHg on [GABA were voltage-independent, suggesting that MeHg binds to the
external surface of the receptor. To determine if MeHg interacts with the GABA or
diazepam binding site of the receptor, effects of MeHg on IGABA suppression and kinetics
were examined in the absence or presence of GABA (10, 100, 500, 1000 M), or

diazepam (0.1, 1, and 10 uM) or flumazenil (10 uM), respectively. The effects of MeHg

on IGABA were not altered in the presence of GABA, diazepam, or ﬂumazenil, indicating
that MeHg does not likely interact with the GABA or diazepam site of the receptor.
Cerebellar granule cells are particularly sensitive to inhibition of GABAergic
synaptic transmission by MeHg as compared with neighboring Purkinje neurons. One
possibility for these differential effects may be the differential expression of the a
subunit, speciﬁcally the (16 subunit, in GABAA receptors in granule cells and Purkinje
cells. The presence of the (16 subunit confers unique pharmacological and kinetic
properties onto the GABAA receptor. To investigate this possibility, whole-cell IGABA
were recorded in the presence of MeHg (O. 1, 1, and 10 M) in cerebellar granule cells,
cerebral cortical cells, and in transfected human embryonic kidney (HEK-293) cells in
culture. Cortical cells were used as a replacement for Purkinje cells as they are more
successfully cultured and express the same ail-containing GABAA receptor phenotype as
do Purkinje cells. To verify the presence of speciﬁc GABAA receptor subtypes of
interest, immunocytochemistry was performed on each cell type. Findings from these
studies reveal that IGABA were suppressed more rapidly by MeHg in granule cells than in
cortical cells. However, IGABA recorded from (16 or Oil-containing granule cells did not
differ in their responsiveness to MeHg. Similarly, the effects of MeHg did not differ in
IGABA recorded from HEK-293 expressing either (16 or ail-containing receptors. These
ﬁndings suggest that the differential effects of MeHg on IGABA in cerebellar cells are not

likely due simply to the expression of the a6-containing GABAA receptor.

DEDICATION

For GABA Girl

iv

ACKNOWLEDGEMENTS

I would like to thank my parents for their love and support, and my mother for
her clerical assistance. I am also greatful to my three sisters, Corinna, Samantha, and
Jamie, for their love and friendship.

I would like to thank Dr. Atchison for his encouragement and support. I would
like to thank Dr. Yukun Yuan for his mentorship, patience, and big heart. Thanks, also to
Dr. Ravindra Hajela for his advice and expertise and help with cell transfections. I would
also like to express my thanks to my guidance committee for their time, input and
suggestions.

I greatly appreciate the friendship, empathy and support of my dear friends James
Otero and Nicole Pardo. Thanks also to Nicole Pardo and Dawn Parsell for their

assistance with immunocytochemistry.

TABLE OF CONTENTS

LIST OF TABLES ................................................................................. viii
LIST OF FIGURES .................................................................................. ix
LIST OF ABBREVIATIONS ...................................................................... xi
CHAPTER ONE: INTRODUCTION .............................................................. 1
A. Methylmercury .......................................................................... 2

B. GABAA receptor

a. Structure .......................................................................... 8
b. Function........ ..................................................................................... 11
c. Pharmacology .................................................................. 13
C. GABAA receptors in the cerebellum ................................................. 21

D. MeHg neurotoxicity

a. Cellular effects of MeHg ...................................................... 30
b. Disruption of synaptic transmission by Mehg ............................. 32
E. Disruption of GABAergic function by MeHg ....................................... 34

CHAPTER TWO: EFFECTS OF METHYLMERCURY ON DIAZEPAM-MEDIATED
ENHANCEMENT OF GABAA RECEPTOR CURRENTS IN RAT CORTICAL
NEURONS IN CULTURE

A. Abstract .................................................................................. 42
B. Introduction ............................................................................. 43
C. Methods ................................................................................. 46
D. Results ................................................................................... 51
E. Discussion ............................................................................... 79

vi

 

CHAPTER THREE: DIFFERENTIAL EFFECTS OF MEHG ON GABAA RECPTOR
CURRENTS IN CEREBELLAR GRAN ULE AND CEREBRAL CORTICAL CELLS IN
CUTLURE

A. Abstract .................................................................................. 85
B. Introduction ............................................................................. 86
C. Methods ................................................................................. 88
D. Results ................................................................................... 94
E. Discussion .............................................................................. 133

CHAPTER FOUR: SUMMARY AND DISCUSSION
A. Summary of research. . . . . . .. ........................................................................... 139
B. Relationship to previous work ...................................................... 143

C. Possible mechanisms underlying the differential effects of MeHg on cerebellar
granule and Purkinje neurons ............................................................ 147

D. Potential mechanisms of interaction between MeHg and the GABAA

Receptor .................................................................................... 158

E. Conclusion ................................................................................ 162
APPENDIX

A. Suppression of [GABA by MeHg in granule cells is not reversible ............... 165

B. MeHg does not compete with GABA for its binding site on the GABAA
receptor ..................................................................................... 168

BIBLIOGRAPHY ................................................................................. 17 1

vii

 

LIST OF TABLES

CHAPTER ONE
1.1. Subunit speciﬁc actions of GABAA receptor agonists and
antagonists .................................................................................... 16

1.2. GABAA receptor subunit subtypes expressed by different cells in the

cerebellum .............................................................................................................. 27
CHAPTER THREE
3.1. Comparative effects of MeHg on [GABA suppression in on- or GIG—containing HEK-
293 cells .................................................................................... 130

viii

LIST OF FIGURES

CHAPTER ONE
1.1. A schematic representation of the GABAA receptor and its subunits ............... 10
1.2. Diagram depicting the general circuitry of the cerebellum ........................... 23
CHAPTER TWO
2.1. MeHg causes a gradual suppression of bicuculline-sensitive IGABA in cortical
cells .......................................................................................... 53
2.2. MeHg causes a complete and irreversible suppression of IGABA in cortical
cells .......................................................................................... 55
2.3. Suppression of IGABA by MeHg in cortical cells is time- and concentration-
dependent .................................................................................... 57
2.4. MeHg prolongs [GABA slow decay constant in a concentration-independent
manner. ...................................................................................... 60
2.5. MeHg affects the time course of prolongation of IGABA slow decay in a
concentration—dependent manner ........................................................ 62
2.6. Diazepam prolongs IGABA slow decay constant in a concentration-dependent
manner ........................................................................................ 65
2.7. Co-application of MeHg and diazepam does not affect MeHg-induced
prolongation of [GABA slow decay ......................................................... 68
2.8. Time course of IGABA suppression by MeHg in cortical cells is not altered by a
pplication of diazepam or ﬂumazenil .................................................... 70
2.9. Diazepam or MeHg pretreatment, following co-application, does not prevent
diazepam- or MeHg-induced prolongation of IGABA slow decay .................... 73
2.10. Co-application of MeHg and flumazenil does not prevent MeHg-induced
prolongation of IGABA slow decay ....................................................... 76
CHAPTER THREE
3.1. GABAA receptor (16 and a1 subunit expression in cerebellar granule and cerebral

cortical cells in culture .................................................................... 96

ix

3.2. MeHg causes a time- and concentration-dependent suppression of IGABA in
granule cells ................................................................................. 98

3.3. MeHg causes a time- and concentration-dependent suppression of [GABA in
cortical cells ............................................................................... 101

3.4. Effects of MeHg on IGABA are voltage—independent in granule and cortical cells.

.............................................................................................. 105
3.5. GABAA receptor (16 and (11 subunit expression in cerebellar granule cells at

different days in culture .................................................................. 107
3.6. Effects of MeHg on granule cells at different days in culture ...................... 110
3.7. GABAA receptor (16 and a1 subunit expression in HEK-293 cells .................. 115
3.8. GABAA receptor membrane localization in (11 subunit-containing HEK-293

cells ......................................................................................... 117
3.9. GABAA receptor membrane localization in (16 subunit-containing HEK-293

cells ......................................................................................... 119
3.10. Control Secondary antibody staining in HEK-293 cells ............................ 121
3.11. Effects of diazepam on IGABA in HEK-293 cells ..................................... 123
3.12. MeHg causes a time- and concentration-dependent suppression of IGABA in HEK-

293 cells .................................................................................... 127
3.13. Effects of MeHg on IGABA are voltage-independent in HEK-293 cells ............ 132
CHAPTER FOUR
4.1. Drawing showing a cerebellar glomerulus ............................................ 150

4.2. Diagram depicting the proposed mechanism of interaction of MeHg with GABAA

receptors of granule cells to alter cerebellar excitation .............................. 153
APPENDIX ,
A.l. Effects of MeHg on IGABA are irreversible in granule cells in culture ............. 167
A.2. MeHg does not interact with the GABA site on the GABAA receptor ............ 170
BIBLIOGRAPHY ............................................................................................................ 171

AMPA
APV
Ara-C
ATP
CNQX
CNS
DMEM
DMSO
DNase I
D-pen
EGTA
EPSC
EPSP

FITC

GABA
GFP

hr
HEK-293

HEPES

IGABA

LIST OF ABBREVIATIONS

a-amino—3-hydroxy-5-methylisoxazole-4— propionic acid
DL-2-amino-5-phosphonopentanoic acid

cytocine B-D-arabino-furanoside

adenosine 5’- triphosphate magnesium salt
6-cyano-7-nitroquinoxaline-2,3-dione disodium salt
central nervous system

Dulbecco’s Modiﬁed Eagle’s Medium

dimethyl sulphoxide

deoxyribonuclease I

D-penicillamine

ethylene glycol-bis(B-aminoethyl ether)-N, N, N', N',-tetraacetic acid
excitatory postsynaptic current

excitatory postsynaptic potential

ﬂuorescein isothiocyanate

gram

y-amino-n-butyric acid

green ﬂuorescence protein

hour

human embryonic kidney cells
4-(2-hydroxyethyl)piperazine- 1 -ethanesulfonic acid

GABAA receptor-mediated current

xi

 

1P3
IPSC
IPSP

MeHg

mV

N MDA

PBS

TRITC

inositol triphosphate

inhibitory postsynaptic current
inhibitory postsynaptic potential
methylmercury

minutes

millivolts

n-mehthyl-d-aspartate
picoamperes

phosphate buffered saline
seconds

tetrarnethylrhodamine

xii

CHAPTER ONE

INTRODUCTION

I

A. Methylmercury

Methylmercury (MeHg) is a ubiquitous environmental contaminant that still poses
a signiﬁcant toxicological threat, particularly in the Great Lakes, where levels of MeHg
in ﬁsh have been found to exceed the estimated dose for adverse health effects in humans
(Rice, 1995; Gerstenberger and Dellinger 2002; Gilbertson, 2004; Weis, 2004). MeHg
toxicity is not limited to the United States; it occurs in populations with high ﬁsh intake
all around the world including, Canada (Mckeown-Eyssen et al., 1983a,b; Forsyth et al.,
2004), the Faroe Islands (Steurwald et al., 2000), the Seychelle Islands (Davidson et al.,
1995, 1998; Shamlaye et al., 1995; Axtell et al., 2000), Peru (Marsh et al., 1995a, 1995b),
Brazil (Kehrig et al., 1998; Grandjean et al., 1999), New Zealand (Crump et al., 1998),
Madeira (Murata et al., 1999b), Greenland (Weihe et al., 2002), French Guiana (Cordier
et al., 2002), France (Claisse et al., 2001), and others. In fact, MeHg was listed by the
International Register for Potentially Toxic Chemicals, part of the World Health
Organization (WHO), as one of the six most dangerous chemicals in the world’s
environment (WHO, 1993). Environmental contamination by MeHg occurs either
through MeHg itself or by conversion of elemental Hg into MeHg. Industries such as
agriculture, paper, lumber, and leather have used MeHg as a preservative because of its
antifungal properties. As a result, contamination has been found in agricultural runoff
and waste water discharge from these industries. Additionally, elemental Hg, which is
used by industries that manufacture electrical equipment, paint, and chloralkali, is vented
directly into the atmosphere or is discharged in waste-water. Furthermore, Hg is released
into the environment by erosion of rock containing Hg or by human activities such as

combustion of fossil fuels. Importantly, aquatic microorganisms in lakes, rivers, and

oceans that receive Hg waste can methylate inorganic Hg, thus converting it to MeHg
(Wood et al., 1968). MeHg tends to bioaccumulate in the aquatic food chain, resulting in
high concentrations in ﬁsh at the top of the food chain, such as tuna and shark
(Toxicological Effects of Methylmercury, 2000). According to the Environmental
Protection Agency (EPA), most ﬁsh that live in US waters have MeHg concentrations of
less than 0.5 parts per million (ppm), but ﬁsh at the top of the food chain may contain
more than 1 ppm (the average concentration in ﬁsh is approximately 0.25 ppm).

The most common route of human exposure to MeHg is through consumption of
ﬁsh and other seafood taken from contaminated waters. Frequent consumption of
MeHg-contaminated food and/or water can result in signiﬁcant accumulation of MeHg in
the body because MeHg has a long half-life in humans (approximately 70 days) and a
low rate of urinary excretion (Magos, 1975). Once consumed, MeHg is transported to
most tissues in the body in plasma and red blood cells and is concentrated in the blood
and the central nervous system (Chang et al., 1972). The current reference dose for
MeHg has been set by the EPA at 0.1 pg/ kg/ day. Taking into consideration the average
MeHg concentration of ﬁsh (0.25 ppm) (EPA, 1997), the reference dose of MeHg
translates into approximately 27 g of ﬁsh per day, which is equivalent to only a few
meals of ﬁsh per week.

The ﬁrst description of the adverse effects of MeHg demonstrated that it produced
sensory and motor disturbances such as impairment of vision, hearing, numbness, and
ataxia, and eventually death (Hunter et al., 1940). Histological examination of the
deceased victim revealed that the brain was a primary target site for MeHg, particularly

the cerebellum (Hunter and Russell, 1945). Within the cerebellum, granule cells were

shown to be especially sensitive to MeHg-induced atrophy as compared to other
surrounding cell types. Several years later, mass episodes of acute and chronic MeHg
poisoning occurred. In the 1950’s, chronic exposure to MeHg in Japan resulted from the
contamination of Minamata and Nigata Bays by local industrial mercurial waste. Over a
period of years, local inhabitants ingested large amounts of ﬁsh and shell-ﬁsh taken from
the Bay. As a result, at least 1500 poisonings and 46 deaths occurred in individuals of all
ages (Takeuchi et al., 1962, 1968, 1970). In contaminated areas, levels of MeHg reached
50 — 85 ppm in ﬁsh and other seafoods. Over 2,000 inhabitants of the Minamata Bay
area that had been exposed to MeHg were found to have neurological defects. Initially,
exposure victims demonstrated slight motor disturbances such as abnormalities in gait
and weakness of lower limbs, resulting in uncoordinated and unstable gait (Takeuchi et
al., 1968, 1989; Takeuchi and Eto, 1999). With longer exposure time to MeHg, however,
victims experienced progressive worsening of symptoms, including visual disturbances,
loss of hearing, and ataxia. The gradual worsening of ataxic symptoms was associated
with progressive insult to cerebellar-based motor pathways, and the onset of ataxia was
closely correlated with granule cell damage in the cerebellum (Takeuchi and Eto, 1999).

In 1972, a mass episode of acute MeHg contamination occurred in Iraq.
Homemade bread was made from seed grain that had been contaminated with MeHg-
based fungicide and eaten by inhabitants of the community. As a result, 450 individuals
lost their lives and over 6500 victims were hospitalized (Bakir et al., 1973). Blood Hg
concentrations in victims in 17 patients ranged from 3-4 mg/ml. At these concentrations
of MeHg signs and symptoms of ataxia were observed and visual and hearing

disturbances were reported. In the majority of pathological investigations of victims

intoxicated with MeHg in these episodes, substantial atrophy of the cerebellum was
observed, including extensive thinning of the gray matter. In the cerebellar cortex, a
severe disintegration of granule cells was detected, but surprisingly, the neighboring
Purkinje cells were relatively resistant to MeHg-induced cell atrophy. Poisoning with
MeHg compounds can occur with even very limited acute effects. At Dartmouth
University in 1997, a chemistry professor who researched the toxicity of heavy metals,
accidentally spilled several drops of dimethylmercury onto one of her latex glove-
covered hands (N ierenberg et al., 1998). Within days after exposure, she began to
experience signs such as weight loss and diarrhea, and weeks after exposure, she
experienced neurologic symptoms such as difﬁculty with speech and abnormal hearing,
vision, and gait. Ultimately, three months after exposure, she entered into a vegetative
state, followed by death. At autopsy, the cerebellum showed diffuse atrophy, including a
widespread loss of granule cells, Purkinje cells, and parallel ﬁbers in the molecular layer.
It is noteworthy to mention that the developing nervous system is particularly
sensitive to the effects of MeHg (Bakir et al., 1973; Grandjean etal., 1997, 1998), and
that there are considerable differences in the distribution of MeHg-induced pathological
changes in young brains as compared with adult brains (Takeuchi et al., 1978). In adults,
the brain damage induced by MeHg is speciﬁc to loss of neurons in the visual cortex and
granule cells of the cerebellum. Purkinje cells of the cerebellum are unaffected.
However, in the developing brain, a more generalized and extensive distribution pattern
of injury is observed, such as loss of neurons in each lobe of the brain (Choi, 1989). This
is due to the fact that MeHg affects many developmental processes including formation

of microtubules (Miura and Imura, 1989; Muira et al., 1999), which leads to disruption of

neuronal migration, cell division, and synaptogenesis (Castoldi et al., 2001, Clarkson,
1993). In addition, MeHg can disrupt protein, DNA, and RNA synthesis and
neurotransmitter production, secretion and uptake, as well as cell-signaling (Castoldi et
al., 2001). Disruption of any of these processes can lead to disruption of cerebellar
organization (Choi et al., 1978).

Many differences exist between cerebellar granule and Purkinje neurons. Thus
there are numerous possible explanations for the differences in neurotoxicity to MeHg
observed in these cells. First, there is the disparity in cell soma size. Purkinje cells of the
cerebellum are much larger (~ 25-30 pm) than granule cells (~ 5-8 pm), thus granule cells
may simply accumulate a greater concentration of MeHg. However, Purkinje cells have
been shown to accumulate as much or more MeHg as compared with granule neurons
(Leyshon-Sorland and Morgan, 1994). Another difference observed between cerebellar
granule cells is their differential expression of potential protective MeHg-sequestering
mechanisms. Purkinje cells, but not granule cells, were shown to synthesize
metallothionein, a protein that is able to sequester mercury (Leyshon-Sorland et al.,
1994). In addition, Purkinje cells have a greater ability to buffer Ca”, as they contain
calbindin-D29K, a Ca+2 binding protein, whereas granule cells do not (Celio, 1990).
Granule and Purkinje cells also show a differential expression of certain types of ion
channels including voltage-gated Ca+2 channels (Mintz et al., 1992; Randall and Tsien,
1995), N-methyl-D-aspartate (NMDA) receptors (Didier etal., 1997), AMPA/kainate
receptors (Keinanen et al., 1990), and GABAA receptors (Takayama and Inoue, 2004).
Most pertinent to the current investigation is the differential expression of GABAA

receptor subtype in cerebellar granule and Purkinje neurons. Cerebellar Purkinje cells

express an subunit—containing GABAA receptors, whereas granule cells express the (16-
containing receptors in addition to a. subunit-containing ones (Wisden et al., 1996; Poltl
et al., 2003).

In the cerebellar slice preparation, GABAergic inhibitory synaptic transmission
appears particularly sensitive to the inhibitory effects of MeHg, especially in granule
cells as compared with Purkinje cells (Yuan and Atchison, 2003). Also, compared to
other targets such as voltage-gated Ca+2 and K+ channels and inwardly rectifying K+
channels, GABAA receptor currents are reduced most rapidly by MeHg (Yuan et al.,
2005). The exact mechanisms by which MeHg differentially interferes with GABAergic
function in granule cells and Purkinje neurons are not known. One possibility for this
differential sensitivity to MeHg may be the fact that granule cells express a unique type
of GABAA receptor that contains the (16 subunit. This receptor subtype is especially
sensitive to inhibition by heavy metals such as La+3 and Zn”, and studies have
demonstrated that MeHg enhances levels of intracellular Ca+2 and Zn+2 at the
concentrations used to inhibit GABAergic function (Saraﬁan, 1993; Denny et al., 1993;
Hare et al., 1993; Hare and Atchison, 1995). Interestingly, no studies comparing the
differential effects of MeHg on subunit subtype-speciﬁc populations of GABAA receptors
found in the cerebellum have been conducted. This thesis was undertaken to enhance
further our understanding of the differential effects of MeHg on the function of different
a-subunit containing GABAA receptors found in either granule or Purkinje cells of the
cerebellum. Speciﬁcally, it was performed to 1) characterize the effects of MeHg on

GABAA receptor-mediated function 2) to determine if MeHg differentially affects the

"1

F

functional properties of GABAA receptors in (16 subunit-containing and non-a6 subunit-

containing cells.

. GABAA Receptor
a. Structure

GABAA receptors are members of the Cys-loop family of receptors which include
the nicotinic acetylcholine, glycine, and serotonin 5-HT3 receptors (Karlin and Akabas,
1995, Le Novere et al., 2002). All subunits in this receptor family have a disulﬁde-linked
loop in the extracellular domain consisting of 15 residues. GABAA receptors are ligand-
gated Cl' ion channels that mediate the majority of fast inhibitory synaptic transmission
in the mammalian central nervous system. Approximately 20-50% of all neuronal
synapses use GABA as a neurotransmitter (Bloom and Iversen, 1971). GABAA receptors
are 275 kDa heteropentameric glycoproteins composed of different subunit families and
their subtypes (cl-(16, [Bl-B3, 71-73, 5, e, p, 1:) (Olsen and Tobin, 1990, Wisden and
Seeburg, 1992; Sieghart, 1995; Hevers and Luddens, 1998) (Fig. 1.1). The exact
stoichiometry of GABAA receptors in their native environment is unknown, but based on
studies in recombinant receptors, the GABAA receptor stoichiometry appears to consist of
two a, two [3, and one y subunit. One phenotype of GABAA receptor predominates in
most regions of the brain (on, 02/3, 72) (Chang et al., 1996). Each subunit has a similar
overall topology consisting of four transmembrane domains (M1, M2, M3, and M4),
which are responsible for forming the ion-conducting pore (Schoﬁeld et al., 1987; Karlin
and Akabas, 1995). Furthermore, each subunit has a large extracellular N-terminus and a

smaller C-terminus, a small loop between domain M2 and M3, and an intracellular loop

Figure 1.1. A schematic representation of the GABAA receptor and its subunits. The (11,
[32/3, and y; phenotype of receptor predominates in most regions of the brain. The binding
site for GABA is located at the interface between the o. and 0 subunits. Each subunit is
comprised of four transmembrane domains (M l-M4), a large extracellular N-terminus

and a smaller C-terminus.

 

   

Intracellular

Figure 1.1

10

 

between M3 and M4 (Schoﬁeld et al., 1987; Karlin and Akabas, 1995). The extracellular
N-terminus is believed to form the agonist binding sites and the loop between M3 and
M4 has been shown to be a site for regulation by phosphorylation and for localization at
synapses (Brandon et al., 2002; Moss and Smart, 1996). The loop between domain M2
and M3 has been implicated in the signal transduction process, speciﬁcally in coupling
conformational changes between the two domains and regulation of channel gating

(Grosman et at., 2000; Beta et al., 2002).

b. Function

GABA is released from presynaptic terminals into the synaptic cleft and reaches
concentrations between 500 - 1000 M in less than 1 ms (Jones and Westbrook, 1995).
GABA concentrations in the cleft decay rapidly (within ms) as a result of diffusion and
reuptake by the presynaptic terminal. Postsynaptically, GABA binds to its receptor at the
interface between the extracellular B and C loops of the [3 subunit and the D loop of the a
subunit (Galzi at al., 1994). Two GABA molecules must bind to trigger a rotation of the
extracellular domains of these receptor subunits, resulting in opening of the channel pore
(Sakmann et al., 1983). However, GABA binding can also result in the entrance of the
receptor into a desensitized state, during which two GABA molecules are bound but the
channel is closed (MacDonald et al., 1989a; Twyman et al., 1990; Twyman and
MacDonald, 1992; Jones and Westbrook; 1995). From the desensitized state, unbinding
of the ligand from the receptor, or deactivation, is slowed because it must backtrack
through each of the preceding states. The number and order of occurrence of open,

closed, and desensitized states of the GABAA receptor channel can vary depending upon

11

which binding model is considered. One model of single channel GABAA receptor
activity at the main conductance level, for instance, incorporates a reaction scheme that
involves two sequential binding sites, three open states, ten closed states, and one
desensitized state (MacDonald et al., 1989a; Twyman et al., 1990; Twyman and
MacDonald, 1992).

GABAA receptor chloride channels open to at least three conductance states (27-
30 pS, 17-19 pS, and 11-12 pS) (MacDonald et al., 1989a). However, most (95%) of the
Cl' current though the channel ﬂows through the highest conductance state (27-30 pS).
At the highest conductance state, two types of GABAA receptor channel openings occur:
brief openings (~ 2.5 ms) and bursts of openings (longer openings interrupted by brief
periods of closure). Bursts of openings are thought to represent fluctuations of the
receptor between the double-1i ganded open and double-liganded closed state. Silent
periods between single brief openings and bursts are believed to correspond to a double-
liganded closed, or desensitized, state of the receptor (Hammond, 2001). Desensitization
of the GABAA receptor increases with increased GABA concentrations and is voltage-
dependent, increasing its rate at more positive potentials (Mellor and Randall, 1998).

The typical GABAA receptor-mediated whole-cell response is characterized by
rapid activation followed by a fast and a slow decay component (Hamill et al., 1983;
MacDonald et al., 1989a). The fast decay component corresponds to the desensitization
of receptors during prolonged GABA exposure, resulting in a decrease in the frequency
of channel opening. The later slow decay phase corresponds to the deactivation, or
GABA unbinding, of receptors. Functional properties of GABAA receptor binding,

desensitization, and deactivation are dependent upon subunit subtype composition of the

12

receptor, in addition to other factors such as protein phosphorylation, voltage, and ligand
concentration (MacDonald and Olson, 1994).

Both the rate and extent of GABAA receptor desensitization are dependent on
subunit composition. Studies using transfected cells expressing different recombinant
GABAA receptors found that receptors containing the (16 subunit (along with 0272
subunits) deactivated more slowly as compared to those receptors containing the at
subunit (Tia et al., 1996). Furthermore, prolonged GABA application produced
desensitization of (110272 receptors that was characterized by a fast exponential decay and
a slow decay component. In contrast, GABAA receptors expressing the (:6 subunit did not
readily desensitize. Cells cotransfected with on and (:6 subunits (along with [5272 subunits)
showed unique deactivation and desensitization kinetics, distinct from those of receptors
containing only one subtype of a. subunit (Tia et al., 1996). Furthermore, studies using
GABAA receptor a] subunit deﬁcient mice revealed the presence of slower current decay
rates in these "(11 knockout" animals, suggesting that the a; subunit is responsible for fast
decay rates of inhibitory synaptic currents (Vicini et al., 2001). Thus, it appears that the
GABAA receptor a subunit plays a critical role in receptor desensitization and

deactivation.

c. Pharmacology

GABAA receptors have drug-binding sites for GABA as well as agonists such as
benzodiazepines, barbiturates, and general anesthetics (Olsen, 1981, 1987; Olsen et al.,
1986). GABAergic function can also be modulated by neurosteroids (Callachan et al.,

1987; Puia et al., 1990, 1993, 2003; Zhu et al., 1996) and La+3 (Ma and Narahashi,

13

1993a) and antagonized by agents such as picrotoxin, bicuculline, furosemide, penicillin,
and Zn+2 (MacDonald and Barker, 1978; Westbrook and Mayer, 1987; Smart and
Constanti, 1990; Twyman et al., 1992; Korpi and Luddens, 1993; Barberis et al., 2000;
Kolbaev et al., 2002). The afﬁnity of GABAA receptors for these agents is highly
dependent upon the subunit composition of the receptor (see Table 1.1). As numerous
drugs act at the GABAAreceptor, only those compounds that are relevant to the present
thesis will be discussed.

GABA binds to the GABAA receptor to regulate the opening or closing, also
known as gating, of the chloride ion channel (Sakmann et al., 1983). Functional GABAA
receptors are formed following expression of only the a and [3 subunits (Baumann et al.,
2001). The nature of the a subunit appears to affect the afﬁnity of the receptor to GABA.
Receptors expressing a6 subunits have a much higher afﬁnity for GABA than do
receptors containing (1; subunits (Fisher et al., 1997).

Benzodiazepines, such as diazepam, are compounds that produce a range of
sedative, anti-convulsant, and/or anxiolytic actions (Bellantuono et al., 1980).
Benzodiazepines potentiate GABAergic responses at submaximal concentrations of
GABA (Study and Barker, 1981). Single-channel analyses of the actions of
benzodiazepines have revealed that they enhance GABAergic responses by increasing
open and burst frequencies without altering channel conductance or open or burst
duration (Study and Barker, 1981). Enhanced open frequency results from an increase in
the probability of Cl' channel opening in response to GABA by increasing the binding
rate of the GABA molecule to the receptor (Costa and Guidotti, 1979; Study and Barker,

1981; Vicini et al., 1987; Sieghart, 1992). Furthermore, a predominant effect of

14

Table 1.1. Subunit speciﬁc actions of GABAA receptor agonists and antagonists.
Excitatory actions of agents on speciﬁc receptor subunits are indicated by a plus (+)
symbol. Enhanced sensitivity to excitation is indicated by a double plus (++) symbol.
Inhibitory responses are indicated by a minus (-) symbol. Enhanced sensitivity to
inhibition is indicated by a double minus (--) symbol. Insensitivity of the subunit to the

agent is depicted by a (0) symbol.

15

 

 

 

 

 

 

 

 

Drug or Drug GABAA Receptor Subunit
Group
0‘1 Cl<5 7
+ 0
Benzodiazepines
+ ++ + 0
Barbiturates
+ ++ + 0
Furosemide
- + + 0
Lanthanum
Zinc

 

 

 

 

 

Table 1.1

16

 

benzodiazepines on GABAergic responses is a prolongation of the time course by
slowing of decay (Segal and Barker, 1984; Vicini et al., 1986; Zhang et al., 1993). The
slowing of decay by benzodiazepines is believed to be due to slowing of current
deactivation and speeding up of current desensitization (Mellor and Randall, 1997). The
actions of benzodiazepines at the GABAA receptor are primarily dependent on a and 7
subunit combinations (Luddens and Wisden, 1991). The presence of the 72 subunit
confers benzodiazepine sensitivity onto the receptor, and the a subunit, when expressed
with the B and 72 subunits, affects the efﬁcacy and afﬁnity of the GABAA receptor for
benzodiazepines (Pritchett et al., 1989). For example, GABAA receptors containing the
0.6 subunit bind the inverse-agonist ligand [3H]Ro 15-4513, but do not bind
benzodiazepines such as diazepam (W ieland et al., 1992). In contrast, the function of
GABAA receptors containing the on subunit is potentiated by benzodiazepines (Makela et
al., 1997).

Barbiturates, such as pentobarbital, are usually anesthetic compounds. Their
effects on GABAergic responses vary depending on the concentration at which they are
applied. At low concentrations, barbiturates enhance GABA-mediated chloride
conductance by increasing the mean channel open time (Nicoll and Wojtowicz, 1980;
Schulz and MacDonald, 1981). At higher concentrations, barbiturates open the GABAA
receptor directly, independent of the presence of GABA (Nicoll and Wojtowicz, 1980;
Schulz and MacDonald, 1981; Thompson et al., 1996). At very high (millimolar)
concentrations, they inhibit GABAA receptor function, most likely by blocking the open
channel (Akk and Steinbach, 2000). Barbiturate sensitivity is dependent on the GABAA

receptor [3 subunit. Receptor mutation studies have shown that [33 subunit-containing

17

GABAA receptors are insensitive to pentobarbital, whereas [31 subunit—containing
receptors are sensitive (Cestari et al., 2000). Moreover, the a subunit affects the
sensitivity of the GABAA receptor to barbiturates. Receptors containing the (16 subunit
are about three times more responsive to pentobarbital activation than are non (1‘ subunit-
containing receptors (Fisher et al., 1997).

