INVOLVEMENT OF THE GABAA RECEPTOR IN CALCIUM-DEPENDENT TOXICITY OF
METHYLMERCURY IN DEVELOPING AND AGED CEREBELLUM
By
Aaron Blackwood Bradford

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Biochemistry and Molecular Biology - Environmental Toxicology - Doctor of Philosophy
2013

ABSTRACT
INVOLVEMENT OF THE GABAA RECEPTOR IN CALCIUM-DEPENDENT TOXICITY OF
METHYLMERCURY IN DEVELOPING AND AGED CEREBELLUM
by
Aaron Blackwood Bradford
Methylmercury (MeHg) is an environmental neurotoxicant of concern because of its bioaccumulation in marine ecosystems and the susceptibility of developing cerebellum to toxicity.
MeHg accumulates in the central nervous system with a long latency to toxicity and excretion. It
induces permanent motor dysfunction through selective toxicity in cerebellar granule cells
(CGCs), with developing CGCs being a particular target. Though the exact pathways by which
MeHg induces neuronal death and dysfunction are not fully characterized, one of the common,
central pathways by which it is hypothesized to act is through disruption of intracellular calcium
2+

2+

concentration ([Ca ]i). [Ca ]i is kept low in neurons and small changes are critical for neuronal
signaling. Disruption of these concentrations occurs in several cell models, characterized as an
2+

2+

irreversible increase in [Ca ]i over time. We sought to characterize disruption of [Ca ]i in
cerebellar slices, which retain many characteristics of whole tissue. One pathway by which
methylmercury may act on intracellular calcium is through the γ-aminobutyric acid type A
receptor (GABAAR), a normally inhibitory receptor that is excitatory in developing CGCs and
2+

possibly in axons and presynaptic terminals. This excitation is coupled to influx of [Ca ]i and is
critical to developmental migration of neurons and to neurotransmitter release. The interaction of
GABAAR s with MeHg is partially characterized by electrophysiology and there are several
experiments that suggest certain subunits of the receptors are associated with cell and
developmental susceptibility. We hypothesized that both MeHg and GABAAR s could have
2+

effects on [Ca ]i in CGCs in acutely-isolated slices of cerebellum and that these effects would
2+

be dependent on age and GABAAR subunit expression. To monitor [Ca ]i we used

concentration-dependent fluorophores and imaging on confocal and epifluorescent microscopes.
2+

We first tested for [Ca ]i changes in neonatal rat, a model frequently used previously for MeHg
toxicity studies. In neonatal rat CGCs, MeHg caused a concentration-dependent increase in
2+

[Ca ]i-dependent fluorescence within minutes. These fluorescence increases included up to
about 2-fold in mature CGCs compared to starting fluorescence and up to about 3-fold increase
in susceptible developing CGCs. Muscimol, a GABAAR agonist, and bicuculline, a GABAAR
2+

antagonist, are both able to reduce the effects of MeHg on [Ca ]i. MeHg is able to increase
2+

[Ca ]i in cerebellar slices from mice as well, and a knockout of the GABAAR α6 subunit is able
to increase the susceptibility of CGCs, particularly developing CGCs, with similar 2-3 fold
increases in fluorescence. Muscimol and bicuculline reduce MeHg effects in these mice as well.
Compared to developmental toxicity, aging and lifelong exposures to MeHg have not been
characterized nearly as fully, and as part of a comprehensive study, the effects of MeHg and
2+

several other [Ca ]i-dependent mechanisms were tested with these fluorescence microscopy
approaches. Acute application of MeHg appears to cause 2-fold increases in fluorescence that are
not dependent on age or chronic treatment. However, chronic treatment with MeHg appears to be
able to increase both spontaneous and induced Ca

2+

activity in slices, and chronic treatment with

isradipine appears to reduce this activity, indicating chronic, non-lethal concentrations can have
2+

subtle effects on Ca -dependent mechanisms. Finally, this dissertation provides some
2+

characterization of spontaneous [Ca ]i transients in slices, indicating age and genotype have
effects on spontaneous CGC activity. Altogether, the results of this dissertation indicate that
GABAAR s have a partial role in the toxicity of MeHg as a pathway by which MeHg appears to
2+

be able to disrupt [Ca ]i as interference by GABAAR modulators and alterations to subunit
2+

expression also alter MeHg effects on [Ca ]i. MeHg exposure may in turn alter GABAAR
signaling as indicated by changes to muscimol responses in chronically-exposed mice.