Steroids, particularly neurosteroids such as pregnanalone, enhance the actions of
GABA at the GABAA receptor by prolonging the open time and burst duration of the
chloride channel and by enhancing the binding afﬁnity of receptor agonists such as
muscimol or benzodiazepines (Barker et al., 1987; Callachan et al., 1997a). The steroids
androsterone and pregnanolone have been shown to increase channel open and burst
durations (Twyman and MacDonald, 1992). The prolonged open duration is due to an
increase in open frequency and an increase in frequency of occurrence of longer open
states. Increased burst duration is concentration-dependent and results in an increase in
proportion of bursts with longer durations. The mechanism for prolongation of open and
burst durations of IGABA is similar to that described for barbiturates, suggesting that both
may regulate the GABAA receptor channel through at least one common mechanism
(MacDonald et al., 1989b). There is no absolute subunit speciﬁcity for neurosteroid
modulation of GABAA receptors, however the actions of neurosteroids seem to be
particularly evident in receptors containing a, y, or 6 subunits (Puia et al., 1990, 1993;
Korpi and Luddens, 1993; Zhu et al., 1996). GABAA receptors that contain the 6 subunit,
for instance, have been shown to have reduced sensitivity to neurosteroids (Zhu et al.,

1996; Vicini et al., 2002)

18

La”, a rare earth metal, potentiates GABAA receptor-mediated currents (Ma and
Narahashi, 1993a). It does so by increasing the sensitivity of the receptor for GABA
(Narahashi et al., 1994) and by increasing the channel open duration (Zhu et al., 1998).
The exact molecular mechanism(s) by which La"3 affects GABAA receptor-mediated
currents in not known. The modulation of GABAergic responses by La+3 is dependent
greatly on the presence of the a subunit. La“3 potentiates a103yzL receptor currents
(Saxena et al., 1997), whereas 06B3Y2L currents are only weakly potentiated. In addition,
(161336 receptor currents are strongly inhibited by La+3 (Saxena et al., 1997). Furthermore,
in autoradiographic binding assays of a6 subunit knockout granule cells, GABA-
antagonistic stimulation of the binding by La+3 was abolished (Makela et al., 1997).

Furosemide is a loop diuretic which acts to reduce the amount of water in the
body by acting on the kidneys to increase the ﬂow of urine. Furosemide non-
competitively antagonizes GABA currents (Makela et al., 1999). Competing evidence
exists for location of the binding site for furosemide on the GABAA receptor. Kolbaev et
al., (2002) demonstrated that furosemide interacted with a site thought to be inside the
channel pore of the receptor, resulting in an open-channel block. However, other
evidence suggests that the main structural domain for furosemide antagonism resides on a
speciﬁc 0.6 subunit domain (J ackel etal., 1998). Interestingly, the effects of furosemide
are unique to GABAA receptors containing (16 or (:4 subunits, in combination with [32/3 and
72 subunits (Korpi et al., 1995). Furosemide rapidly and reversibly antagonizes
GABAergic responses of recombinant 05B2/3Y2 receptors expressed in Xenopus oocytes,
however GABAA receptors containing (110272 subunits are insensitive to the inhibitory

effects of furosemide (Korpi et al., 1995).

19

Zn“2 slows the onset of GABAA responses by decreasing the rate of GABA
binding and the rates of gating transitions from open to closed states (Barberis et al.,
2000). Additionally, Zn+2 enhances the rate of deactivation of GABAA receptor currents
by increasing the unbinding rate of GABA (Barberis et al., 2000). The effects of Zn+2 are
independent of voltage or GABA concentration, indicating that Zn+2 binds to an external
site of the receptor distinct from the GABA binding site (Narahashi et al., 1994).
Speciﬁcally, Nagaya and MacDonald (2001) revealed that the extracellular N-terminus of
the 72L subunit is a major determinant of Zn+2 sensitivity by using 72/6 subunit chimeras.
Sensitivity of GABAA receptor currents to Zn+2 is highly dependent on subunit
composition. The (16036 GABAA receptor currents are the most sensitive to inhibition by
Zn”, whereas replacement of the 6 subunit by a 7 subunit reduces the sensitivity to
inhibition by Zn+2 (Knoﬂach et al., 1996; Saxena and MacDonald, 1996; Makela etal.,
1997). Furthermore, sensitivity to Zn+2 inhibition is reduced when the 06 subunit of the
GABAA receptor is replaced by the al subunit (Draguhn etal., 1990; Saxena and
MacDonald, 1994). Along the same lines, studies using granule cells and Purkinje
neurons in primary culture revealed that GABAA receptors of Purkinje cells, which
contain the on subunit receptor, were signiﬁcantly less sensitive to the inhibitory effects
of Zn”, as compared to granule cells, which have (16 subunit containing receptors
(Zempel and Steinbach, 1995).

Finally, picrotoxin, a plant-derived convulsant, inhibits GABAA receptor currents
by a reduction in the channel's average open and burst durations (Olsen, 1981; Olsen and
Tobin, 1990). It does this via binding to a site inside the channel pore, preventing either

the Cl' channel from opening, or producing a physical block of the channel. Similarly,

20

penicillin, which can also be a convulsant, inhibits GABAA responses by blocking the
open-channel, resulting in a reduction in channel open duration and increase in average
burst duration (MacDonald and Barker, 1978; Twyman et al., 1992).

In sum, GABAA receptors containing the (16 subunit are particularly sensitive to
GABA and pentobarbital activation, insensitive to diazepam potentiation, and especially
sensitive to inhibition by Zn+2 and Ln“3 (Draguhn et al., 1990; Luddens et al., 1990;
Knoﬂach et al., 1996; Saxena and MacDonald, 1996; Makela et al., 1997; Zhu et al.,
1998; Makela et al., 1999). In contrast, GABAA receptors containing the (:1 subunit are
potentiated by benzodiazepines and La”, and are relatively insensitive to inhibition by
Zn+2 and furosemide. Additionally, the on subunit containing receptor has a lower

efﬁcacy and afﬁnity for pentobarbital (Fisher et al., 1997).

C. GABAA Receptors in the Cerebellum

Cerebellar cortex consists of three layers (molecular, Purkinje, and granule) that
contain only ﬁve types of neurons - the granule, Purkinje, Golgi, stellate, and basket
cells. The synaptic circuitry between cell types is well-deﬁned (Wisden et al., 1996)
(Fig. 1.2). In the granular layer, granule cells receive excitatory glutamatergic input from
the mossy ﬁbers, which originate from a variety of brain stem nuclei that receive inputs
from the cerebral cortex and spinal cord. Granule cells receive inhibitory GABAergic
inputs from Golgi cells. In the molecular layer, granule cell axons, or parallel ﬁbers,
form excitatory connections with Purkinje cells and local GABAergic intemeurons (i.e.

basket and stellate cells). These intemeurons form inhibitory synapses with Purkinje

21

Figure 1.2. Diagram depicting the general circuitry of the cerebellum. Five types of
neurons are organized into three layers in the cerebellar cortex: granule cells (GC),
Purkinje cells (PC), Golgi cells (GOL), basket cells (BC), and stellate cells (SC). Mossy
ﬁbers (MF), which originate from brainstem nuclei, form excitatory synapses (+) with
granule cells and Golgi cells. Granule cells axons, or parallel ﬁbers (PF) form excitatory
synapses (+) with Purkinje cells. Golgi cells receive excitatory input (+) from granule
cell parallel ﬁbers, and in turn, form inhibitory synapses (-) with granule cells. Climbing
ﬁbers (CF), originating from the inferior olivary nucleus of the medulla, form excitatory
synapses (+) with Purkinje cells. Stellate and basket cells from inhibitory synapses (-)
with Purkinje cells. Purkinje cells form inhibitory synapses (-) with deep nuclei of the

cerebellum.

22

 

 

 

 

 

 

CF
1] ﬂ L:_l
Deep
cerebellar nuclei

 

Inferior olivary
nucleus

Figure 1.2

23

 

Spinal cord via
brain stem nuclei

cells. Purkinje cells also receive excitatory glutamatergic input from the climbing ﬁbers
originating from the inferior olive. Purkinje cell axons, which project onto nuclei in the
white matter of the cerebellar cortex, serve as the only output of the system. The mossy
ﬁber-granule cell-Purkinje cell pathway through the cerebellum, deﬁned by its high
frequency discharge, correlates with and controls behavior (Bastian and Thach, 1995).
The climbing ﬁber-Purkinje cell pathway does not directly control moment-to—moment
behavior, as its low frequency discharge increases only when errors in motor
performance occur and during adaptation or learning of movement (Bastian and Thach,
1995). Combined electrical stimulation of mossy and climbing ﬁbers depresses parallel
ﬁber Purkinje cell synapses that are concurrently active and spares those that are inactive.
The circuitry of the cerebellum is believed to allow the system to receive sensory
and motor input and adjust the motor output by comparing intended movements with
performed movements, recognizing errors in behavior, and computing corrections
(Kandel et al., 2000). Nearly every motor center in the CNS receives excitatory input
from nuclei located centrally, in the white matter of the cerebellum. One of these centers
is the spinal cord, which receives input from the cerebellum, indirectly, via brain stem
nuclei such as the vestibular and reticular nucleus. Other centers include the primary,
prefrontal, and premotor cerebral cortices, which receive cerebellar input indirectly from
the thalamus (Kandel et al., 2000). Activity in deep cerebellar nuclei correlates with
behavior as these nuclei differentially control motor systems and their respective
functions. The vestibular and fastigial nuclei control equilibrium, upright stance, and
gait; the interpositus controls stretch, contact, placing and other reﬂexes; and the dentate

controls voluntary movements of the extremities, such as those involved in reaching and

24

grasping for objects (Bastian and Thach, 1995). Thus, disruption of the cerebellar
circuitry and its output can result in many dysfunctions including abnormalities of limb
and eye movements, decrease in muscle tone, and production of incoordinated
movement, such as ataxia.

A limited number of GABAA receptor isoforms are expressed in the cerebellum
(Wisden et al., 1996; Poltl et al., 2003) (see Table 1.2). Studies using a mixture of two

subunit-speciﬁc antibodies to immunoprecipitate GABAA receptors from the cerebellum

In“

of rats have found that 28% of the total cerebellar GABAA receptors were the (110,72

I.

isoform, 36% were the (160272 isoform, and 23% were the 0:605 isoform (Quirk etal.,
1994). In addition to having strikingly different pharmacological and biophysical
properties, receptors containing the (11 and/or 0.6 subunits have markedly different cellular
distributions in the cerebellum. Studies using in situ hybridization and
immunocytochemistry of rat brain have shown that cerebellar Purkinje cells express
GABAA receptors containing (11, [32, [33, and 7 subunits (Wisden et al., 1996). Purkinje
cells contain greater levels of B2 mRNA than that of B3. In the adult rat, two splice
variants of the 7 subunit exist, 72L (long) and 725 (short), and the mRNA expression levels
of 721. are stronger than those for 723- Both 'th and 723 subunit-containing receptors are
most likely localized to Purkinje cell dendrites and spines. The predominant GABAA
receptor isoform in Purkinje cells is thought to be on, [32, y. In contrast, the composition
of GABAA receptors in cerebellar granule cells is more diverse and complex. The major
subunit genes expressed in cerebellar granule cells are al, 0.6, B2, [33, 72, and 6 (Laurie et
al., 1992). Interestingly, the 0.6 subunit is only expressed in GABAA receptors of

cerebellar granule cells and its expression is developmentally regulated both in vivo and

25

Table 1.2. GABAA receptor subunit subtypes expressed by different cells in the
cerebellum. The - symbol indicates no expression of the subunit subtype, and the ?
symbol indicates the subtype is unknown or not tested. Note that only cerebellar granule
cells express (:6 and 6 subunit—containing GABAA receptors. Note also that for certain
subtypes expression is limited to the prenatal or the postnatal stage of development.
Contributions to this table include studies by Laurie et al., 1992; Pollard etal., 1993;
Khan et al., 1994; Thompson et al., 1994; Gao et al., 1995; Wisden et al., 1996; Nusser et

al., 1996, 1998; and Poltl et al., 2003.

26

 

 

 

 

 

 

 

 

 

 

 

 

Cell GABAA Receptor Subunit Subtypes
Type a I3 y 5
Purkinje 0‘1 [32/ B3 YZL’YZS '
Granule 0.1 , 0‘6 , [32,133,030 YZS’YZL’CY3) 8
(a4)
Basket/ a1 ,(a3) I32 729th ‘
Stellate
Golgi (12(7), ? Y 1(?) -
(13(7)
Table 1.2

27

 

in vitro (Laurie et al., 1992; Thompson et al., 1994; Gao et al., 1995). Studies using in
situ hybridization, immunocytochemistry, and radioligand binding have shown that there
is a lack of expression of the 0.6 subunit during early cerebellar development and an
increased expression of a6 subunits in rats during maturation (Laurie et al., 1992,
Thompson et al., 1994). In contrast, GABAA receptors containing the al subunit are
expressed throughout cerebellar development and maturation (Laurie et al., 1992,
Thompson et al., 1994).

There is a great degree of segregation of distinct GABAA receptor subtypes on the
surface of cerebellar granule cells (Nusser et al., 1996; Nusser et al., 1998). This
segregation suggests different functional roles of the GABAA receptor depending upon
subunit composition and subcellular location. Receptors containing the (16, [32/3, and 6
subunits are present exclusively in the extrasynaptic membrane. These (166 subunit-
containing receptors have a high afﬁnity for GABA spilled over from Golgi cells (Rossi
et al., 1998), and do not desensitize to the prolonged application of agonist (Saxena et al.,
1994). It is therefore believed that this GABAA receptor subtype plays a role in
mediating tonic inhibition, which originates from the persistent opening of GABAA
receptor channels (Brickley et al., 1996; Nusser et al., 1998; Semyanov et al., 2004;
Hanchar et al., 2005). Receptors containing the al, (16, (32,3, and 72 subunits (01102572 and
(16021372, or (110602372) are primarily present in Golgi cell to granule cell synapses. These
receptors are thought to play a role in mediation of phasic inhibition, which originates
from synchronous opening of GABAA receptors. Furthermore, granule cell axon
terminals that form synapses with Purkinje cells possess 06 subunit containing receptors

which are thought to function in increasing glutamate release onto Purkinje cells (Nusser

28

et al., 1998). A subtype of GABAA receptor containing the (16 subunit ((1602372) is also
present in some granule cells that form synapses with glutamatergic mossy ﬁbers,
suggesting that it may play a functional role in regulation of these excitatory synapses.
GABAA receptors containing combinations of a1, a6, and (ll/0.6 also coexist in cerebellar
granule cells, suggesting that a variety of different GABAA receptor subtypes coexist in
these cells including (110,72, (160,72, (110160.72, a60x6 and a1a60x6 (Pollard et al., 1993;
Khan et al., 1994; Nusser et al., 1996, 1998).

Almost half (42- 45 %) of all GABAA receptors in the cerebellum contain (16
subunits (Khan et al., 1996; Jones et al., 1997; Jechlinger et al., 1998). The
pharmacological properties unique to 06 receptors disappeared in mice lacking the (16
gene; "a6 knockout" animals (u6-/-) were insensitive to the benzodiazepine receptor
ligand [3H]Ro 15-4513, insensitive to furosemide and Zn+2 inhibition, and showed a
reduction in high afﬁnity binding of the GABA agonist [3H]muscimol (Makela et al.,
1997). In these "cu, knockout" mice there was a dramatic loss in the 6 subunit protein,
most likely contributing to the reduction of high afﬁnity GABA agonist binding and to
the decreased Zn+2 inhibition observed in these animals. Along the same lines, rats
containing a point mutation in the (16 gene, also termed alcohol-non-tolerant (ANT) rats,
showed diazepam-mediated potentiation of Ion“. This increased afﬁnity for
benzodiazepines may contribute to the increased susceptibility to benzodiazepine induced
ataxia and impairment of postural reﬂexes observed in these animals (Korpi et al., 1993).
Thus, the (:6 subunit containing GABAA receptors, which are especially sensitive to

inhibition by heavy metals such as Zn+2 and La”, may play a critical role in cerebellar-

29

based motor coordination, the disruption of which can result from MeHg exposure

(Chang, 1977).

D. MeHg Neurotoxicity

(1. Cellular eﬂects of MeHg

The mechanisms by which MeHg produces neurotoxicity are not clear. MeHg
has a high afﬁnity for sulfhydryl and disulﬁde groups on cysteines that are numerous in
proteins and enzymes. Thus, MeHg has the potential to interfere with many cell
processes including disruption of protein synthesis (Syversen, 1981), disruption of
neurotransmitter release (Atchison and Narahashi, 1982; Atchison, 1986; Traxinger and
Atchison, 1987), and disturbance of divalent cation homeostasis (Ca"2 and Zn”)
(Levesque and Atchison, 1991; Hewett and Atchison, 1992; Hare et al., 1993; Denny et
al., 1993; Sirois and Atchison, 1996; Marty and Atchison 1997, 1998; Limke and
Atchison, 2002; Edwards etal., 2005). This is important, as GABAA receptor-mediated
activity can be modulated by numerous divalent cations. For example, Zn+2 inhibits
[GABA (Barberis et al., 2000) and Hg+2 potentiates GABAergic currents (Arakawa et al.,
1991). Furthermore, Ca+2 can activate intracellular signaling cascades that can alter
GABAA receptor function (Moss and Smart, 1996; Brandon et al., 2000). Alteration of
Ca“2 concentrations by MeHg can also have effects on synaptic transmission, such as
increased presynaptic neurotransmitter release and altered postsynaptic response of the
receptor (Sirois and Atchison, 1996; Yuan and Atchison, 1993, 1995, 1999).

Studies using single cell Ca+2 imaging with the ﬂuorescent dye fura-2 have shown

that MeHg induces changes in divalent cation homeostasis in a neuroblastoma- glioma

30

hybrid cell line NG108-15, synaptosomes, and in cerebellar granule and Purkinje cells
(Denny et al., 1993; Hare et al., 1993; Hare et al., 1995, Marty and Atchison 1997, 1998;
Limke and Atchison, 2002; Edwards et al., 2005). Using the ratio of fura—2 ﬂuorescence
at the wavelengths of 340/380 nm as an indicator of relative intracellular Ca“2
concentrations, our lab has demonstrated that MeHg increases levels of intracellular Ca+2
in a multiphasic, kinetically-distinct manner. The initial increase in cytosolic Ca+2
concentration by MeHg is related to Ca+2 release from internal stores (i.e. endoplasmic
reticulum and mitochondria) (Denny et al., 1993; Hare et al., 1993; Marty and Atchison,
1997; Limke and Atchison, 2002), and the ﬁnal phase results from Ca+2 entry into the
cell. Furthermore, a gradual increase in excitation ratio following the initial phase was
also observed, which was inhibited or reversed by the divalent heavy metal chelator,
TPEN, and was shown to be Can-independent. Studies determined that these Ca”-
independent alterations in fura-2 ﬂuorescence resulted from elevations of intracellular
Zn”, resulting from Zn+2 displacement from cytoplasmic proteins (Denny and Atchison,
1994). The response of cerebellar granule cells and NG108-15 cells to MeHg differed
only in their sensitivity to MeHg; granule cells were about ten times more sensitive to
MeHg than were NG108-15 cells. Likewise, the response of granule cells to MeHg was
similar to those of Purkinje cells, however the effects of MeHg in granule cells occured at
earlier time points than those in Purkinje cells (Edwards et al., 2005). In addition,

Purkinje cells were less sensitive to the MeHg-induced increase in [Ca+2]i

and subsequent
cell death. Thus, granule cells appear to be more sensitive to the effects of MeHg as

compared to NG108-15 cells and Purkinje cells.

31

1!,
E
g
E
é

 

b. Disruption of Synaptic Transmission by MeHg

Any of the adverse cellular effects of MeHg described above could inﬂuence
GABAergic synaptic transmission and play an important role in its neurotoxicity. This
is important as GABAA receptors are vital in maintaining the excitability levels of cells.
They accomplish this by mediating inhibitory synaptic transmission and regulating
glutamatergic excitatory synaptic transmission (DeLorey and Olsen, 1992; Olsen and
Avoli, 1997; Kardos, 1999). As cells do not operate in isolation and every cell is part of
a vast network of cells, changes in excitability of one cell can affect the excitability of
many others. Excess excitation can result in exitotoxic cell death (Kardos et al., 1999).

Disruption of synaptic transmission and alteration of channel-mediated synaptic
events by mercurials was ﬁrst demonstrated in the peripheral nervous system in isolated
nerve muscle preparations from cat cervical ganglion, frog sartorius muscle, and rat
diaphram (Kostial and Landeka, 1975; Manalis and Cooper, 1975; Atchison and
Narahashi, 1982). Intracellular recordings from these isolated nerve-muscle preparations
demonstrated that both HgCl2 and MeHg completely reduced the nerve-evoked release of
acetylcholine (ACh). In addition, MeHg affected excitatory endplate potentials (EPPs) by
ﬁrst producing a transient increase in EPP amplitude, followed by complete suppression
of response (Atchison and Narahashi, 1982; Traxinger and Atchison, 1987). MeHg has
been shown to inhibit mammalian neuromuscular synaptic transmission by decreasing the
mean quantal content of neurotransmitter released. This reduction in mean quantal
content is a result of a decrease in the available store of transmitter and an increase in the

probability of its release (by increasing [Ca+2]i) (Atchison and Narahashi, 1982). This

32

suggests the involvement of MeHg in presynaptic mechanisms of synaptic transmission
in the peripheral nervous system.

Acute application of MeHg also disrupts synaptic transmission in the central
nervous system. Similar to the effects of MeHg observed on EPPs, MeHg ﬁrst increases
and later inhibits population spike amplitudes in hippocampal neurons (Yuan and
Atchison, 1993; 1994). These effects of MeHg are also observed on excitatory post-
synaptic potential (EPSP) amplitudes in hippocampal slice (Yuan and Atchison, 1995).
In this preparation, MeHg causes an initial increase followed by a complete reduction of
both extracellular population spikes and of action potentials evoked by stimulation of
Schaffer collaterals (Yuan and Atchison, 1993; 1994; 1995). These effects are time- and
concentration-dependent. Increasing the stimulation intensity following the inhibition of
population spikes by MeHg, causes a complete recovery of responses. This suggests that
MeHg changes the excitability of the membrane, effectively increasing the threshold
needed for ﬁring of an action potential. Furthermore, in cell bodies of CA1 pyramidal
neurons, MeHg inhibits both orthodromically- and antidromically-activated population
spikes over a similar period of time. Because antidromic activation of population spikes
does not involve synaptic connections (because current is carried to the cell body
following axonal stimulation) it should not be as sensitive to the presynaptic effects of
MeHg as would orthodromically-activated population spikes. Orthodromically-activated
responses, which are induced by stimulation of Schaffer collaterals in region CA1,
involve a synaptic mechanism that incorporates activation of CA1 neuron dendrites by
axons of stimulated ﬁbers. These responses should be especially sensitive to any

presynaptic effects of MeHg. Because ﬁndings revealed that both orthodromically- and

33

antidromically-activated population spikes are reduced by MeHg over a similar time-
scale, it suggests that MeHg predominately affects postsynaptic sites in hippocampus.
Similar to the effects observed in hippocampus, MeHg also affects synaptic
transmission in the cerebellum. Using both extracellular and intracellular recording
techniques, acute bath application of MeHg causes an initial stimulation followed by a
depression of cerebellar synaptic transmission between parallel ﬁbers and Purkinje cells
or between climbing ﬁbers and Purkinje neurons (Yuan and Atchison, 1999). These
effects of MeHg are both concentration- and time-dependent because higher
concentrations of MeHg (100 pM) inhibit responses more rapidly than do lower
concentrations (20 M). Furthermore, MeHg also hyperpolarizes and then depolarizes
Purkinje cell membranes and suppresses the spontaneous activity of these cells.
Recently, studies using whole-cell voltage-clamp of granule cells in cerebellar slice have
revealed that spontaneous GABAergic inhibitory currents (sIPSCs) are more sensitive to
MeHg inhibition than are spontaneous excitatory synaptic currents (sEPSCs) because
sIPSCs are reduced more rapidly than are sEPSCs (Yuan and Atchison, 2003). Thus,

GABAA receptors appear to be particularly sensitive to the inhibitory effects of MeHg.

E. Disruption of GABAergic Function by MeHg

The effects of mercury (Hg‘z), the divalent inorganic ion which is methylated to
MeHg, on GABAergic function have been extensively investigated. Studies using the
whole-cell patch clamp technique on dorsal root ganglia neurons of rats have found that
HgCl2 potentiates GABAA receptor-mediated currents in the presence of low

concentrations of GABA (30 11M) (Arakawa et al., 1991). The enhancement by HgCl2

34

was prevented by cysteine and iodoacetamide, suggesting the involvement of sulﬂiydryl
groups in these actions (Huang and N arahashi, 1996). Hg+2 potentiation of GABA-
mediated currents was shown to be concentration-dependent, in that larger concentrations
of Hg+2 caused greater potentiation of responses. The effects of HgCl2 (0.1 - 100 M) on
GABAA receptor mediated currents were voltage-independent, suggesting that the
binding site for Hg2 was located on the external side of the receptor. Contradictory to
these ﬁndings, the response produced by Hg+2 was also shown to be use-dependent, in
that potentiation of GABAergic currents was enhanced with increases in frequency of
channel opening. In addition, Hg+2 accelerated the rate of desensitization of GABAA
receptor currents by decreasing the time constants for the fast and the slow phase of
desensitization (Huang and Narahashi, 1996). Hg+2 was also shown to induce a slow
inward current, insensitive to antagonists of voltage gated Na“, K+, or Ca+2 channels
(Arakawa et al. 1991). Non-speciﬁc cation channels were thought to carry this current.
Examination of HgCl2 potentiation of GABA-induced currents in the presence and
absence of ZnCl2 revealed that the binding site of HgCl2 on the receptor differed from
that of Zn+2 (Huang and Narahashi, 1996).

Evidence has shown that stimulation of G5 protein by choleratoxin and
inhibition of GilGo proteins by pertussis toxin decreased GABAA receptor-mediated
currents (Huang and Narahashi, 1997a). In a follow-up study it was found that protein
kinase A (PKA) negatively regulated GABAA receptor activity and abolished the Hg“2
potentiation of GABAergic responses (Huang and Narahashi, 1997 b). It was suggested
that G-proteins, speciﬁcally - Gi/G0 acting to inhibit PKA - are involved in the Hg”-

induced potentiation of IGABA.

35

In contrast to the extensively investigated effects of Hg+2 on GABAA receptor
currents, little is known about the effects of MeHg on GABAergic function. Whole-cell
patch clamp recordings in dorsal root ganglia neurons have shown that high
concentrations of MeHg (100 pM) suppress the peak currents induced by low
concentrations of GABA (30 M). Additionally, a slow inward current, presumed to be
mediated by non-speciﬁc cation channels, was observed similar to that seen in response
to Hg+2 (Arakawa et al., 1991). In hippocampus, GABAergic inhibitory synaptic
transmission is more sensitive to inhibition by MeHg than is excitatory transmission
(Yuan and Atchison, 1995, 1997). This preferential effect of MeHg on GABAA receptor-
mediated inhibitory synaptic transmission was shown to underlie the early enhancement
of excitatory synaptic transmission (Yuan and Atchison, 1997). Importantly, GABAergic
synaptic transmission was shown to be more sensitive to MeHg inhibition in granule cells
than in Purkinje cells because suppression of granule cell IPSCs occurs much earlier than
does suppression of Purkinje cell currents (Yuan and Atchison, 2003). A spontaneous,
transient, slow inward Cl'-mediated current was also seen in the slice preparation that was
determined not to be GABAA receptor-mediated (Yukun and Atchison, 2005). The role
of this current in MeHg-induced neurotoxicity is unknown.

In primary cultures of cerebellar granule cells, MeHg (0.1 pM - 10 M)
signiﬁcantly suppressed whole cell GABAA receptor-mediated currents in a time- and
concentration- dependent manner. MeHg also induced a slow inward current in these
cells that was insensitive to voltage-activated Na", K”, and Ca+2 channel blockers. The
Zn“2 chelator, TPEN, was able to delay the onset of this non-speciﬁc inward current as

well as delay the time to decrease of GABAA currents (Xu and Atchison, 1997).

36

Several studies have suggested that MeHg interaction with the GABAA receptor
enhances benzodiazepine binding in rodent cerebellum and cerebral cortex (Corda et al.,
1981; Concas et al., 1983; Fonfria et al., 2001). Corda et a1. (1981) observed that a single
administration of MeHg given by gavage produced a long-lasting increase in 3H-
diazepam binding in rat cortex. Similarly, Concas et a1. (1983) showed that long-term
administration of MeHg enhanced 3H-diazepam binding in rat cerebellum. In primary
cultures of mice cerebellar granule cells, Fonfria et a1. (2001) observed that 30 min
exposure to MeHg induced an increase in [3H]ﬂunitrazepam binding at the GABAA
receptor. However, in direct contradiction to these results, Komulainen and Tuomisto
(1985) found no effect of either acute or chronic exposure to MeHg on the number of

benzodiazepine receptors labeled with [3H]ﬂunitrazepam in rat cerebellum.

37

SPECIFIC AIMS
1. What effects does MeHg have on GABAA receptor function in cells in culture? Does
MeHg interact with a known modulatory site(s) on the GABAA receptor? Or does it bind

to a different site(s) on the receptor?

2. Does MeHg differentially affect the functional properties of GABAA receptors in (16
subunit-containing and non-a6 subunit-containing cells? Are these actions subunit
speciﬁc? Are dig-containing GABAA receptors of granule cells especially sensitive to the

inhibitory effects of MeHg?

Speciﬁc Aim 1 is addressed in Chapter 2. Time-course-of-[GABA-suppression
experiments were performed with three different concentrations of MeHg (0.1, 1.0, and
10 uM) in cerebral cortical cells in culture. In addition, the effect of each concentration
of MeHg on GABAA receptor mediated current kinetics was determined. Furthermore,
the effects of diazepam (0.1, 1.0, and 10 uM) on [GABA current kinetics were investigated.
These concentrations of MeHg are clinically relevant, as they are within the range of
concentrations found in the blood of individuals poisoned with MeHg in the mass
poisoning which occurred in Iraq in the 1970’s (Bakir et al., 1973). Following acute
exposure to MeHg, signs of toxicity such as muscle weakness and ataxia occurred in
poison-victims at concentrations ranging between 0.99 -1.56 moles of Hg. The
concentration of MeHg associated with toxicity (1.0 moles of Hg) can be converted to
approximately 12.5 uM, when the average adult male weight (80 kg) and tissue density

(1.0 kg/liter) is considered. The lower concentrations of MeHg (0.1 and 1 uM) were used

38

in the present study to determine the concentrations at which the effects of MeHg become
evident.

To determine whether or not MeHg interacts with the diazepam binding site of the
receptor, effects of MeHg on long». suppression and kinetics were examined in the
presence of diazepam, either with or without pretreatment with MeHg or diazepam.
Experiments which examined the effects of MeHg on IGABA in the presence of a
benzodiazepine antagonist, ﬂumazenil (10 uM), were also performed.

The experiments from Speciﬁc Aim 2 are described in Chapter 3. Sensitivity to
MeHg of cells containing (16 or al phenotyes of GABAA receptor was investigated using
time-course-of-IGABA-suppression experiments performed with three different
concentrations of MeHg (0.1, 1.0, and 10 uM) in both cerebellar granule (an-containing)
and cerebral cortical (on-containing) cells in culture. To verify the presence of speciﬁc
GABAA receptor subtypes of interest, immunocytochemistry was also performed in both
cell types. In addition, the effect of each concentration of MeHg on the current-voltage
relationship in granule and cortical cells was determined. Furthermore, to verify the
ﬁndings obtained in native cells, experiments which examined the effects of MeHg on
IGABA in human embryonic kidney (HEK-293) cells expressing either (16 or (xi-containing
GABAA receptors were performed.

To investigate these aims, several techniques were employed. To examine the
effects of MeHg on functional properties of native GABAA receptors, whole-cell voltage-
patch clamp recordings were performed. The patch-clamp technique is ideal in that it
allows for measurements of whole-cell or single-channel currents at a millisecond time

resolution (Hamill et al., 1981). GABAA receptor-mediated responses typically occur

39

within hundreds of milliseconds (Maconochie etal., 1994). Whole-cell recordings were
performed on rat cerebellar granule and cortical neurons in primary culture. Cortical
cells were used as a substitute for cerebellar Purkinje cells, as they contain the same type
of GABAA receptor phenotype as do Purkinje cells (Hansen et al., 2001) and they are
more successfully cultured. Purkinje cells are difﬁcult to culture because they have an
early mortality rate and often become contaminated with other cell types, particularly
glial cells (Edwards, J ., personal communication). Furthermore, the distribution of
Purkinje cells in the cerebellum is sparse (as compared to granule cells), they do not
produce a large yield (Edwards, J., personal communication), and Purkinje cells have a
very arborized dendritic tree, which is clipped in culture. To verify results obtained in
granule and cortical neurons, experiments were also conducted in HEK-293 cells
transfected with rat cDNA for either (11 or (16 subunit-containing GABAA receptors. To
identify the target site(s) for MeHg, pharmacological characterization of GABAergic
responses was examined using subunit-subtype speciﬁc agonists and antagonists. In
addition, the presence of either on or (16 subunit-containing GABAA receptors in cells was

determined using immunoctyochemistry.

40

CHAPTER TWO

EFFECTS OF METHYLMERCURY ON DIAZEPAM-MEDIATED ENHANCEMENT

OF GABAA RECEPTOR CURRENTS IN RAT CORTICAL NEURONS IN CULTURE

41

 

ABSTRACT

GABAA receptor-mediated synaptic transmission is extremely sensitive to the
effects of MeHg, as it is inhibited more quickly by MeHg than is excitatory synaptic
transmission. The GABAA receptor has many drug binding sites, including those for
barbiturates and benzodiazepines. However, the site at which MeHg interacts with the
GABAA receptor has not yet been determined. Previous work in other labs suggested an
effect of MeHg on benzodiazepine receptors. To characterize the interaction of MeHg
with the GABAA receptor and to test if MeHg interacts with the benzodiazepine binding
site of the receptor, diazepam (0.1, 1.0, and 10 M) or ﬂumazenil (10 M), a
benzodiazepine antagonist, was applied to cerebral cortical cells in primary culture in the
absence and presence of MeHg (1 uM). [GABA were obtained by means of the whole-cell
voltage patch-clamp technique using a symmetrical Cl' concentration. MeHg (0.1, 1.0,
and 10 M) caused a progressive suppression of 15AM. Kinetically, MeHg prolonged
IGABA slow decay time at each concentration tested. Similarly, IGABA slow decay time
was prolonged at each [diazepam] tested. When MeHg was applied in conjunction with
diazepam, the decay time of IGABA was prolonged further as compared to either MeHg or
diazepam alone. Additionally, co-application of MeHg plus diazepam still prolonged
IGABA slow decay following pretreatment with either MeHg or diazepam alone. Finally,
ﬂumazenil did not prevent or reduce the effects of MeHg on IGABA slow decay time.
These results suggest that MeHg and diazepam have independent effects on the GABAA
receptor and that MeHg does not appear to compete with the same binding site on the

receptor as does diazepam, although MeHg does alter decay kinetics.

42

 

 

INTRODUCTION
MeHg is a pervasive environmental contaminant that continues to pose a
signiﬁcant present-day toxicological threat, particularly in the Great Lakes, where levels
of MeHg in ﬁsh have been found to exceed the estimated dose for adverse health effects
in humans (Rice, 1995; Gerstenberger and Dellinger, 2002; Gilbertson, 2004; Weis,

2004). MeHg is highly neurotoxic and, in humans, can cause severe neurological signs

 

including constriction of the visual ﬁeld, speech impairment, mental disturbances, and
ataxia (Chang, 1977). MeHg has a high afﬁnity for sulﬂiydryl groups on cysteines which
are numerous in cell membranes, proteins and enzymes. Thus, MeHg has the potential to
bind to cell membrane proteins, change their conformation, and interfere with many
cellular processes (Atchison and Hare, 1994; Sirois and Atchison, 1996; Denny and
Atchison, 1996; Limke et al., 2004); Among these are disruption of excitatory and
inhibitory synaptic transmission (Yuan and Atchison, 1993, 1997, 1999, 2003).
GABAergic synaptic transmission is inhibited by MeHg. Whole-cell patch
clamp recordings in dorsal root ganglion neurons have shown that high concentrations of
MeHg (100 uM) suppress the peak currents induced by low concentrations of GABA
(Arakawa etal., 1991). In hippocampal slice, MeHg was shown to decrease gradually
inhibitory postsynaptic potential (IPSP) amplitudes to complete suppression (Yuan and
Atchison, 1995). Suppression of inhibitory synaptic transmission by low concentrations
of MeHg occurs earlier than does suppression of glutamatergic synaptic transmission
(Yuan and Atchison, 1995, 1997), suggesting that inhibitory synaptic transmission is
more sensitive to inhibition by MeHg than is excitatory transmission. Similarly, studies

conducted in rats treated chronically with MeHg have suggested that MeHg impairs

 

43

inhibitory GABAergic neurons in cerebral cortex, neostriatum, and caudate putamen
(O'Kusky and McGeer, 1985, 1989). Importantly, some cell types appear to be
particularly sensitive to the effects of MeHg such as granule cells of the cerebellum.
Suppression of inhibitory post synaptic currents (IPSCs) in cerebellar granule cells, for
instance, occurs much earlier than does suppression in neighboring Purkinje cells (Yuan
and Atchison, 2003). Whether or not MeHg differentially interferes with GABAergic
function in different cell types is not known, nor has a speciﬁc mechanism for
suppression of GABAA receptor function by MeHg been identiﬁed. One possibility for
this difference in sensitivity to MeHg may be a differential expression of GABAA
receptor subunits in various cell types, as different GABAA receptor subunit composition
confers GABAA receptors with different pharmacological and electrophysiological
properties. Current carried by GABAA receptors containing the (:1 subunit, for instance, is
potentiated by benzodiazepines and La”, and is relatively insensitive to inhibition by
Zn+2 and furosemide, as compared with that carried by (16 subunit-containing receptors.

In addition, the on subunit-containing receptor has a low efﬁcacy and afﬁnity for
pentobarbital. Furthermore, receptors containing the (1; subunit desensitize more slowly
than do other subtypes of GABA receptors, such as those containing the (1., subunit (Tia et
aL,1996)

Several studies have suggested that MeHg interaction with the GABAA receptor
involves the benzodiazepine binding site. MeHg enhances benzodiazepine binding to
GABAA receptors in rodent cerebellum and cerebral cortex (Corda et al., 1981; Concas et
al., 1983; Fonfria et al., 2001). Corda et a1. (1981) observed that a single administration

of MeHg produced a long-lasting increase in 3H-diazepam binding in rat cortex.

Similarly, Concas et a1. (1983) showed that long-term administration of MeHg enhanced
3H-diazepam binding in rat cerebellum. In primary cultures of mice cerebellar granule
cells, Fonfria et a1. (2001) observed that short-term exposure to MeHg induced an
increase in [3H]ﬂunitrazepam binding at the GABAA receptor. However, in direct
contradiction to these results, Komulainen and Tuomisto (1985) found no effect of either
acute or chronic exposure to MeHg on the number of benzodiazepine receptors labeled
with [3H]ﬂunitrazepam in rat cerebellum. Thus, whether or not MeHg interacts with the
benzodiazepine binding site of the GABAA receptor to affect GABAA receptor function
remains to be determined. The present study was designed speciﬁcally to characterize
the interaction of MeHg with the (11 subunit-containing GABAA receptor, as the diazepam
binding site has been implicated in the interactions of MeHg with the GABAA receptor.
Using the whole-cell voltage patch-clamp technique, results from the present
study show that the effects of MeHg are not changed in the presence of diazepam or a
benzodiazepine antagonist. These ﬁndings suggest that MeHg interacts at a site on the

GABAA receptor different from that used by diazepam.

45

 

METHODS

Solutions and Chemicals

Methylmercuric chloride (MeHg) (ICN Biomedical lnc., Costa Mesa, CA, USA)
was dissolved in deionized water to a ﬁnal concentration of 10 mM, which served as a
stock solution. On the day of experiments, MeHg working solutions (0.1, 1, or 10 uM)
were diluted in extracellular solution consisting of (in mM): NaCl, 125; CaCl2, 2.0; KCl,
2.5; MgCl2, 1.0; KH2PO4, 1.25; NaHC03, 26.0, and D-glucose, 20.0, pH-adjusted to 7.4
using 95% 02/5% CO2. Diazepam, 7-amino-n-butyric acid (GABA), ﬂumazenil,
bicuculline, D-penicillamine, 6-cyano-7-nitroquinoxaline-2,3-dione disodium salt
(CNQX), DL-2-amino-5-phosphonopentanoic acid (APV), glutamine, 4-(2-
hydroxyethyl)piperazine-l-ethanesulfonic acid (HEPES), gentamycin, deoxyribonuclease
I (DNase I), cytosine B-D-arabino-furanoside (Ara-C), adenosine 5'- triphosphate
magnesium salt (ATP), ethylene glycol-bis(B-aminoethyl ether)-N, N, N', N',-tetraacetic
acid (EGTA), trypsin, and poly-D-lysine were all purchased from Sigma Chemical Co.
(St. Louis, MO). Dulbecco’s Modiﬁed Eagle’s Medium (DMEM) was purchased from
Mediatech, Inc. (Hemdon, VA), and heat inactivated horse serum was purchased from
GIBCO-Invitrogen Inc. (Carlsbad, CA). Diazepam and ﬂumazenil were dissolved in
dimethyl sulphoxide (DMSO; < 0.01% (v/v) ﬁnal concentration) and diluted in
extracellular solution prior to use. Other drugs were dissolved in deionized water and

diluted in extracellular solution prior to use.

46

 

Preparation of Primary Cortical Cell Culture

Primary cultures of cortical neurons were obtained from 3-4 newborn Sprague-
Dawley rat pups (Charles River Laboratories, Wilmington, MA) following methods
described by Ingleﬁeld et al. (2001). Brieﬂy, following removal of the brain and
isolation of cortex, cells were digested for 4.5 min at 37° C with trypsin 0.025 % (w/v) in
buffer that contained (in mM): KCI (5.0), KH2PO4 (0.205), NaCl (137.0), Na2HPO4
(0.17), D-glucose (5.0), sucrose (59.0), and gentamycin (0.1 mg/ml), pH 7.3. Trypsin
was inactivated by addition of the above-described buffer containing 0.016 % (w/v)
DNase I for 4.5 min at 37° C. To this mixture, warmed DMEM supplemented with 10 %
(w/v) horse serum, 10 mM HEPES, 2 mM glutamine, and gentamycin (0.1 mg/ml) was
added and the cells were centrifuged (SOOxg, 5 min). The resulting pellet was
resuspended in DMEM containing DNase I and was centrifuged again. The cell
suspension was then resuspended in DMEM and triturated with a Pasteur pipette. Cells
were plated at a density of l x 106 cells/ dish onto 35-mm Petri dishes (Corning 1nc.,
Corning, NY) coated with poly-D-lysine. Cells were maintained in a 37° C incubator
with 95% O2 and 5% CO2. DMEM was replaced every two days. Cytosine [3-
arabinoside (5 M) was added to cell cultures 72 hr after plating to inhibit glial cell
proliferation. Cells were used for recordings 10-14 days after plating. All experiments
were done in compliance with NIH and university standards and were approved by the

Michigan State University Committee on Animal Use and Care.

Whole-cell Recordings in Cortical Cells
Prior to making whole-cell patch voltage-clamp recordings, the Petri dish

containing the cells was rinsed and then covered with approximately 1 ml of recording

47

solution. CNQX (10 M) and APV (50 M) were added to the extracellular solution to
inhibit ionotropic glutamate receptor-mediated responses. Recording patch electrodes
were fabricated from borosilicate capillary glass (0D: 1.5 mm, 1D: 1.0 mm; Garner
Glass Co., Claremont, CA) and had an impedance of 3 - 7 MO when ﬁlled with a pipette
solution consisting of (in mM): CsCl, 140; NaCl, 4.0; CaCl2, 0.5; HEPES, 10.0; EGTA-
CsOH, 5.0; Mg-ATP, 2.0, pH adjusted to 7.3. Following acquisition of a cell-attached
patch and cancellation of capacitative currents, the whole-cell conﬁguration was obtained
by negative pressure applied to the syringe. Cells were voltage clamped at -60 mV so
that GABA application produced inward currents under our approximately symmetric
[Cl'] conditions. The zero current potential was near 0 mV, which is close to the
expected Cl' equilibrium potential. Capacitance transient neutralization and series
resistance compensation were optimized. All experiments were carried out at room
temperature of 22 - 25° C.

GABA was pulsed onto neurons using a picospritzer (Picospritzer II, General
Valve Corporation, Fairﬁeld, NJ) delivered through a glass pipette (impedance of ~ 900
k0) placed close to the cell. GABA (50 M) was applied onto the target cells by a 10 ms
pulse given at intervals of 30 s. This was long enough for GABAA receptors to recover
completely from desensitization (results not shown). Physiological saline alone or
containing MeHg, diazepam and/or ﬂumazenil was delivered into the extracellular
bathing solution at a rate of approximately 0.45 ml/min via a gravity—powered perfusion
system through a series of tubes placed close to the cell. To observe the effects of MeHg
on IGABA, recordings were made in the presence or absence of MeHg. In some

experiments, diazepam (0.1- 10 M) or ﬂumazenil (10 M), a benzodiazepine antagonist,

48

 

was applied simultaneously with MeHg to test for any competitive interactions of the
drugs. In other experiments, cells were pre-treated with either MeHg (1 M) or diazepam
(1 uM) prior to co-application of both agents to allow sufﬁcient time for MeHg or
diazepam to interact with the receptor and affect IGABA. Pretreatment time with MeHg
was approximately 15 — 20 min, the time needed for MeHg to produce prolongation of
IGABA slow decay rate. Pretreatment with diazepam was approximately 1- 2 min, or until
maximum effects on prolongation of IGABA decay rate were evident. In the present

experiments, three MeHg concentrations were used (0.1, 1.0, and 10 pM).

Data Acquisition and Analysis

Data were acquired every 30 s using an IBM compatible computer equipped with
a DigiData 1200 interface and pClamp 8.1 software (Axon Instruments, Union City, CA).
IGABA were recorded using an Axopatch l-D ampliﬁer and low-pass ﬁltered at 1 kHz with
an eight pole ﬁlter before digitization at 10 kHz, storage and display. Off-line analysis
was performed using Clampﬁt 8.1 software (Axon Instruments) and the MiniAnalysis
program 5.6.10 (Synaptosoft Inc., Decatur, GA). Using MiniAnalysis software, effects of
MeHg on kinetics of Iona». were examined. Each GABAA receptor-mediated event was
selected manually and marked. The maximum amplitude for each event was measured
at the peak of the inward current. The decay phase was best ﬁtted using the Levenberg-
Marquardt least-squares method (Levenberg, 1944; Marquardt, 1963) with a double-
exponential function of the form 2 ant", where n is the number of exponential
components, a is the relative amplitude of the component, and r is the time constant.

Statistical analyses of data collected before and during application of MeHg were

49

 

conducted using a Student's paired t-test or a one-way analysis of variance, depending
upon the number of groups being compared. The Tukey-Kramer test was used for post
hoc comparisons. Values were considered statistically signiﬁcant at p<0.05. The data are
presented as mean :SEM and the number of replications is given. The data were
obtained from approximately 46 separate cell cultures, and the number of replications

refers to individual cells obtained from at least 2 separate cell cultures.

50

o
i
t
'—
4.
‘V
T
L
5.
‘.
‘
-i
i

 

RESULTS

MeHg Suppresses [GABA and Prolongs [GABA Slow Decay

Under symmetrical [Cl'], application of a GABA pulse produced an inward
current, which was completely abolished by bath application the GABAA receptor
antagonist bicuculline (10 M) (Fig. 2.1A), indicating that this inward current was a
GABAA receptor-mediated response. Unlike bicuculline, which decreases IGABA rapidly
and reversibly, continuous perfusion of 1 uM MeHg into the bathing solution had little
immediate effect on IGABA (Fig. 2.1B). However, as shown in Figure 2.2, with additional
time of exposure, MeHg caused a progressive reduction of IGABA. The effects of MeHg
never reached a steady-state level less than complete suppression of IGABA (Fig. 2.2A).
This effect was not reversible by washing cells for 5 min with MeHg-free solution
containing D-penicillamine (20 M), a MeHg chelator (Fig. 2.2B). The time course of
[GABA reduction by MeHg is shown in Figure 3A. Onset of reduction typically occurred
between 2 - 8 min, depending upon the MeHg concentration used. Increasing the
concentration of MeHg shortened the time course of suppression. For each concentration
of MeHg tested, the time-course of MeHg suppression was approximately linear. Time
to total reduction of [GABA by MeHg differed signiﬁcantly between each concentration of
MeHg tested (p<0.05) (Fig. 2.3B). At 10 M MeHg, the highest concentration tested,
complete reduction of IGABA occurred within approximately 30 min.

IGABA in cortical cells typically decayed in a bi-exponential manner consisting of a
fast (In...) and slow (15.0w) component as shown in Figure 2.4A. The fast component

decayed at an average of 650 :I: 102 ms and the slow component at

51

Figure 2.1. MeHg causes a gradual suppression of bicuculline-sensitive IGABA in
cortical cells. A, Effects of 10 pM bicuculline on GABAA receptor-mediated currents.
Whole-cell IGABA were evoked from a holding potential of -— 60 mV by a 10 ms pulse of
GABA (50 M, black dot), at intervals of 30 s, in the presence of CNQX (10 M) and
APV (50 pM) in the external solution to block glutamatergic currents. Data were
collected before (control), after 150 s (hash bar) exposure to bicuculline, and after 180 s
(hash bar) wash with bicuculline-free solution. Bicuculline caused a complete and
reversible suppression of IGABA. B, Comparison of early time course of effects of 10 M
bicuculline (top) and 1 uM MeHg (bottom) on consecutive GABAA receptor-mediated
currents. IGABA were obtained under identical conditions to those described above. Data
were collected before (control) and after application of bicuculline (top) or MeHg
(bottom) (as indicated by an arrow). Bicuculline caused immediate reduction of [GABA

whereas over the same time course MeHg did not affect IGABA.

52

A

Control Bicuculline Wash

 

' /o 1[/0

 

 

2000 ms

V

B Control Bicuculline

 

 

 

 

 

 

 

30 s
<
D.
O
8
2000 ms
Control MeHg 30 S r
i
<2:
G.
O
O
1.0
2000 ms

Figure 2.1

53

Figure 2.2. MeHg causes a complete and irreversible suppression of loan in
cortical cells. A, Time course of effects of 1 uM MeHg on GABAA receptor-mediated
currents. IGABA were recorded under identical conditions to those described in Figure 2.1.
Data were collected before (control) and at various time points after exposure to MeHg
(6, 8, 11, 16, and 37 min). MeHg caused a progressive and complete reduction of IGABA.
B, Effects of MeHg on IGABA are irreversible, Currents were collected before (control),
after 35 min (hash bar) exposure to MeHg (1 M), and following 10 min (hash bar) wash

with D-penicillamine (D-pen)-containing (20 uM), MeHg-free solution.

54

GABA
O

37 min

16 min
11 min
8 min

6 min

 

Control 1000 ms
(0 min)

B

Control MeHg D-pen

 

Figure 2.2

55

Figure 2.3. Suppression of IGABA by MeHg in cortical cells is time- and
concentration-dependent. A, IGABA were recorded in rat cortical cells in primary
culture at [MeHg] of 0.1 uM (open squares), 1.0 7.1M (ﬁlled triangles), or 10 M (open
circles). Recording conditions were as described in Figure 2.1. Data were collected
continually before and after MeHg exposure. Current amplitude in response to MeHg was
normalized to that of control. Each datum point represents the mean value recorded from
3-5 cells. B, Comparative effects of different concentrations of MeHg on time to
complete reduction of IGABA. Each bar represents the mean values obtained from 3-5
cells. The times to total reduction of IGABA by MeHg differed signiﬁcantly between each

concentration of MeHg tested, as is indicated by the asterisks (*, p<0.05).

56

 

 

0|
’7; 25
E
8
°\° 50
<
2
.6 75
100

 

III

60r

 

 

Time to Total Reduction (min)

0.1 1.0 10
[MeHg] (HM)

Figure 2.3

57

approximately 1291 i 161 ms. In the presence of MeHg, the fast component appeared to
be unaffected (data not shown), but 15.0,, was prolonged signiﬁcantly (p<0.05) (Fig.
2.4A). However, the percent increase over control in 15.0“, produced by MeHg was
concentration-independent (Fig. 2.4B). The time courses of prolongation of rslow of [GABA
by MeHg are shown in Figure 2.5A. Onset of slow decay prolongation typically occurred
between 1.5 — 18 min, depending upon the MeHg concentration used (Fig. 2.5B).
Increasing the concentration of MeHg shortened the time to onset of slow decay
prolongation. Similarly, time to peak prolongation of slow decay by MeHg was
concentration-dependent (Fig. 2.5C). The higher the concentration of MeHg, the less

time was required to reach the peak prolongation of IGABA slow decay.

Effects of MeHg on [GABA are Not Altered in the Presence of Diazepam

For comparison, effects of diazepam on IGABA in cortical cells were also
examined. Diazepam (0.1, 1.0, or 10 M) was continuously bath-perfused onto cortical
cells during 30 s pulses of GABA. Consistent with previous reports on the effects of
benzodiazepines on IGABA (Vicini et a1, 1986; Sieghart, 1992), diazepam prolonged the
decay of IGABA. In particular, diazepam prolonged the slower component of decay (Tsrow)
(Fig. 2.6A), usually within 30 — 60 s after application. The percent increase over control
in rslow produced by diazepam was concentration-dependent, as it differed signiﬁcantly
between each concentration of diazepam tested (Fig. 2.6B) (p<0.05). Thus, diazepam
and MeHg may act at a similar site via a similar mechanism to prolong IGABA slow decay.
If this is the case, then diazepam and MeHg could compete for the same binding site on

the GABAAreceptor.

58

Figure 2.4. MeHg prolongs [GABA slow decay constant in a concentration-
independent manner. A, Representative IGABA (control, dark line; 1510“,: 981.4 ms)
showing the effect of 1 uM MeHg (dotted line; rsiow: 1451.5 ms) on slow decay rate
(15.0w). Current amplitude in response to MeHg was scaled to that of control. IGABA were
recorded under conditions identical to those described in Figure 2.1. B, Comparative
effects of MeHg on prolongation of Tsjow before (control, black bar) and after exposure to
MeHg at the concentrations indicated (0.1 uM - 10 M, grey bars). Each bar represents
the mean values of rsiow prolongation recorded from 15-16 cells. MeHg prolonged mm in
a concentration-independent manner that differed signiﬁcantly from control, as indicated

by the asterisk (*, p< 0.05).

59

GABA
O

 

   

'1 :981.4 ms

slow

<2

500

1000 ms

 

 

Control 0.1 1,0 10
[MeHg] (MM)

Figure 2.4

60

Figure 2.5. MeHg affects the time course of prolongation of [GABA slow decay in a
concentration-dependent manner. A, IGABA were recorded at [MeHg] of 10 M (dashed
line, open circles), 1.0 7.1M (ﬁlled line, open triangles), or 0.1 uM (dashed and dotted line,
ﬁlled crosses). Recording conditions were as described in Figure 2.1. Data were
collected immediately before and approximately every 60 -90 s after MeHg exposure.
Each trace represents the mean values recorded from 3-5 cells. B, Comparative effects of
MeHg on time to onset of prolongation of tslow, Data were collected after exposure to
MeHg at the concentrations indicated (0.1 uM — 10 M, grey bars). Each bar represents
the mean values obtained from 3-5 cells. Times to onset of rsiow prolongation differed
signiﬁcantly between each concentration of MeHg tested, such that times were lessened
by higher concentrations of MeHg, as indicated by an asterisk (*, p< 0.05). C,
Comparative effects of MeHg on time to peak of prolongation of Tsjow. Data were
collected after exposure to MeHg at the concentrations indicated (0.1 uM — 10 M, grey
bars). Each bar represents the mean values obtained from 3-5 cells. Times to peak of
Tsjow prolongation differed signiﬁcantly between each concentration of MeHg tested, such
that times were reduced at higher concentrations of MeHg, as indicated by the asterisk (*,

p< 0.05).

61

A

—+- 0.1 uM

 

 

200 '

 

50 — "9" 10uM
+ 1pM

coscoo i .3?

40

30

20

Time (min)

10

B

 

30”

2E5 580 2 25

 

[MeHg] (MM)

Figure 2.5

62

O

 

 

E
E
x
(U
G)
o.
.9.
(D
.E
I—

10 1.0 0.1

[MeHg] (71M)

Figure 2.5

63

Figure 2.6. Diazepam prolongs chnn slow decay constant in a concentration-
dependent manner. A, Representative IGABA (control, dark line, rslow: 413 ms) showing
the effects of 1 uM diazepam (dashed line; Tsjowi 722.6 ms) on slow decay rate. Current
amplitude in response to MeHg was scaled to that of control. B, Comparative effects of
diazepam on concentration-dependent prolongation of Tsjow (control, black bar; diazepam,
0.1 uM - 10 M, hashed bars). Each bar represents the averaged values obtained from 14
-15 cells. Prolongation of Tsjow (% control) differed signiﬁcantly among concentrations of

diazepam tested as indicated by the asterisk (*, p<0.05).

A GABA
O

U)

200

150

 

Tsiow (°/o Control)

Figure 2.6

 
  
 

Control

I
I 1.10.; 413.0 ms

’ 1' -722.6 ms

I 4— slow'

0pA

O

10

500 ms

 

0.1 1.0 10
[Diazepam] (11M)

65

 

To test this, the combined effects of MeHg and diazepam on IGABA were
examined. The intermediate [MeHg], 1 7.1M , was chosen for all these studies, inasmuch
as the percent reduction of [GABA and the percent prolongation of Tsjow by MeHg appeared
to be concentration-independent. Co-application of MeHg with diazepam (1 and 10 pM)
prolonged signiﬁcantly Tsjow as compared to the effect of MeHg alone (p<0.05) (Fig. 2.7).
However, prolongation of [GABA Tsjow by MeHg with diazepam (0.1 uM) did not reach
signiﬁcance (p>0.05). Furthermore, co-application of diazepam with MeHg did not alter
the time course of IGABA reduction by MeHg (Fig. 2.8).

Next, cells were pretreated with diazepam prior to co—application of MeHg plus
diazepam to allow sufﬁcient time for diazepam to interact with the receptor.

Pretreatment with diazepam did not interfere with the effects of MeHg on 10AM. As
shown in Figure 2.9A-B, under these conditions MeHg plus diazepam prolonged rsiow to a
signiﬁcantly greater extent as compared to the effect of diazepam pretreatment alone
(p<0.05). At a lower concentration of diazepam (0.1 uM), however, a signiﬁcant
prolongation of rslow following diazepam pretreatment was not observed (p>0.05). In
addition, the time course of [GABA reduction by MeHg was not affected by co-application,

following pretreatment with diazepam (Fig. 2.8) (p>0.05).

Effects of Diazepam on IGABA Are Not Altered in the Presence of MeHg
To allow sufﬁcient time for MeHg to interact with the receptor, cells were then
pretreated with MeHg prior to a combined application of diazepam and MeHg.

Pretreatment with MeHg did not interfere with the effects of diazepam on long... As

66

Figure 2.7. Co-application of MeHg and diazepam does not affect MeHg-induced
prolongation of IGABA slow decay. A, Representative traces depict effects of co-
application of 1 pM MeHg plus 1 uM diazepam (dashed line; rsiow: 817.5 ms) on slow
decay time constant of [GABA (control, dark line, Tslow: 665.4 ms). Current amplitude in
response to MeHg was scaled to that of control. B, Comparison of effects of 1 uM MeHg
(dark grey bar) and co-application of 1 uM MeHg plus diazepam (0.1 uM - 10 M,
hashed bars) on prolongation of Tsiow (control, black bar). Each bar represents the
averaged values obtained from 6 - 8 cells. Prolongation of 1510,. produced by MeHg plus
diazepam differed signiﬁcantly from control and the effects produced by 1 pM MeHg

alone at 1 M and 10 M diazepam as indicated by the asterisk (*, p<0.05).

67

 
    

' rslow: 665.4 ms
‘— Tslow: 817.5 ms

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

B 1000 ms

g 150:: T %/_ /

(C3: 100 — -. % / /
IBIZebnjrll (11:4; +1 ill; Merl;

68

Figure 2.8. Time course of [GABA suppression by MeHg in cortical cells is not altered
by application of diazepam or ﬂumazenil. A, Comparison of time courses of IGABA
reduction by 1 pM MeHg (ﬁlled circles, dotted line), co-application of MeHg plus 1 uM
diazepam, following MeHg pretreatment (open circles), co-application of MeHg plus
diazepam following diazepam pretreatment (ﬁlled triangles, hashed line), co-application
of MeHg plus 10 pM ﬂumazenil (open diamonds), and co-application of MeHg plus
diazepam (open squares). Each datum point represents the averaged values obtained
from 3-5 cells. B, Comparative effects on time to reduction of [GABA by MeHg (ﬁlled
circles), MeHg plus ﬂumazenil (open diamonds), and MeHg plus diazepam, either with
MeHg (open circles) or diazepam pretreatment (ﬁlled triangles) or without (open
squares). Each bar represents the averaged values obtained from 4 - 5 cells. Time to

[GABA reduction by MeHg did not differ signiﬁcantly between any group tested (p>0.05).

69

 
 
  
    

 

 

 

 

o I
-+- MeHg
—9— MeHg Pre-treat

C: —i'- DZ Pre-treat
9 25 7 —°— MeHg+Flu
E MeHg+DZ
O

O r-

°\o 50

<

3

_0 75 -

1oo ' ' 7“. o‘. . .
0 10 20 30 40 50
Time (min)

E 607
E

C l
.9
‘6 40 r

3
'0

(D
0:
E 20 ‘

O
l—
9

(D 0
g MeHg MeHg DZ MeHg MeHg
I-

Pre- Pre- +Flu +DZ

Figure 2.8

70

Figure 2.9. Diazepam or Mel-lg pretreatment, following co-application, does not
prevent diazepam- or MeHg-induced prolongation of IGABA slow decay. A,
Representative traces depict effects of 1 uM diazepam (dashed line; 1510“,: 722.6 ms) and 1
uM MeHg plus diazepam (dotted line; Tslowt 897.2 ms), following diazepam pretreatment
(control, dark line; Tslowi 413.0 ms) on IGABA 15.0w. Current amplitudes in response to
diazepam or diazepam plus MeHg were scaled to those of control. B, Comparative
effects of 1 uM MeHg (dark grey bar) and co-application of 1 7.1M MeHg plus diazepam,
following diazepam pretreatment (0.1 uM — 10 71M, hashed bars) on prolongation of Tsjow
(control, black bar). Each bar represents the averaged values obtained from 4-5 cells.
Prolongation of IGABA slow decay produced by diazepam (0.1, 1.0, 10 M) alone was
signiﬁcantly different from control as indicated by the asterisk (*, p<0.05). Prolongation
of IGABA slow decay produced by the combination of MeHg plus 1 or 10 M diazepam,
following diazepam pretreatment, differed signiﬁcantly from that produced by diazepam
alone as indicated by the cross (T, p<0.05). C, Representative traces depict effects of
application of 1 uM MeHg alone (hashed line; rslow: 1451.5 ms; and co-application of 1
11M MeHg plus 1 uM diazepam, following MeHg pretreatment (dotted line, Tsjow: 1731.0
ms), on slow decay time constant of IGABA (control, dark line, 15.0w: 981.4 ms). Current
amplitudes in response to MeHg or MeHg plus diazepam were scaled to those of control.
D, Comparative effects of 1 uM MeHg (dark grey bar) and co-application of 1 pM MeHg
plus diazepam, following diazepam pretreatment (0.1 uM - 10 pM, hashed bars) on
prolongation of Tsjow (control, black bar). Each bar represents the averaged values

obtained from 6 - 8 cells. Prolongation of Tsjow produced by MeHg plus 1 or 10 M

71

diazepam, following MeHg pretreatment, differed signiﬁcantly from control and that

produced by MeHg alone as indicated by the asterisk (*, p<0.05).