Copyright by
AARON BLACKWOOD BRADFORD
2013

This dissertation is dedicated to everyone that believed I could do it. Having learned stagecraft in
high school and college, I am especially thankful to anyone and everyone that worked behind the
scenes to make sure that I could continue, with or without my knowledge.

v

ACKNOWLEDGMENTS

First I need to thank my family. My parents, grandparents and extended family
encouraged my pursuits of education and science from an early age and always kept me going
even when I faltered. I want to thank my wife for caring for me at my worst and always pushing
me to get on with things. I want to thank my grandpa Blackwood, he didn’t get to see me
accomplish this, but I think he knew it was where I was headed and would be proud to see the
high levels of achievement in the family. I want to thank my little sister Nora for always thinking
the science I am doing is interesting and cool.
I want to thank Dr. William “Bill” Atchison and my thesis committee for carefully
considering my proposal and experiments, and always being able to come up with interesting
suggestions and corrections. I also want to thank everyone for pushing me to produce a complete
dissertation, rather than settling for a Masters’ thesis when the possibility came up.
I want to thank the Center for Integrative Toxicology, including its directors Dr. Roth and
Dr. Kaminsky as well as the support staff, for taking me on as a fellowship student and always
finding a way to support me when I needed it. I want to thank MSU’s Institute for CyberEnabled Research and Dr. Colbry for providing extensive imaging training, and MSU’s Center
for Statistical Training and Consulting for extensive statistical training. I want to thank Dr. Sisk
and Dr. Schneider for training in writing.
I want to thank my friends, particularly my fellow graduate students both inside and
outside the laboratory. I might seem like an introvert, but laughing, socializing, venting and
griping are still needed. I hope I was as generous, kind, honest, fun and loyal with my friends as
they were with me.

vi

TABLE OF CONTENTS

LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x
KEY TO SYMBOLS AND ABBREVIATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xx
CHAPTER ONE: INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
A)
BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
a.
Methylmercury . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
i.
MeHg in the environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
ii.
MeHg in human health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
iii.
MeHg chemistry and mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . 7
iv.
MeHg toxicity susceptibility and resistance . . . . . . . . . . . . . . . . 12
b.
Calcium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
i.
Calcium signaling and regulation . . . . . . . . . . . . . . . . . . . . . . . . 15
ii.
Calcium and neurotransmission . . . . . . . . . . . . . . . . . . . . . . . . . . 17
iii.
Calcium in development and aging . . . . . . . . . . . . . . . . . . . . . . . 18
iv.
MeHg and calcium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
c.
Cerebellum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
i.
Cerebellar circuitry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
ii.
Cerebellum growth and neuron development . . . . . . . . . . . . . . . 23
d.
GABAA receptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
i.
GABAA receptor structure and function . . . . . . . . . . . . . . . . . . . 26
ii.
GABAA receptor signaling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
iii.
GABAA receptors and clacium . . . . . . . . . . . . . . . . . . . . . . . . . . 32
iv.
GABAA receptor pharmacology . . . . . . . . . . . . . . . . . . . . . . . . . 33
v.
GABAA receptors during development . . . . . . . . . . . . . . . . . . . . 35
vi.
GABAA receptor disorders and toxicology . . . . . . . . . . . . . . . . . 37
B)
OBJECTIVES AND RATIONALE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
a.
Hypothesis and aims . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2+
b.
Ca imaging in cerebellar slices from rats and mice as a model for MeHg
toxicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
CHAPTER TWO: METHYLMERCURY-ASSOCIATED INCREASES IN FLUO4
FLUORESCENCE IN NEONATAL RAT CEREBELLAR SLICES ARE DEPENDENT
ON GRANULE CELL MIGRATIONAL STAGE AND MODULATED BY GABAA
RECEPTORS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
A)
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
B)
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
C)
Materials and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Chemicals and solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Acute slice preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Confocal microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Image analysis and statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
D)
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
vii

E)

Visualization of layers and cytotoxicity . . . . . . . . . . . . . . . . . . . . . . . . . 65
Acute MeHg slice treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Muscimol and bicuculline treatment with MeHg . . . . . . . . . . . . . . . . . . 72
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78