72

  

'-
I'I’rsmw: 722.6 ms
ism: 897.2 ms

rm: 413.0 ms

  

 

    

........ .,_ .....,...... ..
in ..w . .. .. . . ..
.1 . ﬂ ...i.. .. . . .

 

. _
000000
55555

coacoo as as

10

1.0
[Diazepam] (uM)

Figure 2.9

73

GABA
O

 
   
    
 

... I'.‘
"' "I‘I‘q r

at .
'r.___;,.i_ Tslow' 1451.5 ms

' a‘” .
”mm“ *— Tsiow- 1731.0 ms

   

, 3,1!“
rslow: 981.4 ms

500 pA

 

 

 

 

 

 

 

 

 

 

D
s: W

 

 

 

0
Control MeHg 0.1 1.0 10

Figure 2.9

[Diazepam] (uM) + 1 11M MeHg

74

 
 

Figure 2.10. Co-application of MeHg and ﬂumazenil does not prevent MeHg-
induced prolongation of [GABA slow decay. A, The effects of diazepam on IGABA were
inhibited by 10 uM ﬂumazenil. IGABA were obtained as described in Figure 2.1. Current
traces were collected before (control), after exposure to diazepam (1 M), and after
exposure to diazepam plus ﬂumazenil. B, Flumazenil itself produced no effects on IGABA.
Currents were recorded before (control), after ﬂumazenil exposure, and after ﬂumazenil
plus MeHg exposure (at times indicated by arrows). In the presence of ﬂumazenil, MeHg
still decreased IGABA amplitude and caused a slowing of the decay rate. C, Representative
traces depict effects of co-application of 1 pM MeHg plus 10 11M ﬂumazenil (dotted line;
Tsjow: 1171.3 ms) on IGABA slow decay time constant (control, dark line; 15.0w: 1007.0 ms).
Current amplitudes in response to ﬂumazenil plus MeHg were scaled to those of control.
D, Comparative effects of 1 uM MeHg (dark grey bar) and co-application of MeHg plus
10 M ﬂumazenil (hashed bar) on prolongation of Tsjow (control, black bar). Each bar
represents the averaged values obtained from 4 - 5 cells. Prolongation of 12510“, produced
by MeHg plus ﬂumazenil did not differ signiﬁcantly from that produced by MeHg alone

(p>0.05).

75

A

 

 

 

 

Diazepam + Flu

 

 

 

 

 

Control Diazepam
603
71
Control Flu
150 s
ﬁr
Figure 2.10

76

 

120 s
i?
<2
0.
§L_
1000 ms
Flu + MeHg
600 s
n
i
<12
0.
O
O
O
" 1000 ms

 

:1171.3 ms

 

 

1000 ms

 

 

 

D
E100” %%

 

 

 

Control MeHg MeHg + Flu

Figure 2.10

77

shown in Figure 2.9C-D, under these conditions MeHg plus diazepam (1.0 and 10 M),
but not 0.1 pM diazepam, caused a signiﬁcantly greater prolongation of rslow as compared
to MeHg treatment alone (p<0.05). However, this effect did not reach signiﬁcance for
co-application with MeHg. Furthermore, co-application, following MeHg pretreatment,

did not alter the time course of [GABA reduction by MeHg (Fig. 2.8).

Flumazenil Does Not Alter the Effects of MeHg on [GABA

To determine further whether or not the MeHg interacts with the action or binding
site of benzodiazepines on the GABAA receptor, the benzodiazepine antagonist,
ﬂumazenil was co-applied with MeHg. The rationale was that if MeHg acts at the
benzodiazepine site of the receptor, then blocking the site with an antagonist should
reduce or prevent the effects of MeHg on IGABA. At the concentration of ﬂumazenil (10
M) used, prolongation of IGABA by diazepam (1 M) was abolished (Fig. 2.10A).
Flumazenil, in and of itself, had no effect on [GABA as shown in Figure 2.10B. However,
co-application of ﬂumazenil did not prevent MeHg from prolonging 1510“,; the effect of
combined treatment was not signiﬁcantly different from that produced by MeHg alone
(Fig. 2.10C-D) (p>0.05). In addition, ﬂumazenil did not alter the time course of IGABA
reduction by MeHg as shown in Figure 2.8 (p>0.05). These results show that ﬂumazenil

did not interfere with the effects of MeHg on 10A“.

78

DISCUSSION

The primary goals of the present study were to characterize the interaction of
MeHg with the at subunit-containing GABAA receptor, and speciﬁcally to determine
whether or not MeHg interacts with the benzodiazepine binding site of the receptor. We
found that MeHg (0.1, 1.0, and 10 uM) caused a progressive suppression of IGABA.
Kinetically, MeHg prolonged 15.0“, of IGABA in a concentration-independent manner.
Furthermore, MeHg did not appear to interact with the diazepam site of the GABAA
receptor, as co-application of MeHg (1 uM) and diazepam (1 and 10 11M) did not alter
the pattern of effects of diazepam alone or MeHg alone on IGABA, nor did the
benzodiazepine antagonist ﬂumazenil prevent or reduce the effect of MeHg on tslow,

At all concentrations of MeHg used, the pattern of effects of MeHg on IGABA was
similar —the current was suppressed completely and irreversibly. These ﬁndings are
qualitatively consistent with effects of MeHg on IGABA reported previously but occurred
at a much lower range of concentrations. In dorsal root ganglion neurons, high
concentrations of MeHg (100 uM) suppressed the peak currents induced by low
concentrations of GABA (Arakawa et al., 1991). Similarly, in hippocampal and
cerebellar slice, MeHg (20 and 100 uM) gradually decreased lPSPs to complete
suppression (Yuan and Atchison, 1997). The time to suppression of [GABA responses
differed between concentrations of MeHg tested, such that the time to suppression of
responses was prolonged at lower concentrations of MeHg as compared with higher
concentrations. The magnitude of effects produced by different concentrations of MeHg,
however, did not differ. This is consistent with what is known about the general pattern

of effects of MeHg. At each concentration of MeHg tested (0.1, 1.0, and 10 uM), the

79

time-course of reduction of IGABA was approximately linear and rsiow was prolonged to
approximately the same extent. The ﬁxed magnitude of effects caused by different
concentrations of MeHg suggests that a consistent series of actions takes place once an
effective or threshold concentration of MeHg is attained.

In addition to causing IGABA suppression, MeHg prolonged the slow component of
IGABA decay. GABAA receptor decay is important as it is a major determinant of IPSC
duration. The faster component of decay is associated with receptor desensitization,
whereas the slower component of decay is typically associated with GABA receptor
deactivation. Deactivation rate is inﬂuenced by unbinding of GABA from its receptor as
well as by transitions of activated receptors between open, closed, and desensitized states
(MacDonald et al., 1989a; Twyman et al., 1990; Twyman and MacDonald, 1992). Even
though removal of agonist is considered the ﬁnal step in current deactivation, it can be
prolonged by detaining the receptor in the desensitized state (Jones and Westbrook, 1995;
Tia et al., 1996; Haas and MacDonald, 1999). Unbinding is prolonged from the
desensitized state because the receptor needs ﬁrst to recover from the desensitized state to
reopen prior to unbinding of GABA. Thus, GABAA receptor deactivation can be shaped
by several different mechanisms. In the present study, the prolongation of Tsjow by MeHg
likely indicates that MeHg slows the rate of unbinding of GABA from its receptor.
Alternatively, it could indicate that MeHg produces an accumulation of desensitized
GABAA receptors, resulting in prolonged unbinding of GABA from its receptor. MeHg
did not appear to affect IGABA fast decay, suggesting that MeHg does not affect GABAA
receptor desensitization. However, the present experiments may not have resolved

appropriately the fast phases of decay, or desensitization, due to the use of a relatively

80

slow perfusion system. Thus, additional studies, such as recovery from desensitization
experiments measured by paired GABA pulses, are needed to determine speciﬁcally if
MeHg affects GABAA receptor desensitization.

MeHg most likely does not compete for the same binding site on the GABAA
receptor as does diazepam, because MeHg was able to maintain its effects on IGABA in the
presence of diazepam. However, it is also possible that diazepam did not interfere with
the actions of MeHg at the GABAA receptor because diazepam and MeHg were not
applied at concentrations high enough to saturate the receptor. As a result, the effects of
diazepam and MeHg would be additive, as neither drug might have reached its maximum
“ceiling” effect. However, even in the presence of ﬂumazenil, a benzodiazepine
antagonist, MeHg maintained its effects on IGABA, indicating that the effects of MeHg on
IGABA are independent of the actions of diazepam.

Although some binding studies showed previously that short-terrn (Corda et al.,
1981; Fonfria et al., 2001) as well as long-term administration (Concas et al., 1983) of
MeHg enhances benzodiazepine binding to the GABAA receptor, the results from the
present study do not necessarily contradict this. Increased binding of diazepam to the
GABAA receptor after MeHg exposure, does not necessarily translate into an increase in
GABAA receptor function or currents. Instead, increased binding of diazepam in
response to MeHg exposure could result from a compensatory response by the cell to
increase GABAA receptor binding sites due to antagonism of the GABAA receptor by
MeHg. Chronic administration of the benzodiazepine antagonist, R015-1788, to mice
caused an upregulation of benzodiazepine binding sites on the GABAA receptor in cortex,

cerebellum, and hippocampus (Miller et al., 1989), supporting the idea that an increase in

81

binding sites may result as a compensatory response to antagonist application.
Furthermore, experiments by Corda et a1. (1981) and Concas et a1. (1983) involved the
use of a whole animal model. The type of model may affect the way in which MeHg
interacts with the GABAA receptor, as MeHg acting in vivo may produce compensatory
effects in the whole animal not seen in cells in culture, resulting in an upregulation of
GABA... receptors in response to MeHg antagonism. In addition, Fonfria et a1. (2001)
used a mouse model, as opposed to rat which was used in the present investigation. This
is important because effects of MeHg can differ between mice and rats, as was
demonstrated by Basu et a1. (2005) in a study in which the inhibition of radioligand
binding to muscarinic acetylcholine receptors by MeHg was more pronounced in rat brain
as compared with mouse.

The present study suggests that MeHg does not interact with the diazepam site of
the GABAA receptor, thus alternative binding sites ought to be considered. Results from
our lab have indicated recently that MeHg does not bind to the channel pore of the
receptor, because stepping the voltage from ~80 mV to +60 mV in the presence of MeHg
produced a linear (voltage-independent) I-V relationship (Herden et al., 2004).
Furthermore, our lab has shown that MeHg does not appear to bind to the GABA site on
the receptor, as time to suppression of IGABA by MeHg was not dependent upon GABA
(10 pM — 1 mM) concentration in cortical cells in culture (Herden et al., 2004).
Moreover, our observation that MeHg prolongs [GABA slow decay may indicate that
MeHg interacts with GABAA receptor extracellular N-terminus structures acting though
transmembrane domain (TMl), as these structures have been shown to modulate current

deactivation (Bianchi et al., 2001). However, detailed molecular studies using GABAA

82

receptor-containing chimeras or exchange mutants between subunits are needed to
determine more accurately the potential binding site for MeHg on the GABAA receptor.
In summary, acute bath application of MeHg to rat cortical cells in primary
culture irreversibly suppressed IGABA and prolonged Tsjow. Co-application of diazepam or
ﬂumazenil either prior to or after MeHg exposure did not alter the patterns of effects of
MeHg on IGABA. Our results suggest that MeHg and diazepam appear to have
independent effects and that they most likely do not compete for the same binding site on
the GABAA receptor in cortical cells. Future studies are needed to determine whether or
not MeHg interacts with a site on the GABAA receptor belonging to a known drug or if it

has a unique binding site on the receptor.

83

CHAPTER THREE

DIFFERENTIAL EFFECTS OF METHYLMERCURY ON GABAA RECEPTOR
CURRENT S IN RAT CEREBELLAR GRAN ULE AND CEREBRAL CORTICAL

NEURONS IN CULTURE

84

ABSTRACT

Granule cells of the cerebellum are particularly sensitive to the inhibitory effects
of MeHg on GABAA receptor-mediated synaptic transmission. Suppression of IPSCs by
MeHg occurs more rapidly in granule cells than in neighboring Purkinje cells. The
mechanism(s) that underlies the differential sensitivity of GABAergic transmission to
MeHg in different cerebellar cell types is not known. A possible explanation is the
differential expression of a6 subunit-containing GABAA receptors in granule and Purkinje
neurons. To test this hypothesis, IGABA were recorded in response to MeHg in cerebellar
granule and cerebral cortical cells in culture. Cortical cells were used as a replacement
for Purkinje cells as they are more successfully cultured and express the same (11-
containing GABAA receptor phenotype as do Purkinje cells (Hansen et al., 2001). IGABA
were obtained by means of the whole-cell voltage patch-clamp technique, using a
symmetrical chloride concentration. MeHg suppressed completely IGABA in both cell
types in a time- and concentration-dependent manner. Suppression in granule cells
occurred more rapidly than did suppression in cortical cells. To verify these results,
IGABA were recorded in response to MeHg in cerebellar granule cells expressing either (16
or a. subunit-containing GABAA receptors. Suppression of IGABA by MeHg was similar
in granule cells containing either subtype of GABAA receptor. Similarly, the effects of
MeHg on [GABA were examined in HEK-293 cells expressing either (16 or (:1 subunit-
containing GABAA receptors. IGABA suppression by MeHg was comparable in both as
and a1 subunit-containing HEK-293 cells. These results suggest that the presence of the
(16 subunit alone may not underlie the differential effects of MeHg on [GABA observed in

cerebellar granule and cortical neurons; other factors may be involved as well.

85

INTRODUCTION

Methylmercury (MeHg) is a wide-spread environmental neurotoxicant. It has a
high afﬁnity for sulfhydryl groups numerous on cysteine-containing proteins. Thus,
MeHg has the potential to bind to cell membrane proteins and interfere with many
cellular processes (Chang, 1977; Atchison and Hare, 1994), including disruption of
excitatory and inhibitory synaptic transmission (Yuan and Atchison, 1993, 1995, 1997,
1999, 2003). GABAA receptor-mediated synaptic transmission is inhibited by MeHg. In
dorsal root ganglion neurons in culture, high concentrations of MeHg (100 uM) suppress
the peak currents induced by GABA (Arakawa et al., 1991). In hippocampal slice, MeHg
decreases gradually inhibitory postsynaptic potential (IPSP) amplitudes to complete
suppression (Yuan and Atchison, 1995). Suppression of inhibitory synaptic transmission
by low concentrations of MeHg occurs earlier than does suppression of glutamatergic
synaptic transmission (Yuan and Atchison, 1995, 1997), suggesting that inhibitory
synaptic transmission is more sensitive to suppression by MeHg than is excitatory
transmission. Interestingly, suppression of inhibitory post synaptic currents (IPSCs)
induced by bath-applied MeHg in cerebellar granule cells in brain slice occurs much
earlier than does suppression in neighboring Purkinje cells (Yuan and Atchison, 2003).
The mechanism(s) by which MeHg differentially affects GABAergic synaptic
transmission in cerebellar cells is not known.

One possibility for this differential sensitivity to MeHg may be the differential
expression of GABAA receptor phenotypes in granule and Purkinje cells. Purkinje cells
express the ail-containing receptor. Granule cells are the only neurons in the cerebellum

which express the (16 subunit-containing GABAA receptor, although they too express the

86

a; subunit. Expression of the (1., subunit is regulated developmentally both in vivo and in
vitro and studies have shown that there is an increased expression of a6 subunits in rats
during maturation (Laurie et al., 1992, Thompson et al., 1994). GABAA receptors
containing the (16 or a] subunit have unique pharmacological and biophysical properties,
including differential sensitivity to agonists, such as benzodiazepines (Luddens and
Wisden, 1991; Sieghart, 1992; Makela et al., 1997) and barbiturates (Fisher etal., 1997;
Cestari et al., 2000) and to antagonists such as, zinc (Draguhn et al., 1990; Zempel and
Steinbach, 1995; Saxena and MacDonald, 1994, 1996), furosemide (Korpi et al., 1995a),
and lanthanum (Saxena et al., 1997; Makela et al., 1999). However, whether or not the
expression of the (16 subunit-containing GABAA receptor contributes to the differential
sensitivity of cerebellar neurons to MeHg remains to be determined. The present study
was designed speciﬁcally to examine the a subunit speciﬁcity of the effects of MeHg by
using the whole-cell patch-clamp technique to compare the effects of MeHg on IGABA in

cells expressing either a. or 0.6 subunit-containing GABAA receptors.

87

METHODS

Solutions and Chemicals

Methylmercuric chloride (MeHg) (ICN Biomedical Inc., Costa Mesa, CA, USA)
was dissolved in deionized water to a ﬁnal concentration of 10 mM, which served as a
stock solution. On the day of experiments, MeHg working solutions (0.1, 1, or 10 uM)
were diluted in extracellular solution consisting of (in mM): NaCl, 125; CaCl2, 2.0; KCl,
2.5; MgCl2, 1.0; KH2PO4, 1.25; NaHCOg, 26.0, and D-glucose, 20.0, pH-adjusted to 7.4
using 95% 02/5% CO2. In the present studies, three different concentrations of MeHg
(0.1, 1, or 10 11M) were used. These concentrations of MeHg were used as they are
within the clinically relevant range of concentrations found in victims poisoned with
MeHg in Iraq in the 1970’s (Bakir et al., 1973) (see Chapter One). Diazepam, 7-amino-
n-butyric acid (GABA), 6-cyano-7-nitroquinoxaline-2,3-dione disodium salt (CNQX),
DL-Z-amino-S-phosphonopentanoic acid (APV), glutamine, 4-(2-hydroxyethyl)piperazine-
l-ethanesulfonic acid (HEPES), gentamycin, deoxyribonuclease (DNase I), cytosine B-D-
arabino-furanoside (Ara-C), adenosine 5'- triphosphate magnesium salt (ATP), ethylene
glycol-bis(B-aminoethyl ether)-N, N, N', N',-tetraacetic acid (EGTA), trypsin, poly-L-
lysine, picric acid, and formaldehyde (37% solution) were all purchased from Sigma
Chemical Co. (St. Louis, MO, USA). Basal Eagle’s Medium, Modiﬁed Eagle’s Medium
and Dulbecco’s Modiﬁed Eagle’s Medium were purchased from Mediatech, Inc.
(Hemdon, VA). Heat inactivated horse serum, fetal bovine serum, sodium pyruvate, non-
essential amino acids, antimycotic and antibiotics, and Optimem were purchased from
GIBCO-Invitrogen Inc. (Carlsbad, CA). Qiagen kits, used for plasmid puriﬁcation, were

purchased from Qiagen Inc. (Valencia, CA) and Fugene 6 was purchased from Roche

88

Molecular Biochemicals (Indianapolis, IN). Secondary antibodies conjugated with
tetramethylrhodamine (TRITC) or ﬂuorescein isothiocyanate (FITC) were purchased
from Jackson ImmunoResearch Inc. (West Grove, PA). Secondary antibody conjugated
with paciﬁc blue was purchased from Invitrogen (Carlsbad, CA). Diazepam was
dissolved in dimethyl sulphoxide (DMSO; < 0.01% (v/v) ﬁnal concentration; a
concentration low enough not to affect 10A“) and diluted in ACSF prior to use. Other

drugs were dissolved in deionized water and diluted in extracellular solution prior to use.

Preparation of Primary Cerebellar Granule Cell Culture

Primary cultures of rat cerebellar granule neurons were prepared from 7-day-old
Sprague-Dawley rats of either gender (Charles River Laboratories, Wilmington, MA)
following procedures described by Gallo et al. (1987). Following extraction of cerebella,
cells were digested for 13 min at 37° C with trypsin 0.025 % (w/v) and plated at a density
of l x 106 cells/ml on 35-mm Petri dishes (Corning Inc., Corning, NY) coated with poly-
L-lysine (0.1 mg/ml). Cells were cultured in Basal Eagle’s Medium supplemented with
10% (w/v) fetal bovine serum, 2 mM glutamine, and 100 ug/ml gentamycin. The ﬁnal
concentration of KCl in the culture medium was adjusted to 25 mM. In order to achieve
functional synapse formation (Chen et al., 2000; Prybylowski et al., 2002), at 4 days in
vitro the medium was replaced with 5 mM KCl-containing Modiﬁed Eagle’s Medium
supplemented with 5 mg/ml glucose, 0.1 mg/ml transferrin, 0.025 mg/ml insulin, 2 mM
glutamine, and 20 jig/ml gentamycin, in addition to cytosine arabinoside (10 M) to
inhibit glial cell proliferation. Cells were maintained in a 37° C incubator with 95% O2

and 5% CO2. Cells were used for recordings 3-8 days after plating, depending upon the

89

aim of the experiment, to allow for maturation and expression of a6 and Ill-subunit

containing GABAA receptors (Laurie et al., 1992a).

Preparation of Primary Cerebral Cortical Cell Culture

Cortical cells were used as a substitute for cerebellar Purkinje cells, as Purkinje
cells are extremely difﬁcult to maintain in culture and produce a low yield. Furthermore,
they frequently become contaminated with other cell types, particularly glial cells, and
MeHg is known to have inhibitory effects on astrocytes (Aschner et al., 2000).
Importantly, cortical cells are easier to maintain in culture and they contain the same type
of GABAA receptor phenotype as do Purkinje cells (Hansen et al., 2001). Primary
cultures of cortical neurons were obtained from newborn Sprague-Dawley rat pups of
either gender (Charles River Laboratories, Wilmington, MA) following methods
described by Ingleﬁeld et a1. (2001). Brieﬂy, following removal of the brain and
isolation of cortex, cells were digested for 4.5 min at 37° C with trypsin 0.025 % (w/v) in
buffer that contained (in mM): KCl (5.0), KH2PO4 (0.205), NaCl (137.0), Na2HP04
(0.17), D-glucose (5.0), sucrose (59.0), and gentamycin (0.1 mg/ml), pH 7.3. Trypsin
was inactivated by addition of the above-described buffer containing 0.016 % (w/v)
DNase I for 4.5 min at 37° C. To this mixture warmed Dulbecco’s Modiﬁed Eagle’s
Medium (DMEM) supplemented with 10 % (w/v) horse serum, 10 mM HEPES, 2 mM
glutamine, and gentamycin (0.1 mg/ml) was added and the cells were centrifuged (500xg,
5 min). The resulting pellet was resuspended in DMEM containing DNase I and was
centrifuged again. The cell suspension was then resuspended in DMEM and triturated
with a Pasteur pipette. Cells were plated at a density of 1 x 106 cells/ dish onto 35-min

Petri dishes (Corning Inc., Corning, NY) coated with poly-L-lysine. Cells were

90

maintained in a 37° C incubator with 95% O2 and 5% CO2. DMEM was replaced every
two days. Cytosine B-arabinoside (5 pM) was added to cell cultures 72 hr after plating to
inhibit glial cell proliferation. Cells were used for recordings 10-14 days after plating.
All experiments were done in compliance with NIH and university standards and were

approved by the Michigan State University Committee of Animal Use and Care.

Culture and Transfection of HEK-293 Cells

As described previously (Peng et al., 2002), HEK 293 cells (#CRL-1573;
American Type Culture Collection, Rockville, MD) were lightly trypsinized, centrifuged
and plated at ~ 50 - 100,000 cells in 2 ml DMEM fortiﬁed with 1 mM sodium puruvate,
0.1 mM non-essential amino acids, 10% (w/v) fetal bovine serum, antibiotics and
antimycotics and grown overnight at 37° C in 5% C02. The transient expressions were
achieved by transfecting the cells one day after plating with 1 pg total DNA (1:1:1 molar
ratio of a, B and 7 and 1/5 of a GFP plasmid, all cloned in pCDNA3.1). Plasmids
containing rat or], B2 and 72 subunit cDNA were obtained from Dr. C. Czajkowski
(University of Wisconsin, Madison,WI) and plasmids containing rat a6 subunit cDNA
were obtained from Dr. W. Wisden (University of Heidelberg, Heidelberg, Germany).
All plasmids were puriﬁed using Qiagen kits (endotoxin free). Optimem (96 p1) was
added together with Fugene 6 (3 pl) and DNA (1 pg) to maintain an anion - cation ratio
of 1:1. The mixture was incubated for 20 min. and then added to dishes. Green
ﬂuorescent cells were seen starting 24 hrs after treatment, and robust expression was
detected approximately 40 - 72 hrs post treatment. On the day of recording, the cells were

lightly trypsinized, centrifuged and replated at 1/3 - 1/6 dilution on poly-L-lysine coated

91

cover glasses in 35 mm culture dishes at a low density with good spatial separation to

permit recordings. Recordings were made 2 - 8 hrs following plating.

lmmunocytochemistry

Cover slips containing HEK-293 cells were incubated overnight in ice-cold
phosphate buffered saline (PBS) containing 0.02% NaN3 to kill cells immediately without
permeabilizing their membranes. Following 2 rinses with PBS, HEK-293 cell were
labeled overnight in the cold room with primary anti-a; (1:7000) antibody (Novus
Biologicals Inc., Littleton, CO) or anti-a6 (1:1000) antibody (Novus Biologicals Inc.,
Littleton, CO), both generated in rabbits. In contrast, cover slips containing granule cells
from 2-8 days in vitro (DIV) and cortical cells (DIV 10) were labeled overnight with anti-
(11 (1:7000) or anti-d6 (1:1000) antibody following a 5 min. ﬁxation in ice-cold acetone
and 3 rinses in PBS. All cells were incubated with a tetramethylrhodamine (TRITC;
1:500)-conjugated anti-rabbit or ﬂuorescein isothiocyanate (FITC; 1:500)-conjugated
anti-rabbit secondary antibody for 2 hrs at 4°C and staining was visualized using a Leitz
Laborlux S (Wetzlar, Germany) epiﬂuorescent microscope and a 20x or 60x oil
immersion objective. Negative controls included incubation of cells with primary or
secondary antibody only, in addition to labeling cells with antibodies against subunits
which they do not express (i.e. labeling cortical cells with anti-06 antibody or labeling
HEK-293 cells expressing only (16 subunit-containing GABAA receptors with antibody
against the a; subunit). No ﬂuorescence was seen in any of the control experiments,
except a minute amount of cross-reactivity was detected with the anti-a1 antibody in

HEK-293 cells expressing (16 subunit-containing GABAA receptors. This cross-reactivity

92

was not detected in granule cells. Absorption controls with the peptide and antibodies
could not be performed, as the peptides were not provided with the antibodies. To
conﬁrm that labeling of anti— a] and anti-a6 antibodies was localized to the cell
membrane, cover slips containing HEK-293 cells were incubated for 1 hr in Zamboni
ﬁxative containing (in ml): picric acid (15), formaldehyde-37% solution (5.5), and PBS
(79.5). Following ﬁxation, cover slips were rinsed 2 times with PBS and labeled
overnight in the cold room with primary anti-a1 (1:7000) antibody or anti—a6 (1:1000)
antibody, both generated in rabbits, or with primary anti-pan cadherin (1:500) antibody,
generated in mice. Cells were incubated with (TRITC; l:200)-conjugated anti-rabbit,
(FITC; 1:200)-conjugated anti-rabbit, or paciﬁc blue (1:200)-conjugated anti-mouse
secondary antibody for 2 hrs at 4°C. Staining was visualized using a Nikon Eclipse
2000-UDiaphot-TMD microscope (Nikon Optics, Tokyo, Japan) and a 20x or 40x

objective, as indicated.

Whole-cell Recordings

Methods are identical to those described in Chapter Two.

Data Acquisition and Analysis

Methods are identical to those described in Chapter Two.

93

 

RESULTS

MeHg Suppresses [GABA in Cerebellar Granule Cells

To determine the subunit speciﬁcity of effects of MeHg on 10AM, whole-cell
recordings were made from (16 subunit-containing granule cells and a1 subunit-containing
cortical cells. First, to verify the GABAA receptor a subunit expression in both cell types,
immunocytochemistry was performed using antibodies against the (11 mm; subunit of the
receptor. As expected, both granule and cortical cells labeled positively for the a; subunit
of the GABAA receptor (Fig. 3.1 A-B), whereas only granule cells labeled positively for
the (16 subunit of the receptor (Fig. 3.1 CD). In both cell types the antibody staining
appeared to be localized primarily to the cell body, and to a lesser extent in neuronal
processes. The cell body is stained in a punctate pattern and in some cells it was evident
that the bulk of the staining is in the membrane periphery, suggesting that staining is
primarily localized to the membrane surface.

Under symmetrical Cl' concentrations, application of a GABA pulse produced
an inward current in granule cells, which was gradually abolished by bath application of
MeHg (0.1 — 10 pM) (Fig. 3.2). The effects of MeHg never reached a steady-state level
less than complete reduction of IGABA (Fig. 3.2 A). Suppression of [GABA by MeHg (1
pM) was not reversible by washing cells for ~ 20 min with a MeHg-free solution or a
solution containing D-penicillamine (20 pM), a MeHg chelator (Appendix, Fig. A. 1).
The time course of reduction of IGABA by MeHg is shown in Figure 3.2 B. Increasing the
concentration of MeHg shortened the time course of suppression. For each concentration
of MeHg tested, the time-course of suppression by MeHg was approximately linear so

that 50% IGABA reduction occurred at about 50% of the total time to reduction. Time to

94

Figure 3.1. GABAA receptpr a6 and a; subunit expression in cerebellar granule and
cerebral cortical cells in culture. Granule (DIV 6) and cortical cells (DIV 10) were
labeled overnight with anti-q] (A-B) or anti-a6 (C-D) antibody following acetone ﬁxation.
Antibody staining was visualized with TRITC using a Leitz epiﬂuorescent microscope
and a 60x oil immersion objective. Note that only cerebellar granule cells express a6
subunit-containing GABAA receptors (C), whereas both granule and cortical cells express

a1 subunit-containing receptors (A -B). Images in this thesis are presented in color.

95

Granule Cortical

 

Figure 3.1

96

Figure 3.2. MeHg causes a time- and concentration-dependent suppression of [GABA
in granule cells. A, MeHg caused a progressive and complete suppression of IGABA in rat
granule cells in primary culture. Whole-cell [GABA were evoked from a holding potential
of - 60 mV by a 10 ms pulse of GABA (50 pM, black dot), at intervals of 30 s, in the
presence of CNQX (10 pM) and APV (50 pM) in the external solution to block
glutamatergic currents. Data were collected before (control) and at various time points
during exposure to 1 pM MeHg (6, 8, 12, and 20 min). B, Time course of effects of
MeHg on [GABA recorded under identical conditions to those described above at [MeHg]
of 0.1 pM (squares), 1.0 pM (triangles), or 10 pM (circles). Data were collected
continually before and after MeHg exposure. Each datum point represents the mean
value recorded from 3-5 cells. C, Comparative effects of various concentrations of
MeHg on time to complete reduction of IGABA. Each bar represents the mean values
obtained from 3-5 cells. The times to total reduction of IGABA by MeHg differed
signiﬁcantly between 0.1 and 1.0 pM as indicated by an asterisks (*), and 0.1 and 10 pM
MeHg, as indicated by a dagger (1'). No signiﬁcant difference was observed between 1.0

and 10 pM MeHg (p>0.05).

97

A

20 min —>

GABA
0

12min —->
8min —>

6 min

 

400 DA

 

2000 ms
Control
(0 min) —>

 

B

|GABA (°/o Control)

 

 

 

Time (min)

Figure 3.2

98

 

10

 

ass

 

 

 

 

 

[MeHg] (MM)

 

 

L
O O O O
6 4 2

Figure 3.2

C 2:5 cozonoom ES 9 9::

99

Figure 3.3. MeHg causes a time- and concentration-dependent suppression of [GABA
in cortical cells. A, MeHg caused a progressive and complete suppression of IGABA in rat
cortical cells in primary culture. Whole-cell [GABA were recorded under identical
conditions to those described in Figure 3.2. Data were collected before (control) and at
various time points during exposure to 1 pM MeHg (5, 8, 11, 18, and 37 min). B, Time
course of effects of MeHg on IGABA recorded at [MeHg] of 0.1 pM (open squares), 1.0
pM (ﬁlled triangles), or 10 pM (open circles). Data were collected continually before
and after MeHg exposure. Each datum point represents the mean value recorded from 3-
5 cells. C, Comparative effects of different concentrations of MeHg on time to complete
reduction of IGABA. Each bar represents the mean (:1: SEM) values obtained from 3-5
cells. The times to total reduction of IGABA by MeHg differed signiﬁcantly between 0.1
and 1.0 pM as indicated by an asterisks (*), and 0.1 and 10 pM MeHg, as indicated by a
dagger (1'). No signiﬁcant difference was observed between 1.0 and 10 pM MeHg
(p>0.05). D, Comparative effects of MeHg on suppression of [GABA in cortical (dark
bars) and granule cells (light bars) in culture. IGABA was suppressed signiﬁcantly faster
(p<0.05) in granule cells as compared with cortical cells for 1.0 and 10 pM MeHg, as

indicated by the asterisks.

100

GABA

37 min
18 min
11 min
8 min

5 min

 

Control —> 1000 m5
(0 min)

|GABA (% Control)

 

 

 

Time (min)

Figure 3.3

101

 

 

 

 

 

_ _
0 0 0 0
4 2

C 3:: cozosoom ES 9 95

 

[MeHg] (11M)

l:l Granule

W Cortical

 

D

 

 

60‘

3:5 cozonoom ES 2 9::

 

[MeHg] (MM)

Figure 3.3

102

total reduction of IGABA by MeHg differed signiﬁcantly between each concentration of
MeHg tested (p<0.05) (Fig. 3.2 C). At 10 pM MeHg, the highest concentration tested,

complete reduction of [GABA occurred within approximately 17 min.

MeHg Suppresses [GABA in Cerebral Cortical Cells

Similar to the effects of MeHg observed in cerebellar granule cells, bath
application of MeHg (0.1 - 10 pM) produced a gradual and complete suppression of
IGABA in cortical cells (Fig. 3.3 A). The time course of reduction of IGABA by MeHg in
cortical cells is shown in Figure 3.3 B. Decreasing the concentration of MeHg prolonged
the time course of suppression. For each concentration of MeHg tested, the time-course
of reduction by MeHg was approximately linear. Time to total reduction of [GABA by
MeHg differed signiﬁcantly between each concentration of MeHg tested (p<0.05) (Fig.
3.3 C). At 10 pM MeHg, the highest concentration tested, complete reduction of IGABA
occurred within approximately 30 min. Interestingly, at 1 and 10 pM MeHg, the time to
suppression of [GABA was signiﬁcantly less in granule cells as compared with cortical
cells (Fig. 3.3 D), however this difference was not observed at 0.1 pM MeHg. In
addition, stepping the voltage from ~80 mV to +60 mV in the presence of MeHg
produced a linear current-voltage relationship in granule cells (Fig. 3.4 A) as well as in
cortical cells (Fig. 3.4 B), suggesting that suppression of IGABA by MeHg is voltage-

independent in both cell types.

103

Figure 3.4. Effects of MeHg on IGABA are voltage-independent in granule and
cortical cells. A, Whole-cell IGABA were recorded from cerebellar granule cells before
(control) and after 5 min. exposure to 1 pM MeHg at holding potentials ranging from -80
to +60 mV. Each datum point represents the mean value recorded from 3-4 cells. MeHg
caused a reduction in IGABA amplitude in a linear fashion, such that it did not change at
different holding potentials. B, IGABA were recorded from cerebral cortical cells before
(control) and after 10 min. exposure to 1 pM MeHg at holding potentials ranging from -
80 to +60 mV. Each datum point represents the mean value recorded from 3 cells.
MeHg caused a reduction in [GABA amplitude in cortical cells in a voltage-independent

manner.

104

.s
U"
1

Control

A
O
r

” MeHg

I I I l

 

O

I
01
I

...L
C
I

 

 

_'.
01
If

-80 -60 -40 -20 0 20 40 60

Voltage (mV)

Normalized Current (pA) >

OD
O

20 . Control

 

 

 

 

Normalized Current (pA) w

-80 so -40 -2o 0 20 40 so
Voltage (mV)

Figure 3.4

105

Figure 3.5. GABAA receptor an and (:1 subunit expression in cerebellar granule cells
at different days in culture. Granule cells (DIV 2-8) were labeled overnight with anti-
0.; or anti-(16 antibody following acetone ﬁxation. Antibody staining was visualized with
TRITC (anti-a1) or FITC (anti-06) using a Leitz epiﬂuorescent microscope and a 60x oil
immersion objective. A, (11 subunit-containing GABAA receptors are expressed in granule
cells grown in vitro throughout days 2-8 (DIV 2-8). B, 016 subunit-containing receptors

are not expressed in granule cells until DIV 6 and 8.

106

DIV 2 DIV 4

1 1 pm

   

38 pm

DIV 6 DIV 8

 

38 pm 15 pm

Figure 3.5

107

B Anti-a6

DIV 2 DIV 4

   

38 pm 38 pm

DIV 6 DIV 8

 

Figure 3.5

108

m ‘

Figure 3.6. Effects of MeHg on granule cells at different days in culture. Time
course of effects of MeHg on IGABA were obtained under identical conditions to those
described in Figure 3.1. IGABA were recorded from granule cells grown in culture for 4 or
6-8 days in the presence of [MeHg] of 0.1 pM (A), 1.0 pM (B), or 10 pM (C). Data were
collected continually before and after MeHg exposure. Each datum point represents the
mean value recorded from 3-4 cells. D, Comparative effects of different concentrations
of MeHg on time to complete suppression of IGABA in granule cells grown for 4 or 6-8
days in culture. Each bar represents the mean (:1: SEM) values obtained from 3-4 cells.
The times to total reduction of IGABA by MeHg did not differ signiﬁcantly between cells

grown for 4 or 6-8 days in culture.

109

>

100 ”

IGABA Reduction (% Control)
CD
0

 

0.1 pM MeHg

—l— DIV4

+ DIV 6-8

I l

 

 

 

40 —
20 "
0
0
2 100 —
9
E _
o 80
O
23 so -
c
.9
*5 40
z
'8
a: 20
<
2
_o 0

Figure 3.6

 

 

8 16 24 32 40
Time (min)
1.0 pM MeHg
——I— DIV4 + DIV 6-8
6 12 18 24
Time (min)

110

10 14M MeHg

C

 

    

+ DIV 6-3

 

_ _,_ DIV4

 

000000
00000

99:50 at cozozumm <96.

       

n\\\\\\\\\\\\\\\\\\\km

 

000000
55555

D Eé 838.com ES 2 9::

6

am
e
m

.015
F

111

The Effects of MeHg Do Not Differ in Granule Cells Expressing Only a; or a
Combination of a1 and a6 Subunit-Containing GABAA Receptors

Following cerebellar granule cell culture, cells were ﬁxed and labeled with
antibodies against GABAA receptor on or (16 subunits at different days in culture to
determine the time-course of a subunit expression. As seen in Figure 3.5 A, expression of
on subunits was observed on the ﬁrst day of antibody labeling (day 2 in culture) and
continued through to the last day of labeling (day 8 in culture). In contrast, expression of
a6 subunits was not observed until day 6 in culture, and expression continued through to
day 8 in culture (Fig. 3.5 B). The staining for either antibody appeared to be primarily
localized to the cell body in a punctate pattern. In some cells, the focal plane shows that
much of the staining is localized to the periphery of the cell body, suggesting cell surface
labeling. Staining was also observed, to a lesser degree, in neuronal processes. GABAA
receptor-mediated currents were examined from cells from day 4 in culture to represent
til-containing GABAA receptor responses, and from day 6-8 in culture to represent (16-
containing responses. At each day in culture tested, MeHg produced a gradual and
complete reduction of IGABA (Fig. 3.6). This suppression was concentration- and time-
dependent. The time course of suppression by MeHg did not differ between days 4 and
6-8 in culture as seen in Figures 3.6 A-D. For 10 pM MeHg, the highest concentration
tested, the average time to suppression of IGABA in granule cells from day 4 in culture was

20.8 :1: 2 min, and for cells from 6-8 days in culture it was 21.3 :1: 3 min.

112

MeHg Suppresses [GABA in HEK-293 Cells Expressing (11 or (16 Subunit-Containing
GABAA Receptors

HEK-293 cells were used in these experiments, as these nonexcitable cells are
commonly used for heterologous expression of membrane proteins including GABAA
receptors (Saxena, 2000; Bianchi and MacDonald, 2002; Hinkle and MacDonald, 2003;
J ones-Davis et al., 2005). First, to verify that HEK-293 cells expressed the appropriate
subunit subtype of interest, cells were treated and labeled with antibody against an or (16
subunit—containing GABAA receptors as seen in Figure 3.7. As expected, only HEK-293
cells transfected with cDN A for the a; subunit-containing GABAA receptor labeled
positive for the a] subunit and not the (16 subunit (Fig. 3.7 A-B). Moreover, HEK-293
cells transfected with cDN A for the (16 subunit-containing GABAA receptor labeled
positive for the (16 subunit and not the (:1 subunit (Fig. 3.7 CD). A slight amount of
ﬂuorescence to anti-d1 was detected for the (16 subunit-containing cells (Fig. 3.7 C),
suggesting that non-speciﬁc binding or a small amount of cross-reactivity for the anti-a1
antibody occurred, perhaps with auxiliary subunits such as [32 or 72. The pattern of
staining of HEK-293 cells for both antibodies appeared punctate and primarily localized
to the periphery of the cell body, suggesting membrane surface staining. To conﬁrm
membrane surface staining, co-localization experiments were conducted in HEK-293
cells with a primary anti—pan cadherin (membrane protein) antibody and a primary anti-a6
or anti-a1 antibody (Fig. 3.8 — 3.10). As shown in Figure 3.8, HEK-293 cells transfected
with mRNA for (xi—containing GABAA receptors in addition to green ﬂuorescence protein
(GFP), stained positive for GFP (Fig. 3.8 B). Note that the transfection was efﬁcient in

approximately 25 % of cells. Most cells, both transfected and untransfected, stained

113

 

Figure 3.7. GABAA receptor «6 and a; subunit expression in HEK-293 cells. HEK-
293 cells were labeled overnight with anti-a1 or anti-r16 antibody following killing by
NaN3, to prevent membrane perrneabilization. Antibody staining was visualized with
TRITC using a Leica epiﬂuorescent microscope and a 60x oil immersion objective.
HEK-293 cells expressing (11 subunit-containing GABAA receptors labeled positive for
the (:1 subunit (A) and negative for the (:6 subunit (B). HEK-293 cells expressing a6
subunit-containing GABAA receptors labeled negative for the on subunit (C) and positive
for the (16 subunit (D). Note, a slight amount of ﬂuorescence was detected in C,
suggesting that a small amount of cross-reactivity or non-selective binding for the anti-u]

antibody may exist.

114

 

   

Anti-0t6

Figure 3.7

115

Figure 3.8. GABAA receptor membrane localization in (11 subunit-containing HEK-
293 cells. HEK-293 cells were labeled overnight with anti-a1 or anti-pan cadherin
antibody following ﬁxation. Antibody staining was visualized with TRITC or paciﬁc blue
using a Nikon epiﬂuorescent microscope and a 20x or 40x (as indicated) objective.
HEK-293 cells expressing a1 subunit-containing GABAA receptors labeled positive for
GFP protein (A) in approximately 25% of cells (B). All cells labeled positive for the
membrane protein cadherin (paciﬁc blue) (C). A lesser number of cells stained positive
for the on subunit (TRITC) (D). Co-localization of cadherin and the al subunit-
containing GABAA receptor in these cells is evident at 20x (E) and (40x) (F)

magniﬁcation.

116

 

      

GFP GFP + Brightfield

 

Cadherin 0‘1

30 pm

Cadherin + a1 Cadherin + a1 (40 x)

   

Figure 3.8

117

Figure 3.9. GABAA receptor membrane localization in an subunit-containing HEK-
293 cells. HEK-293 cells were labeled overnight with anti-a6 or anti-pan caherin antibody
following ﬁxation. Antibody staining was visualized with TRITC or paciﬁc blue using a
Nikon epiﬂuorescent microscope and a 20x or 40x (as indicated) objective. HEK-293
cells expressing r16 subunit-containing GABAA receptors labeled positive for GFP protein
(A) in approximately 25% of cells (B). All cells labeled positive for the membrane
protein cadherin (paciﬁc blue) (C). A lesser number of cells stained positive for the (16
subunit (TRITC) (D). Co-localization of cadherin and the cur, subunit-containing GABAA

receptor in cells is evident at 20x (E) and (40x) (F) magniﬁcation.

118

HEK-oi6

.-

GFP + Brightfield

C

   

Cadherin

   

Cadherin + a6 Cadherin + a6 (40 X)

Figure 3.9

119

Figure 3.10. Control Secondary antibody staining in HEK-293 cells. HEK-293 cells
were labeled overnight with PBS not containing primary antibody following ﬁxation.
Secondary antibody staining (TRITC or paciﬁc blue) was visualized using a Nikon
epiﬂuorescent microscope and a 20x objective. HEK-293 cells expressing a1 subunit-
containing GABAA receptors labeled positive for GFP protein in approximately 30% of
cells (A). Secondary antibody staining was also detected lightly in a few transfected cells
(paciﬁc blue) (B). HEK-293 cells expressing a; subunit-containing GABAA receptors
labeled positive for GFP protein in approximately 25 % of cells (C). Slight secondary

antibody staining was detected in a few transfected cells (TRITC) (D).

120

Controls

 

GFP + Brightfield Pacific blue only

C

      

GFP + Brightfield TRITC only

Figure 3.10

121

Figure 3.11. Effects of diazepam on [GABA in HEK-293 cells. Whole-cell [GABA were
evoked from a holding potential of - 60 mV by a 10 ms pulse of GABA (500 pM, black
dot), at intervals of 30 5, under similar recording conditions to those described in Figure
1. Current traces were collected before (control) and after exposure to diazepam (1 pM)
at times indicated by arrows. A, Diazepam prolonged IGABA slow decay rate in HEK-293
cells expressing (11 subunit-containing GABAA receptors as can be seen readily by the
superimposed control and diazepam traces (right). B, Diazepam did not affect IGABA slow

decay rate in HEK-293 cells expressing a6 subunit-containing GABAA receptors.

122

I-

A

(in-containing

Control 605 Diazepam Overlay<t

1000 ms

B (rs-containing

Control 1205 Diazepam Overlay

V V Vi..-

 

 

 

Figure 3.11

123

positive for the presence of the cadherin membrane protein (paciﬁc blue) (Fig. 3.8 C). In
contrast, a smaller number of cells stained positive for the til-containing GABAA receptor
(TRITC) (Fig 3.8 D). Co-localization of cadherin and til-containing GABAA receptor is
shown in Figure 3.8 E-F, and is discernible by a pink/purple color. Similarly, in HEK-
293 cells transfected with cDN A for dig-containing GABAA receptors, co-localization of
cadherin and (1.5-containing receptors was also evident (Fig. 3.9 A-F). In control '
experiments in which only the secondary antibody was applied (Fig. 3.10 A-D), a few
cells ﬂuoresced, however this ﬂuorescence was localized to transfected cells only.

These ﬁnding suggest that binding of the anti-a6 or anti-a1 antibody is localized to the
cell membrane. In addition to immunocytochemistry, subunit subtype expression was
corroborated pharrnacologically using diazepam (1 pM), a GABAA receptor agonist that
selectively potentiates currents from receptors containing the a; subunit and does not
affect responses from receptors containing the (16 subunit. As seen in Figure 3.11 A,
diazepam prolonged the slow decay time constant of IGABA in HEK-293 cells expressing
the (11 subunit. However, in HEK-293 cells expressing the (16 subunit, no effects of
diazepam on long». were detected (Fig. 3.11 B).

Results obtained in cerebellar neurons in culture suggested that there was no
differential effect of MeHg on [GABA in cells expressing a different compliment of a
subunit. To conﬁrm the results, whole-cell recordings of [GABA in response to MeHg
were made in HEK-293 cells transiently transfected with cDNA for either (11 or (16
subunit—containing GABAA receptors. In HEK-293 cells expressing either GABAA
receptor subtype MeHg produced a time- and concentration-dependent suppression of

IGABA (Fig. 3.12). For each concentration of MeHg tested, the time-course of MeHg

124

 

suppression was relatively linear for HEK-293 cells expressing either on or (:6 subunit-

 

containing GABAA receptors (Fig. 3.12 A). The time to complete reduction of IGABA by
MeHg (1.0 & 10 pM) did not differ signiﬁcantly between HEK-293 cells containing
either a; or (16 subunit—containing receptors as shown in Figure 3.12 B. Because HEK-
293 cells ﬂattened out and could not be maintained in the whole-cell conﬁguration for
long enough to achieve total IGABA suppression by 0.1 M MeHg, 40 % IGABA reduction
was used to compare the effects of all [MeHg] tested. As shown in Fig. 3.12 C, even at
40 % IGABA reduction by MeHg (0.1 ~ 10 pM) no signiﬁcant difference was observed
between HEK-293 cells expressing either (11 or a6 subunit—containing GABAA receptors.
A summary of the times to total or 40% IGABA reduction by MeHg is shown in Table 3.1.
In addition, stepping the voltage from ~80 mV to +60 mV in the presence of MeHg
produced a linear current-voltage relationship in HEK-293 cells expressing a6 subunit—
containing GABAA receptors (Fig. 3.13 A) as well as in HEK-293 cells expressing a;
subunit—containing GABAA receptors (Fig. 3.13 B), suggesting that the effects of MeHg

on IGABA in these cells are voltage-independent.

125

Figure 3.12. MeHg causes a time- and concentration-dependent suppression of
IGABA in HEK-293 cells. A, Time course of effects of MeHg on IGABA recorded under
identical conditions to those described in Figure 8. IGABA were recorded in the presence
of [MeHg] of 0.1 pM (squares), 1.0 pM (triangles), or 10 pM (circles) in HEK-293 cells
expressing either a; (dotted line, open symbols) or (16 (solid line, solid symbols) subunit-
containing GABAA receptors. Data were collected continually before and during MeHg
exposure. Each datum point represents the mean value recorded from 3~4 cells. B,
Comparative effects of 1.0 and 10 pM MeHg on time to complete reduction of IGABA in
HEK-293 cells expressing either (11 (dotted bar) or (16 (solid bar) subunit-containing
GABAA receptors. Each bar represents the mean values obtained from 3~5 cells. The
times to total reduction of [GABA by MeHg did not differ signiﬁcantly between HEK-293
cells expressing either (16 or on subunit-containing GABAA receptors for either [MeHg]
tested. C, Comparative effects of 0.1 - 10 pM MeHg on time to 40% reduction of IGABA
in HEK-293 cells expressing either (11 (dotted bar) or (16 (solid bar) subunit-containing
GABAA receptors. Each bar represents the mean values obtained from 3~5 cells. The
times to 40% reduction of [GABA by MeHg did not differ signiﬁcantly between HEK-293
cells expressing either (16 or (11 subunit-containing GABAA receptors for all [MeHg]

tested.

126

27

18v
Time (min)

 

 

 

29280 as as.

36

   

um

2:5 863.com .98 9 8E

Figure 3.12

127

 

C

 

 

 

0 O O 0

2:5 cozonoom .59. 2 9::

Figure 3.12

 

Table 3.1. Comparative effects of MeHg on IGABA suppression in In- or (u;-

 

containing HEK-293 cells. Table summarizing the comparative effects of MeHg on
IGABA suppression for a] and a6 subunit-containing HEK-293 cells. The time-to-
suppression by each concentration of MeHg did not differ signiﬁcantly between (11 and a6
receptor subtypes. Note that for 0.1 pM MeHg, complete reduction of IGABA was not
achieved in HEK-293 cells expressing either subtype of receptor (as indicated by the
minus symbol), because cells ﬂattened out after ~ 35 min and recordings could not be

maintained.

129

 

Time to Total
Reduction (min)

Time to 40 °/o
Reduction (min)

 

 

 

 

 

 

 

 

 

 

 

0.1 pM - - 27.7 t 5.5 24.5 i .90

1,0 ”M 24 i .70 29.3 i 3.4 11.3 :I: .85 10.75 :I: .28

10 “M 19.3 i .25 22.75 i 1.4 3.5 i .70 6.5 1- .37
Table 3.1

130

 

Figure 3.13. Effects of MeHg on [GABA are voltage-independent in HEK-293 cells.

A, Whole-cell IGABA were recorded from HEK-293 cells expressing a6 subunit-containing

 

GABAA receptors before (control) and after 5 min. exposure to 1 pM MeHg at holding
potentials ranging from ~80 to +60 mV. Each datum point represents the mean value
recorded from 3~4 cells. The effects of MeHg on [GABA in these cells were voltage-
independent. B, Whole-cell IGABA were recorded from HEK-293 cells expressing (11
subunit-containing GABAA receptors before (control) and after 5 min. exposure to 1 pM
MeHg at holding potentials ranging from ~80 to +60 mV. Each datum point represents
the mean (1 SEM) value recorded from 4 cells. MeHg caused a reduction in IGABA

amplitude in a voltage-independent manner.

131

>

N
O

 

0

Control

 

ti:
0
r

 

Normalized Current (pA)
.L

~80 ~60 ~40 ~20 0

Voltage (mV)

W

N
O
I

 

O

is
o

 

Normalized Current (pA)
6: ti)

~80 ~60 ~40 ~20 0

Voltage (mV)

Figure 3.13

132

20 40

Control

 

60

 

 

20 40

DISCUSSION

The primary aim of these studies was to determine if the differential expression
of a6 subunit-containing GABAA receptors in cerebellar granule and Purkinje neurons
underlies the differential sensitivities to MeHg observed in these cells. To test this, the
whole-cell patch-clamp technique was used to record GABAA receptor-mediated currents
in the presence or absence of MeHg in cerebellar granule (cg-containing) and cerebral
cortical (oi-containing) cells in culture and in HEK-293 cells transfected with cDNA for
either a1~ or a6containing GABAA receptors. MeHg (0.1 — 10 pM) suppressed IGABA in a
time- and concentration-dependent manner in each cell type investigated. Suppression of
IGABA occurred more rapidly in granule neurons as compared to cortical neurons.
However, time—to-suppression of [GABA by MeHg did not differ in granule neurons
expressing either a6~containing or Ill-containing GABAA receptors. Furthermore, effects
of MeHg on IGABA were examined in HEK-293 cells transfected with cDNA for the a6- or
Oil-containing GABAA receptor. The effects of MeHg on IGABA did not differ in HEK-
293 cells expressing either GABAA receptor subtype. These results suggest that the
presence of the (16 subunit alone may not be the underlying reason for the differential
effects of MeHg on GABAergic currents observed in cerebellar granule and Purkinje
neurons; other factors be contribute as well.

Previous studies conducted in slice preparations and primary cultures of rat
neurons have revealed that MeHg impairs inhibitory GABAergic neurons in
hippocampus and cerebellum (Yuan and Atchison, 1995, 1997, 2003; Xu and Atchison,
1997). GABAergic synaptic transmission was shown to be more sensitive to the effects

of MeHg as compared with gluamatergic synaptic transmission (Yuan and Atchison,

133

2003). Importantly, some cells types appear to be particularly sensitive to the effects of
MeHg. Suppression of inhibitory post synaptic currents (IPSCs) in cerebellar granule
cells, for instance, occurs much earlier than does suppression in neighboring Purkinje
cells (Yuan and Atchison, 2003). Thus, GABAA receptors in the cerebellum appear to be
particularly sensitive to the inhibitory effects of MeHg. The exact mechanism(s) by
which MeHg differentially interferes with GABAergic function in different cell types are
not known. One possibility for the enhanced sensitivity of GABAergic responses to
MeHg in granule cells, as compared with Purkinje cells, is their differential expression of
the orig-subunit containing GABAA receptor. GABAA receptors containing the (16 subunit,
as compared with non-d6 containing ones, are more sensitive to inhibition by other heavy
metals such Zn+2 (Draguhn et al., 1990; Saxena and MacDonald, 1994, Zempel and
Steinbach, 1995) and La’“3 (Saxena et al., 1997, Makela et al., 1999) and MeHg has been
shown to increase concentrations of intracellular Zn’2 (Denny and Atchison, 1994). To
test this possibility, IGABA were recorded in response to MeHg and compared in granule
cells and in cortical cells (used as a replacement for Purkinje cells). IGABA were more
sensitive to the effects of MeHg in granule cells than in cortical cells, as IGABA were
inhibited more rapidly in granule cells than in cortical cells. These ﬁndings are
qualitatively consistent with effects of MeHg on IGABA reported previously in cerebellar
slice, but occurred at much lower concentrations of MeHg (Yuan and Atchison, 1999,
2003). The results of these experiments support the possibility that the (L6 subunit may
contribute to the differential effects of MeHg on IGABA observed in granule and Purkinje

cells.

134

To conﬁrm the results obtained in granule and cortical cells, the effects of MeHg
on IGABA were investigated in granule cells only at different postnatal days as these cells
can developmentally express either (16- and/or (ll—containing GABAA receptors. The
effects of MeHg on IGABA did not differ in granule neurons expressing either Go-
containing or (ii-containing GABAA receptors. [GABA were suppressed by MeHg in a
time- and concentration-dependent manner in granule cells containing either subtype of
GABAA receptor, suggesting that the (16 subunit alone does not underlie the differential
sensitivity of [GABA to MeHg in granule and cortical cells. Other factors, such as
differential GABAA receptor auxiliary subunit expression in these two cell types, may be
involved. In addition to the on and (:6 subunits, cerebellar granule cells express GABAA
receptors containing either [32 or [13 subunits in combination with 72, 73, or 6 subunits
(Laurie et al., 1992). Cortical cells (and Purkinje cells), on the other hand, express mainly
the (1613272 subtype of GABAA receptor (Wisden et al., 1996). Variability in GABAA
receptor subunit composition can alter the responsiveness of the receptor to drugs. For
instance, the 06-6 subtype of receptor has an increased afﬁnity for GABA and is
extremely sensitive to inhibition by Zn+2 (Saxena and MacDonald, 1994). Thus, Zn+2 may
indirectly contribute to inhibition of the (1.5-6 subunit-containing receptor. In addition,
sensitivity of the receptor to agents such as barbiturates has been shown to be dependent
of the identity of the [3 subunit, as the [33 subunit makes the GABAA receptor insensitive
to the effects of pentobarbital (Cestari et al., 2000). Furosemide inhibition is enhanced
when the [31 subunit of the GABAA receptor is replaced with either the B2 or [33 subunit

subtypes (Thompson et al., 1999). Similarly, replacement of the [3, subunit of the

135

receptor with the B2 or [33 subunit, results in enhanced sensitivity to potentiation by
ethanol (Mihic et al., 1997) or etomidate, a general anesthetic (Belelli et al., 1997).

To examine the effects of MeHg on (16 or (xi-containing receptor-mediated
currents in isolation, whole-cell recordings were also conducted in HEK-293 cells
containing either (16 or (11 subtype of GABAA receptor. Results showed that effects of
MeHg on [GABA suppression did not differ between HEK-293 cells containing either
subtype of GABAA receptor. Time to suppression of long». by MeHg (1 and 10 pM)
occurred at similar times in Go or (xi-containing HEK-293 cells. These ﬁndings suggest
that differential expression of a6 subunit-containing GABAA receptors in granule and
Purkinje cells does not underlie the differential sensitivities of IGABA to MeHg observed
in granule and Purkinje cells. Furthermore, as both subtypes of GABAA receptor in
HEK-293 cells contained the same [32 and 72 subunits, it suggests that sensitivity to MeHg

by these subunit subtypes does not differ. An alternate explanation for the lack of

difference in sensitivity to MeHg observed between 016- and ail-containing HEK-293 cells,

is the speciﬁc sequence of chNA used. Plasmids (pRK5) containing GABAA-a6
cDN A (Luddens et al., 1990) were used for HEK-cell transfections in these studies.
When cDNA was released from its vector using digestion, several bands were detected,
including a 1.3 kb band, which was determined to be the correct (116-containing cDNA
insert (Hajela, personal communication). Expression of the (116-containing GABAA
receptor was veriﬁed immunocytochemically and pharmacologically, by evaluating the
responsiveness to diazepam. However, after completion of the experiments described in
this paper, it was observed that the same GABAA-a6 cDNA injected into Xenopus

oocytes, produced GABAergic responses that did not appropriately respond to the Cl‘

136

channel blocker, niﬂumic acid. Instead of being inhibited by niﬂumic acid, GABAergic
responses in oocytes expressing the 016- subunit were potentiated. Therefore, the (16-
containing cDN A insert sequence used in these studies may not have been spliced at
precisely the right site, or the receptor isoform simply responded differently when
expressed in oocytes as compared with in HEK-293 cells.

In sum, acute bath application of MeHg to rat cerebellar granule and cerebral
cortical cells completely and irreversibly suppressed 10AM. IGABA in granule cells were
more sensitive to the effects of MeHg than were IGABA in cortical cells. However, IGABA
in granule cells expressing 0.5-containing GABAA receptors were approximately equally
as sensitive to the effects of MeHg as [GABA in (xi-containing cells. Moreover,
suppression of [GABA by MeHg was similar in HEK-293 cells expressing either 0.6- or al-
containing receptors. These results suggest that the expression of the (16 subunit alone
does not underlie the differential effects of MeHg on IGABA observed in cerebellar granule
and Purkinje neurons; additional factors may be involved as well. Further chimeric
studies are needed to determine the relevant contribution, if any, of each GABA receptor
subunit and its subtypes, to the differential sensitivity of GABAAergic responses to

MeHg in cerebellar cells.

137

 

 

 

CHAPTER FOUR

SUMMARY and DISCUSSION

138

 

 

 

A. Summary of Research

The studies described in this thesis were aimed at characterizing the interactions
of MeHg with the GABAA receptor and at determining if the differential expression of (16-
containing GABAA receptors in cerebellar granule and Purkinje neurons underlies the
differential sensitivities to MeHg observed in these cells. This is due to the observations
that MeHg inhibits GABAergic synaptic transmission, and does so more rapidly in
cerebellar granule than in Purkinje cells (Yuan and Atchison, 1999, 2003). A possible
explanation for this differential effect of MeHg on [GABA may be the differential
expression of a6 subunit-containing GABAA receptor in these cell types. To address this
hypothesis, 1 performed whole-cell voltage-patch clamp experiments in cerebellar granule
and cerebral cortical cells in primary culture and in HEK-293 cells expressing speciﬁc
GABAA receptor subunits of interest. I showed that GABAA receptor-mediated currents
in cerebral cortical cells in culture are completely and irreversibly suppressed by MeHg
(0.1, 1.0, and 10 pM) in a time- and concentration-dependent manner. Because an effect
less than complete suppression of current was not seen, the variable measured was time
to suppression. This is a general response of cells to MeHg (Arakawa et al., 1991; Yuan
and Atchison, 1993, 1995, 1997, 1999, 2003; Xu et al., 1997). The magnitude of effects
produced by different concentrations of MeHg did not differ. At each concentration of
MeHg tested (0.1, 1.0, and 10 pM) the time-course of suppression of [GABA was
approximately linear and complete suppression was always attained (if recordings could
be maintained long enough). This is consistent with the effects of MeHg reported on
GABAergic synaptic transmission in granule and Purkinje cells in the cerebellar slice

preparation (Yukun and Atchison, 2003). In cerebellar slice, the magnitude of changes in

139

frequency or amplitude of postsynaptic currents did not change, irrespective of the
concentration of MeHg used. The ﬁxed magnitude of effects caused by different
concentrations of MeHg suggests that a consistent series of actions takes place once an
effective or threshold concentration of MeHg is attained.

I also showed that MeHg affects the kinetics of IGABA by prolonging the slow
decay rate. These effects were concentration-independent, in that at each concentration
of MeHg tested, 15.0,. was enhanced to approximately the same extent. GABAA receptor
current decay is important as it is a major determinant of IPSC duration (Twyman et al,
1990; MacDonald and Twyman, 1992; Jones and Westbrook, 1995). The faster
component of decay is associated with receptor desensitization, whereas the slower
component of decay is typically associated with GABA receptor deactivation
(MacDonald et al., 1989a; Twyman et al., 1990; Twyman and MacDonald, 1992).
Deactivation rate is inﬂuenced by unbinding of GABA from its receptor as well as by
transitions of activated receptors between open, closed, and desensitized states. Even
though removal of agonist is considered the ﬁnal step in current deactivation, it can be
prolonged by detaining the receptor in the desensitized state (Jones and Westbrook, 1995;
Tia et al., 1996; Haas and MacDonald, 1999). Unbinding is prolonged from the
desensitized state because the receptor needs ﬁrst to recover from the desensitized state to
reopen prior to unbinding of GABA. Thus, GABAA receptor deactivation can be shaped
by several different mechanisms. In the present study, the prolongation of Tslow by MeHg
likely indicates that MeHg slows the rate of unbinding of GABA from its receptor.
Alternatively, it could indicate that MeHg produces an accumulation of desensitized

GABAA receptors, resulting in prolonged unbinding of GABA from its receptor. To

140

 

 

investigate the effects of MeHg receptor desensitization, further studies are needed, such
as ones investigating the effects of MeHg on the recovery of the GABAA receptor from
desensitization using paired GABA pulses applied in the absence and in the presence of
MeHg. Additional studies at the single channel level would also be useful in examining
the effects of MeHg on GABAA receptor kinetics in more detail. These experiments
should be performed with the use of a fast perfusion system, as a relatively slow
perfusion system, such as the one used to perform experiments in this thesis, may not
appropriately resolve the fast phases of decay, or desensitization. For the experiments
described in this dissertation, physiological saline solution was delivered via gravity into
the extracellular solution at a rate of approximately 0.45 mllmin and the 10~90% rise time
for solution exchange was ~ 15 8.

Furthermore, I have shown that the effects of MeHg are voltage-independent,
suggesting that MeHg does not interact with the pore of the GABAA receptor. 1 have
shown that the effects of MeHg on IGABA are GABA concenUation-independent,
suggesting that MeHg does not compete with GABA for its binding site on the receptor.
Moreover, MeHg prolonged IGABA decay and completely suppressed IGABA in the
presence of diazepam (10 pM) or the benzodiazepine antagonist, ﬂumazenil (20 pM),
suggesting that MeHg does not compete with diazepam for its binding site on the
GABAA receptor. Thus, MeHg may interact with a unique site on the receptor or one
belonging to another known drug. It is also possible that MeHg may have non-speciﬁc
effects on the GABAA receptor.

To address the second goal of these studies, I examined and compared the

effects of MeHg on IGABA in cerebellar granule (rm-containing) and cortical cells (al-

141

 

 

containing) in culture. Suppression of IGABA occurred more rapidly in granule neurons as
compared to cortical neurons, suggesting that (1.5-containing granule cells are more
sensitive to the effects of MeHg. To verify these ﬁndings in granule cells, which can
express either subtype of GABAA receptor, I initially attempted to distinguish the two
GABAA receptor subtypes by pharmacological means. By using the benzodiazepine,
diazepam, I predicted that I would observe differences in IGABA responsiveness between
cells expressing 016- and non- (1.5-containing receptors. However, in almost all cells tested,
whole-cell IGABA were similarly potentiated by diazepam. Thus, it was determined that it
would be extremely difﬁcult to differentiate the receptor subtypes pharmacologically in
granule cells, as they contain GABAA receptors that have a wide variety of subunit
combinations. Importantly, these subunit arrangements can consist of a1, (16, or a mixture
of a; and (16, in addition to other subunit variations, producing receptors with a multitude
of different pharmacological proﬁles. Consequently, immunocytochemistry was used to
identify a1- and (1.5-containing granule cells. As GABAA receptor subtypes are expressed
developmentally, granule cells were grown for different days in culture and their
responsiveness to MeHg was compared at DIV 4 (a1 expressing) and at DIV 6—8 (a; and
a6 expressing). Subunit subtype expression was determined using antibodies against each
a subunit of interest. Time-to-suppression of IGABA by MeHg did not differ in granule
neurons expressing either dig-containing or III-containing GABAA receptors. To conﬁrm
the results obtained in native cerebellar granule neurons, effects of MeHg on IGABA were
examined in HEK-293 cells transfected with cDNA for the a6~ or (Ii-containing GABAA
receptor. The effects of MeHg on IGABA did not differ in HEK-293 cells expressing either

GABAA receptor subtype. The results obtained in cerebellar granule and cortical neurons

142

and those reported in HEK-293 cells suggest that the presence of the (16 subunit alone
may not be the underlying reason for the differential effects of MeHg on GABAergic

currents observed in cerebellar granule and Purkinje neurons.

. Relationship to Previous Work

In hippocampal slice, GABAergic inhibitory synaptic transmission is more
sensitive to suppression by MeHg than is excitatory transmission, because suppression of
inhibitory synaptic transmission by MeHg occurrs before suppression of glutamatergic
synaptic transmission (Yuan and Atchison, 1995, 1997). These studies also showed that
MeHg ﬁrst increased excitatory postsynaptic potential (EPSP) amplitudes prior to their
suppression. This early enhancement of EPSP amplitude prior to inhibition was shown to
occur as a result of MeHg suppressing GABAA receptor-mediated inhibitory synaptic
transmission prior to suppressing excitatory transmission (Yuan and Atchison, 1997). In
the cerebellum, GABAergic synaptic transmission is also more sensitive to the effects of
MeHg as compared with glutamatergic synaptic transmission (Yuan and Atchison, 2003).
Importantly, these effects of MeHg on GABAergic synaptic transmission in cerebellum
were cell type preferential. IPSCs were more sensitive to suppression by MeHg in
granule cells than in Purkinje cells (Yuan and Atchison, 2003). Similarly in this thesis, in
primary neurons in culture, MeHg suppressed IGABA more rapidly in cerebellar granule
cells (org-containing) than in cerebral cortical cells (cl-containing). In both cell types,
MeHg suppressed IGABA completely and irreversibly in a time- and concentration-
dependent manner. Granule cells were more sensitive to the effects of MeHg than were

cortical cells, as [GABA were inhibited more rapidly in granule cells than in cortical cells.

143

 

These ﬁndings are qualitatively consistent with effects of MeHg on IGABA reported
previously in cerebellar slice, but occurred at much lower concentrations of MeHg (Yuan
and Atchison, 1999, 2003). Differences in concentrations of MeHg needed to produce
inhibition in cerebellar slice as compared with cells in culture may be due to the
difference in thickness of the preparations (Meacham et al., 2005). The slice preparation
is much thicker, resulting in a limited surface area available for accumulation of MeHg.
To reach its target cells, MeHg must diffuse through the slice, encountering a greater
number of sites of interaction for MeHg. However, for cells grown in culture, most cells
have a majority of their surface area directly exposed to MeHg in the bathing solution,
thus a much lower concentration is needed to produce effects. These ﬁndings suggest
that the differential expression of the (16 subunit may contribute to the differential
sensitivities of IGABA to MeHg in granule and cortical or Purkinje cells.

However, additional data in this thesis do not support this hypothesis. IGABA in
cerebellar granule cells expressing (1.5-subunit containing receptors were equally as
sensitive to inhibition by MeHg (0.1, 1.0, and 10 pM) as were IGABA in granule cells not
expressing dis-containing receptors. It is important to keep in mind that cells expressing
(1.5-containing receptors also express Ill-containing ones, and a1 and (:6 subunits are often
co-expressed in the same receptor. Thus, complete isolation of (16 responses was most
likely not achieved. In the HEK-293 cell expression system, responses mediated solely
by (11 or (1.5-containing GABAA receptors can be recorded. Findings from this thesis
revealed that IGABA in HEK-293 cells were abolished completely and over a similar time-
course by MeHg (0.1, 1.0, and 10 pM) in both 016- or ail-containing cells. Effects of

MeHg on IGABA were voltage-independent in cells containing either receptor phenotype.

144

 

 

These results suggest that the presence simply of the dis-containing GABAA receptor may
not underlie the differential effects of MeHg on IGABA observed in cerebellar granule and
Purkinje neurons, but that other factors are likely involved. One contributing factor may

be the differential expression of auxiliary GABAA receptors subunits in these cells.

 

Other factors may include the alteration of intracellular concentrations of Ca+2 by MeHg.
This is important because within the cerebellum, granule cells express the highest density
of cholinergic muscarinic (M3) receptors as compared with other cell types (Neustadt et
al., 1988). These receptors are involved in synaptic transmission between granule and
Purkinje cells, in that acetylcholine increases the glutamate-mediated excitation of
Purkinje cells (Takayasu et al., 2003). M3 receptors play an important role in increasing
[Can] from intracellular stores, as they activate 1P3 (inositol triphosphate) receptors
resulting in release of Ca+2 into the cytoplasm from the endoplasmic reticulum (Kandel et
al., 2000), Notably, M3 receptors have been shown to be upregulated by MeHg (Coccini
et al., 2000), and in cerebellar granule cells in culture, a decrease in M3 receptors protects
against neurotoxicity by MeHg (Limke et al., 2004). Another contributing factor may
include the alterations of intracellular concentrations of Zn+2 by MeHg. This is important
because Zn+2 can be released from glutamate-containing vesicles (Assaf and Chung,
1984; Howell et al., 1984). Granule cell axon terminals that from synapses with Purkinje
cells have dis-containing GABAA receptors, which are extremely sensitive to inhibition by
Zni2 (Gallo et al., 1981). Thus, Zn+2 my be co-released with glutamate from granule cell
axons onto Purkinje cells, and may feed back onto dis-containing GABAA receptors in

granule cells and inhibit them.

145

 

A small number of studies have found that administration of MeHg, both short
term and long term, enhances benzodiazepine binding in rodent brain via interaction with
the GABAA receptor (Corda et al., 1981; Concas et al., 1983; Fonfria et al., 2001).
However, Komulainen and Tuomisto (1985) showed that both acute as well as chronic
exposure to MeHg does not affect the number of benzodiazepine receptors labeled with
[3H]ﬂunitrazepam in rat cerebellum. To determine if MeHg interacts with the
benzodiazepine site of the GABAA receptor, studies in this thesis were performed to
examine the effects of MeHg on IGABA in the presence or absence of diazepam. In the
presence of diazepam, the effects of MeHg on [GABA were maintained, even following
pretreatment with diazapem. The actions of MeHg on IGABA were also unaffected by
treatment with the benzodiazepine antagonist ﬂumazenil. These results suggest that
MeHg interacts at a site on the GABAA receptor different from that used by diazepam.
They are consistent with those reported by Komulainen and Tuomisto (1985). However,
these results do not necessarily contradict those described by others (Corda et al., 1981;
Concas et al., 1983; Fonfria et al., 2001). This is because increased binding of diazepam
to the GABAA receptor after MeHg exposure, does not necessarily indicate that MeHg
alters the function of the receptor in response to diazepam. Instead, increased binding of
diazepam in response to MeHg exposure could occur as result of the cell trying to
compensate for the antagonistic affects of MeHg by increasing GABAA receptor
benzodiazepine binding sites. In addition, exposure to MeHg could alter GABAA
receptor subunit expression, perhaps increasing expression of benzodiazepine sensitive

subunits, such as on.

146

 

C. Possible Mechanisms Underlying the Differential Effects of MeHg on Cerebellar

Granule and Purkinje Neurons

The circuitry of the cerebellum may play a role in the differential effects of
MeHg-induced neurotoxicity observed in granule and Purkinje cells. Excitatory mossy
ﬁbers release glutamate on granule cell dendrites, which results in the propagation of
action potentials down granule cell axons, or parallel ﬁbers. These axons form excitatory
synapses with Purkinje cells and Golgi cells. Upon Golgi cell activation by granule cell
parallel ﬁbers, Golgi cells release GABA onto the synapse between granule cells and
mossy ﬁbers, forming an inhibitory feed-back loop. MeHg is known to increase Ca"2
concentrations within cells (Denny et al., 1993; Hare et a1, 1993; Denny and Atchison,
1996; Marty and Atchison, 1997; Limke and Atchison, 2002) which may result in
enhanced spontaneous GABA release from Golgi, stellate, and basket cells, and increased
glutamate release from granule cell parallel ﬁbers. Because the membrane excitability of
neurons is highly dependent on the combined activity from both excitatory and inhibitory
inputs, disruption of GABAergic synaptic transmission in cells can be damaging. Loss of
inhibition can cause an imbalance in excitability to occur, resulting in cell
overexcitability. Thus, because GABAergic transmission is inhibited preferentially in
granule cells by MeHg (Yuan and Atchison, 2003), it may be possible that disruption of
the granule cell-Golgi cell inhibitory feedback loop could occur. As a result, granule cells
would have diminished abilities to respond to the inhibitory input from Golgi cells, and in
turn become overexcited by glutamatergic input from mossy ﬁbers. In addition, MeHg
could suppress dig-containing GABAA receptor-mediated responses, resulting of

impairment of tonic inhibition mediated by this receptor subtype. GABAA receptors

147

 

 

containing (16 subunits can also contain 6 subunits and are localized at the extrasynaptic
membrane of granule cells, typically at synapses with Golgi cell axons (Nusser et al.,
1996; Nusser et al., 1998). This is important because one of the main contributors in the
regulation of granule cell excitability is the tonic inhibitory Cl' conductance. The Golgi
cell granule cells synapse is specialized to promote transmitter spillover, as this synapse
is within a glomerulus (Hamori and Zsentagothai, 1966; Jakab and Hamori, 1988).
Glomeruli are specialized synapses that become ensheathed by glial cells, forcing
neighboring dendrites to share a limited space, thus facilitating the build-up of
neurotransmitter (Fig. 4.1). Each glomerulus contains glutamatergic mossy ﬁber
terminals surrounded by numerous granule cell dendrites intermingled with Golgi cell
axon terminals that form GABAergic synapses with granule cells. GABA released at
Golgi to granule cell synapses is conﬁned by the glial sheath and is removed by uptake
into Golgi terminals and by surrounding glia (Itouji et al., 1996). As a result, any GABA
that escapes from the synaptic cleft is likely to spill over (Rossi et al., 1998) onto both
postsynaptic and extrasynaptic sites on granule cell dendrites, as both types of synaptic
sites are expressed within the glomerulus (Nusser et al., 1995; 1996). Increased Ca+2
concentrations caused by MeHg may initially contribute to enhanced GABA release from
Golgi cells and increased GABA spillover. Extrasynaptically located (16 (and 6)~

containing receptors are likely to be activated by low spillover concentrations of GABA,

148

Figure 4.1. Drawing showing a cerebellar glomerulus. This specialized synapse is
ensheathed by glial cells and contains glutamatergic mossy ﬁber terminals surrounded by
numerous granule cell dendrites. It also contains Golgi cell axon terminals that form

GABAergic synapses with granule cells. Figure 4.1, borrowed from Kandel et al., 2000.

149

Glomerulus

   
   
 

Mossy fiber
terminal

Golgi cell axon Granule cell
dendrite

Mossy fiber

Figure 4.1.

150

 

as these receptors have a high afﬁnity for GABA and do not readily desensitize (Saxena
et al., 1994; Tia et al., 1996b). Spillover GABA acting at (1.5-containing receptors results
in persistent opening of GABAA receptor channels producing a tonic form of inhibition
(Nusser et al., 1998). Phasic inhibition produced by direct activation of granule cells by
Golgi cells is thought to produce more synchronous rhythmic ﬁring, and tonic inhibition
is thought to reduce synchronicity and rhythmicin of ﬁring patterns of granule and Golgi
cells (De Schutter and Maex 1996; De Schutter, 2002; Maex and De Schutter, 2003).
MeHg antagonizes both (16 and (xi-containing GABAA receptors, which could produce a
reduction in tonic and phasic inhibition, resulting in heightened levels of excitability of
granule cells (Fig. 4.2.). Increased intracellular Ca+2 concentrations caused by MeHg
may further contribute to increased glutamate release from granule cell parallel ﬁbers.
Enhanced glutamate could also occur in parallel with increased Zn+2release, as both are
contained and released from glutamatergic synaptic vesicles (Assaf and Chung, 1984;
Howell et al., 1984; Xie and Smart, 1991). As a result, Zn+2 can inhibit a66~containing
GABAA receptors in granule cells, which are extremely sensitive to its effects,
contributing to the suppression of tonic inhibition. However, (rm-containing receptors in
Purkinje cells, which have a reduced sensitivity to Zn”, are less likely to be affected. The
excessive glutamatergic output from granule cell parallel ﬁbers onto Purkinje neurons
could be detrimental in regulation of cerebellar output, as the mossy ﬁber-granule cell-
Purkinje cell pathway through the cerebellum has been shown to be the operational path,
as it correlates with and controls moment-to-moment behavior (Kandel et al., 2000).
Under normal conditions, Purkinje cells are excited by glutamatergic input from granule

cell axons and climbing ﬁber terminals, and are inhibited by GABAergic input from

151

Figure 4.2. Diagram depicting the proposed mechanism of interaction of MeHg with
GABAA receptors of granule cells to alter cerebellar excitation. A. Climbing ﬁbers (CF)
release glutamate and excite Purkinje cells (PC). Mossy ﬁbers (MF) release glutamate
and excite both granule cells (GC) and Golgi cells (GOL). GCs release glutamate and
and excite GOL and PC. Zn+2 may also be co-released with glutamate, and it may
feedback to inhibit extrasynaptic GABAA receptors. GOL release GABA and form an
inhibitory feedback loop with GC. GCs express both do and on containing GABAA
receptors, whereas PC express only (xi-containing ones. PC express only AMPA-type
glutamate receptors, whereas GCs and GOL express both AMPA and NMDA subtypes of

+211 and

receptor. M3 receptors are abundant in GCs and play a role in increasing [Ca
enhancement of glutamate release. Nearby astroctyes (A) contain glutamate transporters
that remove glutamate from the synaptic space. PC provide the only output from the
cerebellar cortex; it is inhibitory. For simplicity, basket and stellate intemeurons have
been excluded. B. MeHg increases [Can], and release of glutamate and GABA, and
inhibits both (16 and a1-containing GABAA receptors as well as astrocyte glutamate
transporters. As a result, GCs are unable to respond to GOL inhibition and become

overexcited. GCs, in turn, overexcite PC, thus enhancing the inhibitory output of the

cerebellar cortex.

152

engage .
<m<o .

56 5

ms— a

<n_s_< .
$253.22 a
:0 <<m<o -

co <<m<o .

 

 

 

 

 

 

153

 

@‘AIVu—U

 

 

 

 

sneeze .
<m<o .
So a

n2 Q

<nz<2n22 .
.o .55 .
on <<m<e .

N... 8&3

«

 

154

stellate and basket cells. Purkinje cells provide the only output signals of the cerebellum
by forming inhibitory GABAergic synapses with deep nuclei of the cerebellum. These
nuclei, in turn, excite brain pathways receiving input from the cerebellum. Excessive
excitatory drive from granule cells could lead to overexcitability of Purkinje cells (Fig
4.2). In addition, increased intracellular concentrations of Ca+2 caused by MeHg could
further lead to enhanced excitability of these cells. A possible contributing factor leading
to enhanced excitation of granule cells by MeHg may be its effects on astrocytes situated
nearby granule and Purkinje cells. Under normal conditions, these astrocytes contribute
to the removal of some transmitters during synaptic transmission (Hosli et al., 1995),
including removal of glutamate via a glutamate carrier (Kimelberg et al., 1989). MeHg,
however, has been shown to inhibit the transport of glutamate into astrocytes (Aschner et
al., 2000), thus possibly making Purkinje cells even more vulnerable to increased
concentrations of glutamate released from granule cells (Fig 4.2). As a result, the
inhibitory output of Purkinje cells may be disturbed and regions of the brain receiving
cerebellar signals, such as the motor cortex, via the thalamus, and the spinal cord, via
brain stem nuclei, could be disrupted. This disruption could produce abnormalities in
motor function, such as tremors and ataxic gait, similar to what has been observed in
victims of MeHg poisoning.

Data from this thesis in granule cells suggest that differential expression of a6
subunit-containing GABAA receptors is not the underlying factor involved in the
differential effects of MeHg in cerebellar cells. However, these ﬁndings are complicated
by the fact that GABAA receptors in granule cells can express a variety of different

GABAA receptor phenotypes. Thus, expression of auxiliary GABAA receptor subunits

155

 

 

 

may contribute to the differential effects of MeHg observed. This is important as the
GABAA receptor subunit composition can inﬂuence its pharmacology (Luddens and
Wisden, 1991; Saxena and MacDonald, 1994, 1996; Korpi et al., 1995; Fisher et al.,
1997, 1998; Makela etal., 1997, 1999; Saxena et al., 1997; Cestari et al., 2000). GABAA
receptor subtypes in cerebellar granule cells are numerous and diverse (ale72, (166,72,
aia6Bx72, a6Bx6 and alaéBx6) (Pollard et al., 1993; Khan et al., 1994; Nusser et al., 1996,
1998). However, Purkinje cells express GABAA receptors containing mainly (11, B2, and
72 subunits (Wisden et al., 1996). At the granule cell-Purkinje cell synapse, Purkinje
cells express a1B2,72, whereas granule cells express a1B2/3,72, 06B2/372, (11061321372 receptors
synaptically and a6B2/36 and a1a6B2/36 receptors extrasynaptically (Schmid et al., 1999;
Nusser et al., 1998). An auxiliary subunit that may contribute to the differential
sensitivity to MeHg between these cell types is the 6 subunit. Receptors that contain the
6 subunit, in combination with the 0.6 subunit, are present only in the extrasynaptic
membrane of granule cells. 6-containing GABAA receptors are highly sensitive to the
inhibitory effects of Zn+2 (Draguhn et al., 1990). In addition, mice with targeted
inactivation of the 6 subunit have been shown to have reduced sensitivity to the
behavioral actions of neurosteroids (Mihalek et al., 1999). Similarly, recombinant
GABAA receptors containing the 6 subunit show an enhanced sensitivity to neurosteroids
(Adkins et al., 2001; Wohlfarth et al., 2002).

Another auxiliary GABAA receptor subunit of interest is the B subunit. Purkinje
cells (and cortical cells) express GABAA receptors that mainly have B2~containing
receptors; B3~containing receptors are expressed much less frequently (Wisden et al.,

1996). Granule cells have GABAA receptors that express either the B2 or B3 subunit. The

156

identity of the B subunit can alter the pharmacology of the receptor. Sensitivity of the
receptor to agents such as barbiturates has been shown to be dependent of the identity of
the B subunit, as the B3 subunit makes the GABAA receptor insensitive to the effects of
pentobarbital (Cestari et al., 2000). Furosemide inhibition is enhanced when the B1
subunit of the GABAA receptor is replaced with either the B2 or B3 subunit subtypes
(Thompson et al., 1999). Similarly, replacement of the B1 subunit of the receptor with the
B2 or B3 subunit, results in enhanced sensitivity to potentiation by ethanol (Mihic et al.,
1997) or etomidate, a general anesthetic (Belelli et al., 1997).

In addition, data from this thesis show that the effects of MeHg on IGABA
suppression did not differ between HEK-293 cells containing either (16 or a] subtype of
GABAA receptor, suggesting that differential expression of a6 subunit-containing
GABAA receptors in granule and Purkinje cells does not underlie their differential
sensitivities of [GABA to MeHg. One possibility for the lack of difference of effects of
MeHg observed in HEK-293 cells, may be the lack of co-expression of the (16 and 6
subunits in these cells. To determine if the 6 subunit plays a role, HEK-293 cells could
be transfected with cDNA with GABAA receptors containing either the (16 or (16 6
subunits. Responsiveness of IGABA in these cells to MeHg could then be compared.
However, the lack of differential responses of IGABA to MeHg in HEK-293 cells
expressing either (16 or a; subunits in this thesis could also suggest that other mechanisms
are involved. A plausible explanation for the differential effects of MeHg on IGABA
observed in cerebellar granule and Purkinje cells is the differential Ca+2 buffering
abilities between these cells. Purkinje cells express the binding protein, calbindin,

whereas granule cells do not (Celio, 1990). Thus, Purkinje cells are more able to resist the

157

increases of [Ca+2]i, caused by MeHg (Edwards et al., 2005). As a result, proteins
activated by Ca”, such as protein kinase C (PKC), are less likely to be set into motion.
This is important as PKC is known to phosphorylate the GABAA receptor and decrease
its function (Moss and Smart, 1996; Brandon et al., 2000). Furthermore, intracellular
Ca+2 ions can also suppress the GABA-activated Cl' conductance, by decreasing the
apparent afﬁnity of the GABAA receptor (Inoue et al., 1986). Thus, the effects of MeHg
on IGABA could be Ca"2 dependent. To test this, [GABA could be recorded in cells in '
response to MeHg in the presence and absence of the Ca+2chelator, BAPT A, and Ca”-
free solution. This type experiment has been performed in cerebellar slice (Yuan and
Atchison, 2003), and ﬁndings showed that suppression of GABAergic currents was
delayed, but not prevented, by addition of BAPT A or Can—free solution, suggesting that

the effects of MeHg on IGABA are Ca+2-independent.

. Potential Mechanisms of Interaction Between MeHg and the GABAA Receptor
The mechanisms by which MeHg interacts with the GABAA receptor are not

clear. However, it is known that MeHg has a high afﬁnity for sulﬂrydryl and disulﬁde
groups on cysteines that are present in proteins including the GABAA receptor. GABAA
receptors are members of the Cys-loop family of receptors, which include the nicotinic
acetylcholine receptors, (Karlin and Akabas, 1995, Le Novere etal., 2002). All subunits
in this receptor family have a disulﬁde-linked cysteine loop in the extracellular domain
consisting of 15 residues. Thus, it is likely that MeHg might bind to these disulﬁde
groups on the external side of the GABAA receptor to produce its effects, as this site has

been proposed to play a role in GABAA receptor subunit expression and assembly (Amin

158

et al., 1994) as well as in the transduction mechanism between agonist binding and
channel opening (Kash et al., 2003). In nicotinic acetylcholine receptors, cysteine residue
mutation studies have shown that the disulﬁde loop is required for antagonist binding and
proper subunit assembly and binding (Blount and Merlie, 1990; Sunrikawa and Gehle,
1992). However, the exact site(s) and subunit(s) with which MeHg may interact with the
receptor have yet to be revealed. Data from this thesis have not determined a site by
which MeHg might interact with the GABAA receptor, however it has eliminated a few
possible candidates. Data from the present thesis showed that MeHg does not likely
interact with the diazepam site of the GABAA receptor. Furthermore, studies in this
dissertation were conducted to determine whether or not MeHg interacts with the GABA
site of the receptor. The GABA binding sites are distributed between residues on both
the a and B subunits of the receptor (Sigel et al., 1992; Anrin and Weiss, 1993). On the
B2 subunit, amino acids Y157-Y160 and T202~Y205 are involved, and on the a; subunit
residues F64, R66, S68, and R120 are involved in GABA binding; on the (:6 subunit,
residue S68 is replaced by T68 (Sigel et al., 1992; Boileau et al., 1999; Westh-Hansen et
al., 1999). To determine if MeHg interacts with the GABA binding site of the receptor,
whole-cell IGABA were recorded in the presence of increasing concentrations of GABA, in
the presence of MeHg. If MeHg competes with GABA for its binding site, then greater
concentrations of GABA should interfere with the actions of MeHg. However, ﬁndings
from this thesis showed that time to suppression of IGABA by MeHg was not dependent
upon GABA (10 pM - 1 mM) concentration in cortical cells in culture, suggesting that
MeHg does not appear to compete with GABA for its site on the receptor (Appendix, Fig.

A. 2). These data are consistent with ﬁnding from other studies (Narahashi et al., 1994;

159

Fonfria et al., 2001). In rat dorsal root ganglia neurons in primary culture, the effects of
MeHg on Iona». were not prevented by bicuculline, a GABAA receptor antagonist known
to bind to the GABA site (Narahashi et al., 1994). Furthermore, in cultured cerebellar
granule neurons, MeHg did not inhibit [3H]GABA binding, suggesting that MeHg does
not directly act on the GABA recognition site (Fonﬁia et al., 2001). To determine if
MeHg interacts with the channel pore of the GABAA receptor, the effects of MeHg on
IGABA were examined at different holding potentials. If MeHg binds to the pore, it will be
subjected to the electrical ﬁeld of the membrane, resulting in voltage-sensitivity.
Stepping the voltage from ~80 mV to +60 mV in the presence of MeHg produced a linear
(voltage-independent) current-voltage relationship, suggesting that MeHg does not bind
to the channel pore of the receptor. These ﬁndings are consistent with the fact that the
GABAA receptor is anion-conducting because it has a channel pore that is lined with
positively charged residues at the extracellular and intracellular ends (Akabas et al.,
1994). Because MeHg is a monovalent cation, it is unlikely to interact with the channel
pore of the GABAA receptor. In agreement with this, are ﬁndings from Narahashi and his
colleagues (1994) that showed that picrotoxin, a GABAA receptor antagonist that binds to
the channel pore, did not prevent the effects of MeHg on IGABAo

In addition, comparison of a6 and a; subunit sequences identiﬁed a single
residue that affected the benzodiazepine sensitivity of the receptor (Wieland et al., 1992).
Mutation of the (16 subunit amino acid Arg~100 in the extracellular domain to His-100,
found in on subunits, transformed the normally benzodiazepine insensitive receptor to a
benzodiazepine sensitive one. As these structural changes make the 016- and (11-

containing receptor differentially sensitive to agents such as benzodiazepines and Zn”,

160

perhaps it may play a role in the differential effects of MeHg on GABAergic responses.
However, this is unlikely, as the majority of the ﬁndings in this thesis have suggested that
the ab subunit of the GABAA receptor is not more sensitive to MeHg than is (11.

The observation in this thesis that MeHg affects the decay kinetics of IGABA may
provide some clues as to the possible binding site for MeHg on the GABAA receptor.
[GABA in cortical cells typically decayed in a bi—exponential manner consisting of a fast
(rm) and slow (mow) component, and Tslow was prolonged by MeHg in a concentration-
independent manner. MeHg did not appear to affect the was. at any time or at any
concentration tested. Because MeHg affects decay kinetics of IGABA, a likely region of
interaction of MeHg is the membrane domain M2, as much evidence exists for the role of
the threonine 266 residue (Im et al., 1995) or the 9’ leucine residue (Dalziel et al., 2000)
of membrane domain M2 in gating of GABAA receptors. Because the highly conserved
9’ leucine residue of domain M2 is hypothesized to form the channel gate (Unwin, 1993),
many mutagenesis studies have been conducted at this region. Mutations at this position
have shown to have an impact on channel gating, including alterations of rates of entry
into desensitized states (Im et al., 1995; Dalziel et al., 2000). However, studies by
Bianchi and colleagues (2001) revealed that in GABAA receptor subunits that have the 9’
leucine residue of domain M2 conserved (72L and 6 subunits), dramatically different
desensitization kinetics are observed, suggesting that other domains are involved in the
regulation of desensitization. Using chimeric strategies, Bianchi et al., (2001) mutated all
non-identical sites in the 72L (a splice variant of 72) and 6 subunit to the corresponding
amino acids in the other subunit to create “M2-swap” subunits (72L(M2S) and 6(M2S)).

Neither the desensitization nor the deactivation kinetics were altered by changes in the

161

M2 domain sequence. Next, these investigators mutated the M1-M2 linking loop of the
two subunits and found that the N-terminus and extracellular end of domain M1 can
modulate fast desensitization independent of the identity of M2. Based on these ﬁndings
in conjunction with data from cysteine-scanning studies (Akabas et al., 1994; Karlin and
Akabas, 1995), Bianchi and colleagues (2001) suggested that the M1 domain may be
interconnected with the M2 domain toward the extracellular end of the channel. It was
hypothesized that the interaction between M1 and M2 could serve to couple
conformational changes in these two domains, resulting in altered channel kinetics.

Thus, it may be possible that MeHg interacts with the cysteine-loop on the N-terminus of
M1, which is linked to M2, to produce effects on decay and desensitization kinetics of
IGABA. However, detailed molecular studies using GABAA receptor exchange mutants are
needed to determine more accurately the regions or residues with which MeHg might

interact to produce its effects.

. Conclusion

In sum, ﬁndings in this thesis have revealed that MeHg inhibits GABAA receptor-
mediated currents, possibly by altering its response kinetics. This is important as
disruption of GABAergic activity can have deleterious effects in the adult as well as in
the immature brain. In the adult CNS, GABAA receptors maintain basal excitability
levels by mediating inhibitory synaptic transmission, regulating glutamatergic activity
and preventing hyperexcitation (DeLorey and Olsen, 1992; Olsen and Avoli, 1997;
Kardos, 1999). If the main excitatory inputs to the cerebellum (climbing and mossy

ﬁbers) are not kept in balance by GABA, motor and sensory regions of the brain

162

receiving input from the cerebellum can become improperly excited. As a result,
alteration of excitation in the brain can result in movement and sensory disturbances,
such as those observed with MeHg intoxication, or seizures, such as those associated with
epilepsy (Jones-Davis and MacDonald, 2003). In the immature CNS, GABAA receptor-
mediated depolarization induces various steps in brain development, such as cell
proliferation, cell migration and neural maturation, including synaptogenesis (Barker et
al., 1998; Kardos, 1999; Varju et al., 2001; Ben-Ari, 2002; Owens and Kriegstein, 2002).
Cell migration is an important process during brain development by which neurons reach
their ﬁnal location. Such physical contact between cells is important for the construction
of complex circuits, and any interference with cell migration has profound detrimental
effects on the developing brain. To achieve mature function as signal transmitters,
neurons must form connections, and this occurs during the process of synaptogenesis.
Though neurons retain the ability to make new synapses throughout life, the
developmental period is critical for the formation of the basic circuitry of the nervous
system (Rodier, 1995). Establishment of GABAergic synapses is crucial for the
expression of normal and higher brain functions, such as learning and memory
(Takayama and Yoshiro, 2004). As the prenatal stage appears to be the period of life
when sensitivity to MeHg is at its greatest, interference of neurotransmission by MeHg
during development could be pathogenic during pregnancy and could cause permanent
birth defects in the CNS such as mental retardation, cerebral palsy, IQ loss, deafness,
blindness, and seizures (Bakir et al., 1973; Amin-Zaki et a1. 1974; Costa etal., 2004;
Trasande et al., 2006). This is signiﬁcant as contamination by MeHg remains a

contemporary problem. According to the EPA, health departments in 45 states have

163

 

 

issued warnings for contamination of freshwater and coastal ﬁsh by mercury.
Amazingly, $8.7 billion dollars per year is spent by the US. in lost worker productivity
due to mercury-related IQ loss (EPA, 2000). The EPA has also reported that 1 in 6
women of childbearing age have higher than normal levels of mercury in their blood.
Based on a recent National Health and Nutrition Examination Survey (l999-2000), more
than 300,000 newborns each year may be exposed in utero to MeHg at concentrations
high enough to produce neurodevelopmental effects (Mahaffey et al., 2004). The
estimated cost to society, attributable to environmental exposures, is in excess of $50
billion (Landrigan et al., 2002). Consequently, it is of vital importance to protect
ourselves as well as our children from MeHg exposure to prevent dysfunction of the

[ICI'VOUS SYSICITI.

164

APPENDIX

A. Suppression of [GABA by MeHg in Granule Cells Is Not Reversible

MeHg is a “sulfur seeking” alkyl-metal and forms covalent bonds with sulﬂrydryl
groups on cysteines, numerous in proteins such as membrane-spanning receptors. The
effects of MeHg are GABAergic synaptic transmission in cerebellar slice can be
irreversible or reversible (Yuan and Atchison, 2003). In ~ 67% of granule cells, sIPSCs
were recovered completely with 20 - 25 min wash with D~pen (lmM) following
suppression by 10 ~ 20 pM MeHg. In Purkinje cells, however, wash with D~pen did not
reverse the effects of MeHg on sIPSCs. To determine if the inhibitory effects of MeHg
( 1 pM) on IGABA are reversible in cerebellar granule neurons in primary culture, cells
were washed for ~ 20 min with MeHg-free recording solution containing D~pen (20 pM)

following complete suppression of IGABA.

Cell Culture Conditions: Cell culture protocol for cerebellar granule cells was similar

to that presented in chapter three.

Whole-Cell Recordings: Recording conditions were similar to those described in

chapter two.

165

Figure A.1. Effects of MeHg on [GABA Are Irreversible in Granule Cells in Culture.
Whole-cell IGABA were evoked from a holding potential of - 60 mV by a 10 ms pulse of
GABA (50 pM, black dot), at intervals of 30 s, in the presence of CNQX (10 pM) and
APV (50 pM) in the external solution to block glutamatergic currents. Data were
collected before (control), after ~ 25 min (hash bar) exposure to 1 pM MeHg, and
following 10 min (hash bar) wash with D~pen-containing (20 pM), MeHg-free solution.

Wash with D~pen-containing solution did not reverse the suppression of IGABA by MeHg.

166

Control MeHg D~pen

' 251/? 5.51 .55

 

500 pA

2000 ms

Figure A.1.

167

B. MeHg Does Not Compete With GABA for Its Binding Site on the GABAA
Receptor

The GABAA receptor has many drug binding sites, including those for
barbiturates, benzodiazepines, and Zn”. However, the site(s) at which MeHg interacts
with. the GABAA receptor has yet to be established. I have shown in the previous
chapters in this thesis that MeHg is not likely to interact with the channel pore or the
diazepam binding site of the GABAA receptor. To determine if MeHg competes with
GABA for its site on the GABAA receptor, whole—cell voltage patch-clamp experiments
were performed in cerebral cortical cells in the presence of 1 pM MeHg and 10, 100, 500,
or 1000 pM GABA. The concentrations of GABA used included both non-saturating (10
and100 pM) and saturating (500 and 1000 pM) concentrations. If MeHg competes with
GABA for its site on the GABAA receptor, it would be expected that the inhibitory
effects of MeHg should be reduced or abolished in the presence of higher concentrations

of GABA.

Cell Culture Conditions: Cell culture protocol for cerebral cortical cells was similar to

that presented in chapter two.

Whole-Cell Recordings: Recording conditions were similar to those described in

chapter two.

168

Figure A.2. MeHg Does Not Interact With the GABA Site on the GABAA Receptor.
Whole-cell IGABA were obtained as described in Fig. A. 1. Data were collected during
exposure to 1 pM MeHg at the concentrations at GABA indicated (10 — 1000 pM, black
bars). Each bar represents the mean values obtained from 3~5 cells. Time-to-complete-
suppression of IGABA by MeHg did not differ between any concentration of GABA tested

(p> 0.05).

169

 

 

500

100

10

 

 

 

wwwmmo

EEC 558.35 2 met.

GABA (11M)

Figure A.2.

170

BIBLIOGRAPHY

Adkins CE, Pillai GV, Kerby J, Bonnert TP, Haldon C, McKeman RM, Gonzalez JE,
Oades K, Whiting PJ and Simpson PB (2001) Alpha4beta3delta GABA(A) receptors

characterized by ﬂuorescence resonance energy transfer-derived measurements of
membrane potential. J Biol Chem 276(42):38934~9.

Akabas MH, Kaufmann C, Archdeacon P and Karlin A (1994) Identiﬁcation of
acetylcholine receptor channel-lining residues in the entire M2 segment of the alpha
subunit. Neuron l3(4):919~27.

Akabas MH, Kaufmann C, Cook TA and Archdeacon P (1994) Amino acid residues
lining the chloride channel of the cystic ﬁbrosis transmembrane conductance regulator.
J Biol Chem 269(21): 14865-8.

Akk G and Steinbach JH (2000) Activation and block of recombinant GABA(A)
receptors by pentobarbitone: a single-channel study. Br J Pharmacol 130(2):249~58.

Amin J, Dickerson IM, and Weiss DS (1994) The agonist binding site of the gamma-

aminobutyric acid type A channel is not formed by the extracellular cysteine loop. Mol
Pharmacol 45(2):3 17-23.

Amin J and Weiss DS (1993) GABA(A) receptor needs two homologous domains of the
beta-subunit for activation by GABA but not by pentobarbital. Nature 366(6455):565~9.

Amin-Zaki L, Elhassani S, Majeed MA, Clarkson TW, Doherty RA and Greenwood M
(1974) Intra-uterine methylmercury poisoning in Iraq. Pediatrics 54(5):587~95.

Arakawa O, Nakahiro M and Narahashi T (1991) Mercury modulation of GABA-
activated chloride channels and non-speciﬁc cation channels in rat dorsal root ganglion
neurons. Brain Res 551(1-2):58-63.

Aschner M and Allen JW (2000) Astrocytes in methylmercury, ammonia, methionine
sulfoximine and alcohol-induced neurotoxicity. Neurotoxicology 21(4):573~9.

171

Aschner M, Yao CP, Allen JW and Tan KH (2000) Methylmercury alters glutamate
transport in astrocytes. Neurochem Int 37(2~3):199~206.

Assaf SY and Chung SH (1984) Release of endogenous Zn2+ from brain tissue during
activity. Nature 308(5961):734~6.

Atchison WD (1986) Extracellular calcium-dependent and ~independent effects of
methylmercury on spontaneous and potassium-evoked release of acetylcholine at the
neuromuscular junction. J Pharmacol Expt Ther 237 :672-80.

Atchison WD and Hare MF (1994) Mechanisms of methylmercury-induced
neurotoxicity. FASEB J 8(9):622-29.

Atchison WD and Narahashi T (1982) Methylmercury-induced depression of
neuromuscular transmission in the rat. Neurotoxicology 3:37-50.

Axtell CD, Cox C, Myers GJ, Davidson PW, Choi AL, Cernichiari E, Sloane-Reeves J,
Shamlaye CF and Clarkson TW (2000) Association between methylmercury exposure
from ﬁsh consumption and child development at ﬁve and a half years of age in the

Seychelles Child Development Study: an evaluation of nonlinear relationships. Environ
Res 84(2):71~80.

Bakir F, Damluji S, Amin-Zaki L, Murtadha M, Khalidi A, Al-Rawi NY, TikritiS,
Dhahir HI, Clarkson TW, Smith JC and Doherty RA (1973) Methylmercury poisoning in
Iraq. Science 181:230~40.

Barberis A, Cherubini E and Mozrzyma SJW(2000) Zinc inhibits miniature GABAergic
currents by allosteric modulation of GABA(A) receptor gating. J Neurosci 20:8619-27.

Barker JL, Behar T, Li YX, Liu QY, Ma W, Maric D, Marie 1, Schaffner AE, Seraﬁni R,
Smith SV, Somogyi R, Vautrin JY, Wen XL and Xian H (1998) GABAergic cells and
signals in CNS development. Perspect Dev Neurobiol S(2-3):305~22.

Barker JL, Harrison NL, Lange GD and Owen DG (1987) Potentiation of gamma-
aminobutyric-acid-activated chloride conductance by a steroid anaesthetic in cultured rat
spinal neurones. J Physiol (Land) 386:485-501.

172

Bastian AJ and Thach WT (1995) Cerebellar outﬂow lesions: a comparison of movement
deﬁcits resulting from lesions at the levels of the cerebellum and thalamus. Ann Neurol
38(6):881~92.

Basu N, Stamler CJ, Loua KM, and Chan HM (2005) An interspecies comparison of
mercury inhibition on muscarinic acetylcholine receptor binding in the cerebral cortex
and cerebellum. Toxicol Appl Pharmacol 205:71-6.

Baumann SW, Baur R and Sigel E (2001) Subunit arrangement of gamma-aminobutyric
acid type A receptors. J Biol Chem 276(39):36275~80.

Belelli D, Lambert J], Peters JA, Wafford K and Whiting PJ (1997) The interaction of the
general anesthetic etomidate with the gamma-aminobutyric acid type A receptor is
inﬂuenced by a single amino acid. Proc Natl Acad Sci U S A 94(20):] 1031-6.

Bellantuono C, Reggi V, Tognoni G and Garattini S (1980) Benzodiazepines: clinical
pharmacology and therapeutic use. Drugs 19(3): 195-219.

Ben-Ari Y (2002) Excitatory actions of GABA during development: the nature of the
nurture. Nat Rev Neurosci 3(9):728~39.

Bera AK, Chatav M and Akabas MH (2002) GABA(A) receptor M2~M3 loop secondary
structure and changes in accessibility during channel gating. J Biol Chem
277(45):43002-10.

Bianchi MT and MacDonald RL (2001) Mutation of the 9' leucine in the GABA(A)
receptor gamma2L subunit produces an apparent decrease in desensitization by
stabilizing open states without altering desensitized states. Neuropharmacology 41:737-
44.

Bianchi MT, Haas KF and MacDonald RL (2001) Structural determinants of fast
desensitization and desensitization-deactivation coupling in GABA(A) receptors.
J Neurosci 15: 1 127-36.

Bianchi MT, MacDonald RL (2002) Slow phases of GABA(A) receptor desensitization
structural determinants and possible relevance for synaptic function. J Physiol (Lond)
544(1):3-18.

173

 

 

 

Bloom FE and Iversen LL (1971) Localizing 3H-GABA in nerve terminals of rat cerebral
cortex by electron microscopic autoradiography. Nature 229(5287):628-30.

Blount P and Merlie JP (1990) Mutational analysis of muscle nicotinic acetylcholine
receptor subunit assembly. J Cell Biol 111(6 Pt 1):2613-22.

Boileau AJ, Evers AR, Davis AF and Czajkowski C (1999) Mapping the agonist binding
site of the GABA(A) receptor: evidence for a beta-strand. J Neurosci l9(12):4847-54.

Brandon NJ, J ovanovic JN, Smart TG and Moss SJ (2002) Receptor for activated C
kinase-1 facilitates protein kinase C-dependent phosphorylation and functional

modulation of GABA(A) receptors with the activation of G-protein-coupled receptors
J Neurosci 22(15):6353-61.

Brickley SG, Cull-Candy SG and Farrant M (1996) Development of a tonic form of
synaptic inhibition in rat cerebellar granule cells resulting from persistant activation of
GABA(A) receptors. J Physiol (Land) 497 :753-759.

Callachan H, Cottrel] GA, Hather NY, Lambert J], N ooney JM and Peters JA (1987)
Modulation of the GABA(A) receptor by progesterone metabolites. Proc R Soc Lond B
Biol Sci 231(1264):359-69.

Castoldi AF, Coccini T, Ceccatelli S and Manzo L (2001) Neurotoxicity and molecular
effects of methylmercury. Brain Res Bull 55:197—203.

Cestari IN, Min KT, Kulli JC and Yang J (2000) Identiﬁcation of an amino acid deﬁning
the distinct properties of murine beta] and beta3 subunit-containing GABA(A) receptors.
J Neurochem 74(2):827-38.

Chang LW (1977) Neurotoxic effects of mercury. Environ Res l4(3):329-73.

Chang LW, Desnoyers PA and Hartmann HA (1972) Quantitative cytochemica] studies
of RNA in experimental mercury poisoning 1. Changes in RNA content. J Neuropathol
Exper Neurol 31:489-501.

174

Chang Y, Wang R, Barot S and Weiss DS (1996) Stoichiometry of a recombinant
GABA(A) receptor. J Neurosci 16:5415-24.

Choi BH (1989) The effects of methylmercury on the developing brain. Prog Neurobiol
32(6):447-70.

Choi BH, Lapham LW, Amin-Zaki L and Saleem T (1978) Abnormal neuronal migration
deranged cerebral cortical organization and diffuse white matter astrocytosis of human

fetal brain a major effect of methylmercury poisoning in utero. J Neuropathol Exp
Neurol 37(6):7l9~33.

Claisse D, Cossa D, Bretaudeau-Sanjuan J, Touchard G and Bombled B (2001)
Methylmercury in molluscs along the French coast. Mar Pollut Bull 42(4):329-32.

Clarkson TW (1993) Mercury major issues in environmental health. Environ Health
Perspect 100:31-8.

Coccini T, Randine G, Candura SM, Nappi RE, Prockop LD and Manzo L (2000) Low-
level exposure to methylmercury modiﬁes muscarinic cholinergic receptor binding
characteristics in rat brain and lymphocytes: physiologic implications and new
opportunities in biologic monitoring. Environ Health Perspect 108(1):29-33.

Concas A, Corda MG, Salis M, Mulas ML, Milia A, Corongiu PP and Biggio G (1983)
Biochemical changes in the rat cerebellar cortex elicited by chronic treatment with
methyl mercury. T oxicol Lett 18:27-33.

Corda MG, Concas A, Rossetti Z, Guarneri P, Corongiu F P and Biggio G (1981) Methyl

mercury enhances [3H]diazepam binding in different areas of the rat brain. Brain Res
229:264-9.

Cordier S, Gare] M, Mandereau L, Morce] H, Doineau P Gosme-Seguret S, Josse D,
White R and Amiel-Tison C (2002) Neurodevelopmental investigations among
methylmercury-exposed children in French Guiana. Environ Res 89(1): 1-1 1.

Costa E and Guidotti A (1979) Recent studies on the mechanism whereby
benzodiazepines facilitate GABA-ergic transmission. Adv Exp Med Biol 123:371-8.

175

Costa LG, Aschner M, Vitalone A, Syversen T and Soldin OP (2004) Developmental
neuropathology of environmental agents. Annu Rev Pharmacol T oxicol 44:87-110.

Crump KS, Kjellstrom T, Shipp AM, Silvers A and Stewart A (1998) Inﬂuence of
prenatal mercury exposure upon scholastic and psychological test performance
benchmark analysis of a New Zealand cohort. Risk Anal l8(6):701~13.

Dalziel JE, Cox GB, Gage PW and Bimir B (2000) Mutating the highly conserved second
membrane-spanning region 9' leucine residue in the alpha(1) or beta(1) subunit produces

subunit-speciﬁc changes in the function of human alpha(1)beta(1) gamma-aminobutyric
Acid(A) receptors. Mol Pharmacol 57(5):875-82.

Davidson P W, Myers G J, Cox C, Shamlaye C F, Marsh DO, Tanner MA, Berlin M,
Sloane-Reeves J, Cemichiari E, Choisy O, Choi A, and Clarkson T W (1995)
Longitudinal neurodevelopmental study of Seychellois children following in utero
exposure to methylmercury from matema] ﬁsh ingestion outcomes at 19 and 29 months.
Neurotoxicology 16:677-88.

Davidson P W, Myers G J, Cox C, Axtell C, Shamlaye C, Sloane-Reeves J, Cemichiari
E, Needham L, Choi A, Wang Y, Berlin M, and Clarkson T W (1998) Effects of prenatal
and postnatal methylmercury exposure from ﬁsh consumption on neurodevelopment:
outcomes at 66 months of age in the Seychelles Child Development Study. JAMA
280:701-7.

De Schutter E (2002) Cerebellar cortex: computation by extrasynaptic inhibition? Curr
Biol 12(10):R363~5.

De Schutter E and Maex R (1996) The cerebellum: cortical processing and theory. Curr
Opin Neurobiol 6(6):759-64.

DeLorey TM and Olsen RW (1992) Gamma-aminobutyric acidA receptor structure and
function. J Biol Chem 267(24): 16747-50.

Denny MF and Atchison WD (1994) Methylmercury-induced elevations in
intrasynaptosoma] zinc concentrations an 19F-NMR study. J Neurochem 63:383-6.

176

Denny MF and Atchison WD (1996) Mercurial-induced alterations in neuronal divalent
cation homeostasis. Neurotoxicology l7(l):47-61.

Denny MF, Hare MF and Atchison WD (1993) Methylmercury alters intrasynaptosomal
concentrations of endogenous polyvalent cations. Toxicol Appl Pharmacol 122:222-32.

Didier M, Xu M, Berrnan SA, Saido TC and Bursztajn S (1997) Involvement of three
glutamate receptor epsilon subunits in the formation of N-methyl-D-aspartate receptors
mediating excitotoxicity in primary cultures of mouse cerebellar granule cells.
Neuroscience 78(4):1 129-46.

Draguhn A, Verdoom TA, Ewert M, Seeburg PH and Sakmann B (1990) Functional and

molecular distinction between recombinant rat GABA(A) receptor subtypes by Zn2+-
Neuron 5:781-88.

Edwards JR, Marty MS and Atchison WD (2005) Comparative sensitivity of rat
cerebellar neurons to dysregulation of divalent cation homeostasis and cytotoxicity
caused by methylmercury. Toxicol Appl Pharmacol 208:222-32.

EPA (1997) Mercury report to Congress. Washington, DC: US Environmental
Protection Agency, Ofﬁce of Air Quality Standards.

EPA (2000) The mercury research strategy. Washington, DC: US Environmental
Protection Agency, Ofﬁce of Research and Development.

Eto K and Takeuchi T (1978) A pathological study of prolonged cases of Minamata
Disease. With particular reference to 83 autopsy cases. Acta Pathol Jpn 28(4):565-84.

Fisher JL and MacDonald RL (1998) The role of an a subtype M2-M3 his in regulating

inhibition of GABA(A) receptor current by zinc and other divalent cations. J Neurosci
18:2944-53.

Fisher JL, Zhang J and MacDonald RL (1997) the role of alphal and alpha6 subtype
amino-terminal domains in allosteric regulation of ~aminobutyric acida receptors. Mol
Pharmacol 52:714-24.

177

 

Fonfria E, Rodriguez-Farre E and Sunol C (2001) Mercury interaction with the
GABA(A) receptor modulates the benzodiazepine binding site in primary cultures of
mouse cerebellar granule cells. Neuropharmacology 41:819-33.

Forsyth DS Casey V, Dabeka RW and McKenzie A (2004) Methylmercury levels in
predatory ﬁsh species marketed in Canada. Food Addit Contam 21(9):849-56.

Gallo V, Kingsbury A, Balazs R and Jorgensen OS (1987) The role of depolarization in
the survival and differentiation of cerebellar granule cells in culture. J Neurosci

7(7):2203~ l 3.

Gallo V, Levi G, Raiteri M and Coletti A (1981) Enhancement by GABA of glutamate
depolarization-induced release from cerebellar nerve endings. Brain Res 205(2):431-5.

Gao B and Fritschy J-M (1995) Cerebellar granule cells in vitro recapitulate the in vivo
pattern of GABA(A)-receptor subunit expression. Develop Brain Res 88: 1-16.

Gerstenberger SL and Dellinger J A (2002) PCBs, mercury, and organochlorine
concentrations in lake trout, walleye, and Whiteﬁsh from selected tribal ﬁsheries in the
Upper Great Lakes region. Environ Toxicol 17:513-9.

Gilbertson M (2004) Male cerebral palsy hospitalization as a potential indicator of
neurological effects of methylmercury exposure in Great Lakes communities.
Environ Res 95:375-84.

Grandjean P, Weihe P, White RF and Debes F (1998) Cognitive performance of children
prenatally exposed to "safe" levels of methylmercury. Environ Res 77(2): 165-72.

Grandjean P, Weihe P, White RF, Debes F, Araki S, Yokoyama K, Murata K, Sorensen
N, Dahl R and Jorgensen PJ (1997) Cognitive deﬁcit in 7-year-old children with prenatal
exposure to methylmercury. Neurotoxicol Teratol l9(6):417-28.

Grandjean P, White RF, Nielsen A, Cleary D and de Oliveira Santos EC (1999)
Methylmercury neurotoxicity in Amazonian children downstream from gold mining.
Environ Health Perspect 107(7):587-91.

178

Grosman C, Salamone FN, Sine SM and Auerbach A (2000) The extracellular linker of
muscle acetylcholine receptor channels is a gating control element. J Gen Physiol
116(3):327-40.

Haas KF and MacDonald RL (1999) GABA(A) receptor subunit gamma2 and delta
subtypes confer unique kinetic properties on recombinant GABA(A) receptor currents in
mouse ﬁbroblasts. J Physiol (Land) 514:27-45.

Hamill OP, Borrnann J and Sakmann B (1983) Activation of multiple-conductance state
chloride channels in spinal neurones by glycine and GABA. Nature 305(5937):805-8.

Hamill OP, Marty A, Neher E, Sakmann B and Sigworth FJ (1981) Improved patch-
clamp techniques for high-resolution current recording from cells and cell-free membrane
patches. Eur J Physiol 391(2):85-100.

Hamori J and Szentagothai J (1966) Participation of Golgi neuron processes in the
cerebellar glomeruli: an electron microscope study. Exp Brain Res 2(1):35-48.

Hanchar HJ, Dodson PD, Olsen RW, Otis TS and Wallner M (2005) Alcohol-induced
motor impairment caused by increased extrasynaptic GABA(A) receptor activity. Nature
Neurosci 8:339-45.

Hansen SL, Ebert B, Fjalland B and Kristiansen U (2001) Effects of GABA(A) receptor
partial agonists in primary cultures of cerebellar granule neurons and cerebral cortical
neurons reﬂect different receptor subunit compositions. Br J Pharmacol l33(4):539-49.

Hare MF and Atchison WD (1995) Methylmercury mobilizes Ca2+ from intracellular
stores sensitive to inositol 1.4.5-trisphosphate in NG108-15 cells. J Pharmacol Exp Ther
272: 1016-23.

Hare MF and Atchison WD (1995) Nifedipine and tetrodotoxin delay the onset of
methylmercury-induced increase in [Ca2+]i in NG108-15 cells. T oxicol Appl Pharmacol
135:299-307.

Hare MF, McGinnis KM and Atchison WD (1993) Methylmercury increases intracellular

concentrations of Ca2+ and heavy metals in NG108-15 cells. J Pharmacol Exp Ther
266:1626-35.

179

Herden CJ, Yuan Y and Atchison WD (2004) Effects of methylmercury (MeHg) on
GABA(A) receptor-mediated currents in rat cortical cells in culture. Toxicologist
78: 1 122.

Hevers W and Lttddens H (1998) The diversity of GABA(A) receptors. Pharmacological
and electrophysiological properties of GABA(A) channel subtypes. Molec Neurobiol
18:35-86.

Hewett SJ and Atchison WD (1992) Effects of charge and lipophilicity on mercurial-

induced reduction of 45Ca2+ uptake in isolated nerve terminals of the rat. Toxicol Appl
Pharmacol 113(2):267-73.

Hosli E and Hosli L (1995) Autoradiographic studies on the uptake of 3H-noradrenaline
and 3H-serotonin by neurones and astrocytes in explant and primary cultures of rat CNS:
effects of antidepressants. Int J Dev Neurosci 13(8):897-908.

Howell GA, Welch MG and Frederickson CJ (1984) Stimulation-induced uptake and
release of zinc in hippocampal slices. Nature 308(5961):736-8.

Huang CS and Narahashi T (1996) Mercury chloride modulation of the GABA(A)
receptor-channel complex in rat dorsal root ganglion neurons. Toxicol Appl Pharmacol
140:508-20.

Huang CS and Narahashi T (1997) The role of G proteins in the activity and mercury
modulation of GABA-induced currents in rat neurons. Neurapharmacalagy 36(11-
12): 1623-30.

Huang CS and N arahashi T (1997) The role of phosphorylation in the activity and

mercury modulation of GABA-induced currents in rat neurons. Neurapharmacalagy
3601-12): 1631-40.

Hunter D and Russell D (1945) Focal cerebral and cerebellar atrophy in a human subject
due to organic mercury compounds. Neurol Neurosurg Psychiat 17:235-41.

Hunter D, Bomford RR and Russell DS (1940) Poisoning by methyl mercury compounds.
Q J Med 35: 193-218.

180

Irn WB, Binder JA, Dillon GH, Pregenzer JF, Im HK Altman RA (1995) Acceleration of
GABA-dependent desensitization by mutating threonine 266 to alanine of the alpha 6
subunit of rat GABA(A) receptors. Neurosci Lett l86(2-3):203-7.

Ingleﬁeld JR, Mundy WR and Shafer, TJ (2001) Inositol 1,4,5-triphosphate receptor-
sensitive Ca2+ release, store-operated Ca2+ entry, and CAMP responsive element binding
protein phosphorylation in developing cortical cells following exposure to
polychlorinated biphenyls. J Pharmacol Exp Ther 297 :762-73.

Inoue M, Oomura, Y, Yakushiji, T, and Akaike, N (1986) Intracellular calcium ions
decrease the afﬁnity of the GABA receptor. Nature (324): 156-8.

Itouji A, Sakai N, Tanaka C and Saito N (1996) Neuronal and glial localization of two
GABA transporters (GATl and GAT3) in the rat cerebellum. Brain Res Mol Brain Res
37(1-2):309-16.

Jackel C, Kleinz R, Makela R, Hevers W, Jezequel S, Korpi ER and Luddens H (1998)
The main determinant of furosemide inhibition on GABA(A) receptors is located close to
the ﬁrst transmembrane domain. Eur J Pharmacol 357(2-3):251-6.

Jakab RL and Hamori J (1988) Quantitative morphology and synaptology of cerebellar
glomeruli in the rat. Anat Embryol (Berl) 179(1):81-8.

Jechlinger M, Pelz R, Tretter V, Klausberber T and Sieghart W (1998) Subunit
composition and quantitative importance of heterooligomeric receptors GABA(A)
receptors containing (16 subunits. J Neurosci 18:2449-57.

Jones A, Korpi ER, McKeman RM, Pelz R, Nusser Z, Makela R, Mellor JR, Pollard
S, Bahn S, Stephenson FA, Randall AD, Sieghart W, Somogyi P, Smith AJH and
Wisden W (1997) Ligand-gated ion channel subunit partnerships GABA(A) receptor (16
subunit gene inactivation inhibits 6 subunit expression. J Neurosci 17:1350-62.

Jones MV and Westbrook GL (1995) Desensitized states prolong GABA(A) channel
responses to brief agonist pulses. Neuron 15:181-91.

Jones-Davis DM and MacDonald RL (2003) GABA(A) receptor function and
pharmacology in epilepsy and status epilepticus. Curr Opin Pharmacol 3(1): 12-8.

181

Kandel E, Schwartz J, and Jessell, T (2000) Principles of Neural Science 4‘h ed.
McGraw-Hill, New York.

Kardos J (1999) Recent advances in GABA research. Neurochem Int 34(5):353-8.

Karlin A and Akabas MH (1995) Toward a structural basis for the function of nicotinic
acetylcholine receptors and their cousins. Neuron 15(6): 1231-44.

Kash TL, Jenkins A, Kelley JC, Trudell JR and Harrison NL (2003) Coupling of agonist
binding to channel gating in the GABA(A) receptor. Nature 421(6920):272~5.

Kehrig HA, Malm O, Akagi H, Guimaraes JR and Torres J P (1998) Methylmercury in
ﬁsh and hair samples from the Balbina Feservoir Brazilian Amazon. Environ Res 77:84-
90.

Keinanen K, Wisden W, Sommer B, Werner P, Herb A, Verdoom TA, Sakmann B and
Seeburg PH (1990) A family of AMPA-selective glutamate receptors. Science
249(4968):556-60.

Khan ZU, Gutierrez A and De Blas AL (1994) The subunit composition of a
GABA(A)/benzodiazepine receptor from rat cerebellum. J Neurochem 63:371-4.

Khan ZU, Gutierrez A and De Blas AL (1996) The (11 and a6 subunits can coexist in the
same cerebellar GABA(A) receptor maintaining their individual benzodiazepine-binding
speciﬁcities. J Neurochem 66:685-91.

Kimelberg HK, Pang S, and Treble DH (1989) Excitatory amino acid-stimulated uptake
of 22Na“ in primary astrocyte cultures. J Neurosci 9(4):1 141-9.

Knoﬂach F, Benke D, Wang Y, Scheurer L, Luddens H, Hamilton B], Carter DB, Mohler
H and Benson JA (1996) Pharmacological modulation of the diazepam-insensitive
recombinant gamma-aminobutyric acidA receptors alpha 4 beta 2 gamma 2 and alpha 6
beta 2 gamma 2. Mol Pharmacol 50(5): 1253-61.

182

Kolbaev SN, Sharonova IN, Vorobjev VS and Skrebitsky VG (2002) Mechanisms of
GABA(A) receptor blockade by millimolar concentrations of furosemide in isolated rat
Purkinje cells. Neurapharmacalogy 42:913-21.

Komulainen H and Tuomisto J (1985) 3H-dopamine uptake and 3H-haloperidol binding in
striatum after administration of methyl mercury to rats. Arch Toxicol 57:268-71.

Korpi ER and Luddens H (1993) Regional gamma-aminobutyric acid sensitivity of t-
butylbicyclophosphoro[35S]thionate binding depends on gamma-aminobutyric acid A
receptor alpha subunit. Mol Pharmacol 44:87-92.

Korpi ER, Kuner T, Seeburg PH and Luddens H (1995) Selective antagonist for the
cerebellar granule cell-speciﬁc 7-aminobutyric acid Type A receptor Mal Pharmacol
47 :283-9.

Korpi ER, Kleingoor C, Kettenmann H and Seeburg PH (1993) Benzodiazepine-induced

motor impairment linked to point mutation in cerebellar GABA(A) receptor. Nature
361(6410):356-9.

Kostial K and Landeka M (1975) The action of mercury ions on the release of
acetylcholine from presynaptic nerve endings. Experientia 31(7):834-5.

Landrigan PJ, Schechter CB, Lipton JM, Fahs MC and Schwartz J (2002) Environmental
pollutants and disease in American children: estimates of morbidity, mortality, and costs
for lead poisoning, asthma, cancer, and developmental disabilities. Environ Health
Perspect 110(7):721-8.

Laurie DJ, Seeburg PH and Wisden W (1992) The distribution of 13 GABAA receptor
subunit mRNAs in the rat brain. 11. Olfactory bulb and cerebellum. J Neurosci
12(3): 1063-76.

Laurie DJ, Wisden W and Seeburg PH (1992) The distribution of thirteen GABA(A)
receptor subunit mRN As in the rat brain. III. Embryonic and postnatal development.
J Neurosci 12(11):4151-72.

183

Le Novere N, Corringer PJ and Changeux JP (2002) The diversity of subunit composition
in nAChRs evolutionary origins physiologic and pharmacologic consequences. J
Neurobiol 53(4):447-56.

Levenberg KA (1944) Method for the solution of certain problems in least squares. Quart
Appl Math 2:164-8.

Levesque PC and Atchison WD (1991) Disruption of brain mitochondria] calcium
sequestration by methylmercury. J Pharmacol Exp Ther 256(1):236-42.

Leyshon-Sorland K, Jasani B, Morgan AJ (1994) The localization of mercury and
metallothionein in the cerebellum of rats experimentally exposed to methylmercury.
Histochem J 26(2): 161-9.

Limke TL and Atchison WD (2002) Acute exposure to methylmercury opens the
mitochondrial permeability transition pore in rat cerebellar granule cells. Toxicol Appl
Pharmacol l78(1):52~6l.

Limke TL, Bearss JJ and Atchison WD (2004) Acute exposure to methylmercury causes
Ca2+ dysregulation and neuronal death in rat cerebellar granule cells through an M3
muscarinic receptor-linked pathway. Toxicol Sci 80(1):60-8.

Luddens H and Wisden W (1991) Function and pharmacology of multiple GABA(A)
receptor subunits. Trends Pharmacol Sci l2(2):49-51.

Luddens H, Pritchett DB, Kohler M, Killisch I, Keinanen K, Monyer H, Sprengel R and
Seeburg PH (1990) Cerebellar GABA(A) receptor selective for a behavioural alcohol
antagonist. Nature 346(6285):648-51.

Ma J Y and Narahashi T (1993) Enhancement of gamma-aminobutyric acid-activated
chloride channel currents by lanthanides in rat dorsal root ganglion neurons. Neurosci
13:4872-9.

MacDonald RL and Barker JL (1978) Different actions of anticonvulsant and anesthetic
barbiturates revealed by use of cultured mammalian neurons. Science 200(4343):775-7.

184

MacDonald RL and Olsen RW (1994) GABA(A) receptor channels. Annu Rev Neurosci
17:569-602.

MacDonald RL and Twyman RE (1992) Kinetic properties and regulation of GABA(A)
receptor channels. Ion Channels 3:315-43.

MacDonald RL, Rogers CJ and Twyman RE (1989) Kinetic properties of the GABA(A)
receptor main conductance state of mouse spinal cord neurones in cell culture. J Physiol
(Land ) 410:479-99.

MacDonald RL, Rogers CJ and Twyman RE (1989) Barbiturate regulation of kinetic
properties of the GABA(A) receptor channel of mouse spinal neurones in culture. J
Physiol (Land) 417:483-500.

Maconochie DJ, Zempel JM and Steinbach JH (1994) How quickly can GABA(A)
receptors open? Neuron 12(1):61~71.

Maex R and De Schutter E (2003) Resonant synchronization in heterogeneous networks
of inhibitory neurons. J Neurosci 23(33): 10503-14.

Mahaffey KR, Clickner RP, Bodurow CC (2004) Blood organic mercury and dietary
mercury intake: National Health and Nutrition Examination Survey, 1999 and 2000.
Environ Health Perspect 112(5):562-70.

Magos L (1975) Letter; Variation of biological half-life of methylmercury in man. Arch
Environ Health 30(7):371.

Makela R, Uusi-Okari M, Homanics GE, Quinlan JJ, Fierstone LL, Wisden W and
Korpi ER (1997) Cerebellar 7-aminobutyric acid type A receptors pharmacological
subtypes revealed by mutant mouse lines. Mal Pharmacol 53:380-8.

Makela R, Uusi-Oukari M, Oja SS, Alho H, Anghelescu I, Klawe C, Luddens H and
Korpi ER (1999) Furosemide action on cerebellar GABA(A) receptors in alcohol-
sensitive ANT rats. Alcohol 19(3): 197-205.

Makela R, Wisden W and Korpi ER (1999) Loreclezole and La3+ differentiate cerebellar
granule cell GABA(A) receptor subtypes. Eur J Pharmacol 367(1): 101-5.

185

Manalis RS and Cooper GP (1975) Evoked transmitter release increased by inorganic
mercury at frog neuromuscular junction. Nature 257(5528):690~1.

Marquardt D (1963) An algorithm for least-squares estimation of nonlinear parameters.
SIAM J Appl Math 11:431-41.

Marsh DO, Clarkson TW, Myers GJ, Davidson PW, Cox C, Cemichiari E, Tanner MA,
Lednar W, Shamlaye C, Choisy O, Horeau C and Berlin M (1995) The Seychelles study
of fetal methylmercury exposure and child development. Neurotoxicalagy 16:583-96.

Marsh DO, Turner MD, Smith JC, Allen P and Richdale N (1995) Fetal methylmercury
study in a peruvian ﬁsh-eating population. Neurataxicalagy 16:717-26.

Marty MS and Atchison WD (1997) Pathways mediating Ca2+ entry in rat cerebellar
granule cells following in vitro exposure to methyl mercury. Toxicol Appl Pharmacol
147:319-30.

Marty MS and Atchison WD (1998) Elevations of intracellular Ca2+ as a probable
contributor to decreased viability in cerebellar granule cells following acute exposure to
methylmercury. Toxicol Appl Pharmacol 150:98-105.

McKeown-Eyssen GE and Ruedy J (1983) Methylmercury exposure in northern Quebec.
1. Neurologic ﬁndings in adults. Am J Epidemiol 118(4):461-9.

McKeown-Eyssen GE, Ruedy J and Neims A (1983) Methyl mercury exposure in
northern Quebec. 11. Neurologic ﬁndings in children. Am J Epidemiol 118(4):470—9.

Meacham CA, Freudenrich TM, Anderson WL, Sui L, Lyons-Darden T, Barone S Jr,
Gilbert ME, Mundy WR and Shafer TJ (2005) Accumulation of methylmercury or
polychlorinated biphenyls in in vitro models of rat neuronal tissue. Toxicol Appl
Pharmacol 205(2): 177-87.

Mellor JR and Randall A (1997) Frequency-dependent actions of benzodiazepines on
GABA(A) receptors in cultured murine cerebellar granule cells. J Physiol (Land)
503:353-69.

186

Mellor JR and Randall AD (1998) Voltage-dependent deactivation and desensitization of

GABA responses in cultured murine cerebellar granule cells. J Physiol (Land)
506(2):377-90.

Mihalek RM, Banerjee PK, Korpi ER, Quinlan JJ, Firestone LL, Mi ZP, Lagenaur C,
Tretter V, Sieghart W, Anagnostaras SG, Sage JR, Fanselow MS, Guidotti A, Spigelman
1, Li Z, DeLorey TM and Olsen RW, and Homanics GE (1999) Attenuated sensitivity to
neuroactive steroids in gamma-aminobutyrate type A receptor delta subunit knockout
mice. Proc Natl Acad Sci U S A 96(22): 12905-10.

Mihic SJ, Ye Q, Wick MJ, Koltchine VV, Krasowski MD, Finn SE, Mascia MP,
Valenzuela CF, Hanson KK, Greenblatt EP, Harris RA and Harrison NL (1997) Sites of

alcohol and volatile anaesthetic action on GABA(A) and glycine receptors. Nature
389(6649):385~9.

Miller LG, Greenblatt DJ, Roy RB, Gaver A, Lopez F and Shader RI (1989) Chronic
benzodiazepine administration. III. Upregulation of gamma-aminobutyric acidA receptor
binding and function associated with chronic benzodiazepine antagonist administration. J
Pharmacol Exp Ther 248: 1096-1 101.

Mintz 1M, Adams ME and Bean BP (1992) P-type calcium channels in rat central and
peripheral neurons. Neuron 9(1):85-95.

Moss SJ and Smart TG (1996) Modulation of amino acid-gated ion channels by protein
phosphorylation. Int Rev Neurobiol 39: 1-52.

Miura K and Imura N (1989) Mechanism of cytotoxicity of methylmercury. With special
reference to microtubule disruption. Biol Trace Elem Res 21:313-6.

Miura K, Koide N, Himeno S, Nakagawa I, and Imura N (1999) The involvement of
microtubular disruption in methylmercury-induced apoptosis in neuronal and
nonneuronal cell lines. Toxicol Appl Pharmacol. 160(3):279-88.

Murata K, Weihe P, Renzoni A, Debes F, Vasconcelos R, Zino F, Araki S, Jorgensen PJ,
White RF and Grandjean P (1999) Delayed evoked potentials in children exposed to
methylmercury from seafood. Neurataxicol T eratal 21(4):343-8.

187

N agaya N and MacDonald RL (2001) Two gamma2L subunit domains confer low Zn2+
sensitivity to ternary GABA(A) receptors. J Physiol (Land) 532217-30.

Narahashi T, Ma J Y, Arakawa O, Reuveny E and Nakahiro M (1994) GABA receptor-
channel complex as a target site of mercury, copper, zinc, and lanthanides. Cell Mal
Neurobiol 14:599-621.

Neustadt A, Frostholm A and Rotter A (1988) Topographical distribution of muscarinic
cholinergic receptors in the cerebellar cortex of the mouse, rat, guinea pig, and rabbit: a
species comparison. J Comp Neural 272(3):317-30.

Nicol] RA and Wojtowicz J M (1980) The effects of pentobarbital and related compounds
on frog motoneurons. Brain Res 19l(1):225-37.

Nierenberg DW, Nordgren RE, Chang MB, Siegler RW, Blayney MB, Hochberg F,
Toribara TY, Cemichiari E and Clarkson T (1998) Delayed cerebellar disease and death
after accidental exposure to dimethylmercury. NEJM 338: 1672-6.

Nusser Z, Roberts J DB, Baude A, Richardson JG and Somogyi P (1995) Relative
densities of synaptic and extrasynaptic GABA(A) receptors on cerebellar granule cells as
determined by quantitative immunogold method. J Neurosci 15:2948-60.

N usser Z, Sieghart W and Somogyi P (1998) Segregation of different GABA(A)
receptors to synaptic and extrasynaptic membranes of cerebellar granule cells. J
Neurosci 18: 1693-1703.

Nusser Z, Sieghart W, Stephenson FA and Somogyi P (1996) The (16 subunit of the
GABA(A) receptor is concentrated in both inhibitory and excitatory synapses on
cerebellar granule cells. J Neurosci 16:103-14.

O'Kusky JR and McGeer EG (1985) Methylmercury poisoning of the developing nervous
system in the rat: decreased activity of glutamic acid decarboxylase in cerebral cortex and
neostriatum. Brain Res 353:299-306.

O'Kusky JR and McGeer EG (1989) Methylmercury-induced movement and postural
disorders in developing rat: high-afﬁnity uptake of choline, glutamate, and gamma-

188

aminobutyric acid in the cerebral cortex and caudate-putamen. J Neurochem 53:999-
1006.

Olsen RW (1981) The GABA postsynaptic membrane receptor-ionophore complex. Site
of action of convulsant and anticonvulsant drugs. Mal Cell Biochem 39:261-79.

Olsen RW (1987) GABA-drug interactions. Prag Drug Res 31:223-41.

Olsen RW and Avoli M (1997) GABA and epileptogenesis. Epilepsia 38(4):399-407.

Olsen RW and Tobin AJ (1990) Molecular biology of GABA(A) receptors. FASEB J
4(5): 1469-80.

Olsen RW, Yang J, King RG, Dilber A, Stauber GB, Ransom RW (1986) Barbiturate and
benzodiazepine modulation of GABA receptor binding and function. Life Sci
39(21): 1969-76.

Owens DF and Kriegstein AR (2002) Is there more to GABA than synaptic inhibition?
Nat Rev Neurosci 3(9):715~27.

Peng S, Hajela RK and Atchison WD (2002) Effects of methylmercury on human
neuronal L-type calcium channels transiently expressed in human embryonic kidney cells
(HEK-293). J Pharmacol Exp Ther 302(2):424-32.

Pollard S, Duggan MJ, Stephenson FA (1993) Further evidence for the existence of alpha
subunit heterogeneity within discrete gamma-aminobutyric acidA receptor
subpopulations. J Biol Chem 268(5):3753~7.

Poltl A, Hauer B, Fuchs K, Tretter V and Sieghart W (2003) Subunit composition and
quantitative importance of GABA(A) receptor subtypes in the cerebellum of mouse and
rat. J Neurochem 87(6): 1444-55.

Pritchett DB, Luddens H and Seeburg PH (1989) Type I and type II GABA(A)-
benzodiazepine receptors produced in transfected cells. Science 245(4924): 1389-92.

189

 

Puia G, Ducic I, Vicini S and Costa E (1993) Does neurosteroid modulatory efﬁcacy
depend on GABA(A) receptor subunit composition? Receptors Channels 1(2): 135-42.

Puia G, Mienville JM, Matsumoto K, Takahata H, Watanabe H, Costa E and Guidotti A
(2003) On the putative physiological role of allopregnanolone on GABA(A) receptor
function. Neuropharmacology 44(1):49-55.

Puia G, Santi MR, Vicini S, Pritchett DB, Purdy RH, Paul SM, Seeburg PH and Costa E
(1990) Neurosteroids act on recombinant human GABA(A) receptors. Neuron 4(5):759~
65.

Quirk K, Gillard NP, Ragan CI, Whiting PJ, and McKeman RM (1994) Model of subunit
composition of gamma-aminobutyric acid A receptor subtypes expressed in rat

cerebellum with respect to their alpha and gamma/delta subunits. J Biol Chem
269(23) 1 6020-8.

Randall A and Tsien RW (1995) Pharmacological dissection of multiple types of Ca2+
channel currents in rat cerebellar granule neurons. J Neurosci 15(4):2995~3012.

Rice DC (1995) Neurotoxicity of lead, methylmercury, and PCBs in relation to the Great
Lakes. Environ Health Perspect 103:71-87.

Rodier PM (1995) Developing brain as a target of toxicity. Environ Health Perspect
103(6):73-6.

Rossi P, De Filippi G, Armano S, Taglietti V and D'Angelo E (1998) The weaver
mutation causes a loss of inward rectiﬁer current regulation in premigratory granule cells
of the mouse cerebellum. J Neurosci 18:3537-47.

Sakmann B, Hamill OP and Bormann J (1983) Patch-clamp measurements of elementary
chloride currents activated by the putative inhibitory transmitter GABA and glycine in
mammalian spinal neurons. J Neural Trans Suppl 18:83-95.

Saraﬁan TA (1993) Methyl mercury increases intracellular Ca2+ and inositol phosphate
levels in cultured cerebellar granule neurons. J Neurochem 61(2):648-57.

190

Saxena NC and MacDonald RL (1994) Assembly of GABA(A) receptor subunits role of
the delta subunit. J Neurosci 14:7077-86.

Saxena NC and MacDonald RL (1996) Properties of putative cerebellar 7-aminobutyric
acid A receptor isoforms. Mal Pharm 49:567-79.

Saxena NC, Neelands TR and MacDonald RL (1997) Contrasting actions of lanthanum
on different recombinant Gamma-aminobutyric acid receptor isoforms expressed in L929
ﬁbroblasts. Mal Pharmacol 51(2):328-35.

Schmid G, Bonanno G, Raiteri L, Sarviharju M, Korpi ER and Raiteri M (1999)
Enhanced benzodiazepine and ethanol actions on cerebellar GABA(A) receptors

mediating glutamate release in an alcohol-sensitive rat line. Neuropharmacalagy
38: 1273-9.

Schoﬁeld PR, Darlison MG, Fujita N, Burt DR, Stephenson FA, Rodriguez H, Rhee LM,
Ramachandran J, Reale V, and Glencorse TA (1987) Sequence and functional expression

of the GABA A receptor shows a ligand-gated receptor super-family. Nature
328(6127):221-7.

Schulz DW and MacDonald RL (1981) Barbiturate enhancement of GABA-mediated
inhibition and activation of chloride ion conductance: correlation with anticonvulsant and
anesthetic actions. Brain Res 209(1): 177-88.

Segal M and Barker JL (1984) Rat hippocampal neurons in culture Potassium
conductances. J Neuraphysial 51: 1409-33.

Semyanov A, Walker MC, Kullmann DM and Silver RA (2004) Tonically active
GABA(A) receptors modulating gain and maintaining the tone. Trends Neurosci 27 :262-
9.

Shafer TJ, Contreras ML and Atchison WD (1990) Characterization of interactions of
methylmercury with Ca2+ channels in synaptosomes and pheochromocytoma cells
radiotracer ﬂux and binding studies. Molec Pharmacol 38:102-13.

Shamlaye CF, Marsh DO, Myers GJ, Cox C Davidson PW, Choisy O, Cemichiari E,
Choi A, Tanner MA and Clarkson TW (1995) The Seychelles child development study

191

on neurodevelopmental outcomes in children following in utero exposure to
methylmercury from a maternal ﬁsh diet background and demographics.
Neurataxicalogy l6(4):597-6l2.

Sieghart W (1992) GABAA receptors: ligand-gated Cl‘ ion channels modulated by
multiple drug-binding sites. Trends Pharmacol Sci 13:446-50.

Sieghart W ( 1995) Structure and pharmacology of gamma-aminobutyric acidA receptor
subtypes. Pharmacol Rev 47(2): 181-234.

Sigel E, Baur R, Kellenberger S and Malherbe P (1992) Point mutations affecting
antagonist afﬁnity and agonist dependent gating of GABA(A) receptor channels. EMBO
J ll(6):2017-23.

Sirois JE and Atchison WD (1996) Effects of mercurials on ligand- and voltage-gated ion
channels a review. Neurataxicalagy 17:63-84.

Smart TG and Constanti A (1990) Differential effect of zinc on the vertebrate GABA(A)-
receptor complex. Br. J Pharmacol 99:643-54.

Smart TG, Moss SJ, Xie X and Huganir RL (1991) GABA(A) receptors are differentially
sensitive to zinc dependence on subunit composition. Br J Pharmacol 103: 1837-9.

Steuerwald U, Weihe P, Jorgensen PJ, Bjerve K, Brock J, Heinzow B, Budtz-Jorgensen E
and Grandjean P (2000) Maternal seafood diet methylmercury exposure and neonatal
neurologic function. J Pediatr l36(5):599-605.

Study RE and Barker JL (1981) Diazepam and (~~)-pentobarbital: ﬂuctuation analysis
reveals different mechanisms for potentiation of gamma-aminobutyric acid responses in
cultured central neurons. Proc Natl Acad Sci U S A 78(11):7180-4.

Sumikawa K and Gehle VM (1992) Assembly of mutant subunits of the nicotinic
acetylcholine receptor lacking the conserved disulﬁde loop structure. J Biol Chem

267(9):6286-90.

Syversen TLM (1981) Effects of methyl mercury on protein synthesis in vitro. Acta
Pharmacol Toxicol 49:422-6.

192

Takayasu Y, Iino M, Furuya N and Ozawa S (2003) Muscarine-induced increase in
frequency of spontaneous EPSCs in Purkinje cells in the vestibulo-cerebellum of the rat.
J Neurosci 23(15):6200-8.

Takeuchi T, and Eto K (1999) The pathology study of Minamata Disease: a tragic story
of water pollution. Kyushu University Press, INC. Okinawa, Japan.

Takeuchi T, Norikawa N, Matsumoto H and Shiraishi Y (1962) A pathological study of
Minamata Disease in Japan. Acta Neuropathol 2:40-57.

Takeuchi T, Eto K and Tokunaga H (1989) Mercury level and histochemical distribution
in a human brain with Minamata Disease following a long-term clinical course of twenty-
six years. Neurataxicolagy 10(4):651-7.

Takeuchi T, Matsumoto H, Sashi M, Kambara T, Shiraishi Y, Hirata Y, Nobukiro M and
Ito H (1968) Pathology of Minamatma disease. Kumamata Med J 34:521-4.

Thompson CL and Stephenson FA (1994) GABA(A) receptor subtypes expressed in
cerebellar granule cells: a developmental study. J Neurochem 62:2037-44.

Thompson SA, Arden SA, Marshall G, Wingrove PB, Whiting PJ and Wafford KA
(1999) Residues in transmembrane domains I and II detemrine gamma-aminobutyric

acid type A receptor subtype-selective antagonism by furosemide. Mol Pharmacol
55(6):993-9.

Thompson SA, Whiting PJ and Wafford KA (1996) Barbiturate interactions at the human
GABA(A) receptor: dependence on receptor subunit combination. Br J Pharmacol
117(3):521-7.

Tia S, Wang JP, Kotchabhakdi N and Vicini S (1996) Distinct deactivation and
desensitization kinetics of recombinant GABA(A) receptors. Neurapharmacalagy
35:1375-82.

Toxicological Effects of Methylmercury (2000) Committee on the Toxicological Effects
of Methylmercury. Board on Environmental Studies and Toxicology Commission on
Life Sciences. National Research Council National Academy Press: Washington, DC.

193

Trasande L, Schechter CB, Haynes KA and Landrigan PJ (2006) Mental retardation and
prenatal methylmercury toxicity. Am J Ind Med 49(3): 153-8.

Traxinger DL and Atchison WD (1987) Reversal of methylmercury-induced block of

nerve-evoked release of acetylcholine at the neuromuscular junction. Toxicol Appl
Pharmacol 90(1):23-33.

Twyman RE and MacDonald RL (1992) Neurosteroid regulation of GABA(A) receptor
single-channel kinetic properties of mouse spinal cord neurons in culture. J Physiol
456:2]5-45.

Twyman RE, Green RM and MacDonald RL (1992) Kinetics of open channel block by
penicillin of single GABA(A) receptor channels from mouse spinal cord neurones in
culture. J Physiol 445:97-127.

Twyman RE, Rogers CJ and MacDonald RL (1990) Intraburst kinetic properties of the
GABA(A) receptor main conductance state of mouse spinal cord neurones in culture, J
Physiol 423: 193-220.

Varju P, Katarova Z, Madarasz E and Szabo G (2001) GABA signalling during
development: new data and old questions. Cell Tissue Res 305(2):239-46.

Vicini S and Mienville JM (1987) Costa E. Actions of benzodiazepine and beta-carboline
derivatives on gamma-aminobutyric acid-activated Cl- channels recorded from
membrane patches of neonatal rat cortical neurons in culture. J Pharmacol Exp Ther

243(3):1195-201.

Vicini S, Ferguson C, Prybylowski K, Kralic J, Morrow AL and Homanics GE (2001)
GABA(A) receptor alpha] subunit deletion prevents developmental changes of inhibitory
synaptic currents in cerebellar neurons. J Neurosci 21:3009-16.

Vicini S, Losi G and Homanics GE (2002) GABA(A) receptor delta subunit deletion
prevents neurosteroid modulation of inhibitory synaptic currents in cerebellar neurons.
Neurapharmacolagy 43(4):646-50.

Vicini S, Mienville JM, and Costa E (1986) Flunitrazepam action on the GABA-Cl
ionophore complex in rat cortical neurons in culture: a patch-clamp study.
Clin Neurapharmacol 9(4):395-7.

194

Weihe P, Hansen JC, Murata K, Debes F, Jorgensen P, Steuerwald U, White RF and
Grandjean P (2002) Neurobehavioral performance of Inuit children with increased
prenatal exposure to methylmercury. Int J Circumpolar Health 61(1):41-9.

Weis IM (2004) Mercury concentrations in ﬁsh from Canadian Great Lakes areas of
concern: an analysis of data from the Canadian Department of Environment database.
Environ Res 95:341-50.

Westbrook GL and Mayer ML (1987) Micromolar concentrations of Zn2+ antagonize
NMDA and GABA responses of hippocampal neurons. Nature 328(6131):640-3.

Westh-Hansen SE, Witt MR, Dekerrnendjian K, Liljefors T, Rasmussen PB and Nielsen
M (1999) Arginine residue 120 of the human GABA(A) receptor alpha 1, subunit is

essential for GABA binding and chloride ion current gating. Neurarepart 10(11):2417-
21.

World Health Organization (1993) Evaluation of certain food additives and contaminants.
Forty-ﬁrst report of the joint FAO/WHO expert committee on food additives. WHO
Technical Report Series No. 837 (Geneva: WHO).

Wieland HA, Luddens H and Seeburg PH (1992) A single histidine in GABA(A)
receptors is essential for benzodiazepine agonist binding. J Biol Chem 267(3): 1426-9.

Wisden W and Seeburg PH (1992) GABA(A) receptor channels from subunits to
functional entities. Curr Opin Neurobiol 2(3):263~9.

Wisden W, Korpi ER and Bahn S (1996) The cerebellum a model system for studying
GABA(A) receptor diversity. Neuropharmacalagy 35:1139-60.

Wohlfarth KM, Bianchi MT and MacDonald RL (2002) Enhanced neurosteroid
potentiation of ternary GABA(A) receptors containing the delta subunit. J Neurosci
22(5): 1541-9.

Wood JM, Kennedy FS and Rosen CG (1968) Synthesis of methyl-mercury compounds
by extracts of a methanogenic bacterium. Nature 220(163):173-4.

195

Xie XM and Smart TG (1991) A physiological role for endogenous zinc in rat
hippocampal synaptic neurotransmission. Nature 349(6309):521-4.

Xu M and Akabas MH (1996) Identiﬁcation of channel-lining residues in the M2
membrane-spanning segment of the GABA(A) receptor alphal subunit. J Gen Physiol
107(2): 195-205.

Xu YF and Atchison WD (1997) Methylmercury blocks 7- anrinobutyric acid (GABA(A)
current and induces a nonspeciﬁc inward current in rat cerebellar granule cells. Sac
Neurosci Abstr 23: 373.

Yuan Y and Atchison WD (1993) Disruption by methylmercury of membrane excitability
and synaptic transmission of CA1 neurons in hippocampal slices of the rat. Toxicol Appl
Pharmacol 120:203-15.

Yuan Y and Atchison WD (1994) Comparative effects of inorganic divalent mercury
methylmercury and phenylmercury on membrane excitability and synaptic transmission
of cal neurons in hippocampal slices of the rat. Neurataxicalagy 15:403-12.

Yuan Y and Atchison WD (1995) Methylmercury acts at multiple sites to block
hippocampal synaptic transmission. J Pharmacol Exp Ther 275:1308-16.

Yuan Y and Atchison WD (1997) Action of methylmercury on GABA(A) receptor-
mediated inhibitory synaptic transmission is primarily responsible for its early
stimulatory effects on hippocampal CA1 excitatory synaptic transmission. J Pharmacol
Exp Ther 282264-73.

Yuan Y and Atchison WD (1999) Comparative effects of methylmercury on parallel-
ﬁber and climbing-ﬁber responses of rat cerebellar slices. J Pharmacol Exp Ther
288: 1015-25.

Yuan Y and Atchison WD (2003) Methylmercury differentially affects GABA(A)
receptor-mediated spontaneous IPSCs in Purkinje and granule cells of rat cerebellar
slices. J Physiol (Land) 550: 191-204.

196

Yuan Y and Atchison WD (2005) Methylmercury induces a spontaneous transient slow
inward chloride current in Purkinje cells of rat cerebellar slices. J Pharmacol Exp Ther
313:751-764.

Yuan Y, Otero-Montanez J KL, Yao A, Herden CJ, Sirois J E, Atchison WD (2005)
Inwardly rectifying and voltage-gated outward potassium channels exhibit low sensitivity
to methylmercury. Neurataxicalagy 26:439-54.

Zempel JM and Steinbach JH (1995) Neonatal rat cerebellar granule and Purkinje
neurons in culture express different GABA(A) receptors. Eur J Neurosci 7(9): 1895-905.

Zhang S-P, Natsukari N, Bai G, Nichols RA and Weiss B (1993) Localization of the
multiple calmodulin messenger RNAs in differentiated PC12 cells. Neuroscience 55:571-
82.

Zhu W], Wang JF, Corsi L and Vicini S (1998) Lanthanum-mediated modiﬁcation of
GABA(A) receptor deactivation, desensitization and inhibitory synaptic currents in rat
cerebellar neurons. J Physiol ( Land) 51 l(3):647-61.

Zhu W], Wang JF, Krueger KB and Vicini S (1996) Delta subunit inhibits neurosteroid
modulation of GABA(A) receptors. J Neurosci l6(21):6648~56.

197