CHAPTER THREE: ACUTE METHYLMERCURY TREATMENT INCREASES FLUO4
FLUORESCENCE IN NEONATAL MOUSE CEREBELLAR SLICES AND IS
DEPENDENT ON GABAAR α6 SUBUNIT EXPRESSION, MIGRATIONAL STAGE AND
GABAAR MODULATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
A)
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
B)
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
C)
Materials and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
Chemicals and solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
Animal care and genotyping of transgenic animals . . . . . . . . . . . . . . . . . 89
Acute slice preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Confocal microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Image analysis and statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
D)
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Visualization of layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Acute MeHg paired treatment in (+/+) and (-/-) mice . . . . . . . . . . . . . . 100
Muscimol and bicuculline paired treatments with MeHg . . . . . . . . . . . 102
E)
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
2+

CHAPTER FOUR: COMPARATIVE Ca EFFECTS OF CHRONIC EXPOSURE OF
ADULT MICE TO MeHg OR ISRADIPINE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
A)
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
B)
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
C)
Materials and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
Chemicals and solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
Animal care and treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
Tissue collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
High-speed imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
Confocal microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
D)
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
2+
Confocal Ca imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
2+
High-speed Ca imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
E)
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
2+

2+

CHAPTER FIVE: HIGH-SPEED Ca IMAGING REVEALS SPONTANEOUS Ca
TRANSIENTS IN CEREBELLAR SLICES FROM YOUNG AND OLD MICE . . . . . . . 158
A)
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
B)
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
C)
Materials and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Chemicals and solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Animal care and use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Image acquisition and treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
viii

D)
E)

Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
Visualization of CGCs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
2+
Spontaneous Ca transients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175

CHAPTER SIX: SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
Summary and conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195

ix

LIST OF FIGURES

Figure 1.1

Schematic representation of neuronal MeHg targets and the central action of
2+
[Ca ]i in normal neuronal function and toxicity. This schematic represents an
axon terminal, but many of these mechanisms are present throughout a neuron.
2+
MeHg acts on Ca channels (such as VGCCs), metabotropic receptors that signal
through IP3 (such as the M3 muscarinic receptor), and the GABAA receptor at the
surface of neurons (red arrows). MeHg is also able to enter cells and act on
proteins such as the mitochondrial transition pore (mPTP), but the exact
2+
mechanisms of entry are unclear (dashed red arrow). [Ca ]i is normally increased
2+
2+
transiently through several mechanisms, such as Ca -dependent Ca release
2+
from ER through ryanodine receptors, but MeHg increasec [Ca ]i beyond normal
2+
levels (black arrows). The action by which MeHg can increase [Ca ]i through
GABAA receptors is unclear, but may be dependent on VGCCs through
2+
depolarization or other mechanisms (dashed black line). Some Ca may be
2+
2+
buffered by Ca binding proteins in the cytosol, but excess Ca is normally
2+
taken up by the ER through Ca -ATPases such as the smooth endoplasmic
2+
reticulum Ca ATPase (SERCA), and may be expelled from the cytoplasm by
other ATP-dependent mechanisms, both of which require ATP from the
2+
mitochondria (blue arrows). The normal actions of increased [Ca ]i include
regulating vesicle release, neuronal motility during development, and regulation
2+
of gene expression, but the excess of Ca from MeHg actions can cause excess
neurotransmitter release, aberrant gene expression, and can activate several
enzymes involved in apoptosis, such as lipases and calpains (magenta arrows).
For interpretation of the references to color in this and all other figures, the reader
is referred to the electronic version of this dissertation. . . . . . . . . . . . . . . . . . . . 10

Figure 1.2

Schematic of CGC migration and maturation in the cerebellum. CGCs migrate
from the base of the developing cerebellum through several stages (gray arrows)
during the third trimester in humans and the first two postnatal weeks in rodents,
both of which are considered to be the most susceptible periods for MeHg toxicity
in each group. CGC precursors (orange) migrate along the surface of the
cerebellum, accumulate in a temporary layer termed the external granule (or
germinal) layer (EGL) and divide. Division terminates for these CGC precursors,
and the resulting CGCs then migrate down Bergmann glial processes (green)
through the molecular layer (ML) where PCs are developing extensive dendrites
(blue). Migration stops, and the CGCs put out axons that travel back through the
ML and produce parallel fibers which synapse onto PC dendrites. All stages of
2+
migration and maturation are dependent on finely-tuned Ca oscillations and
regulation by expression of surface receptors such as GABAARs (Komuro and
Rakic, 1998). CGCs express mostly α2 subunit-containing GABAARs during
development, while expressing mostly α6 in maturation (Takayama and Inoue,
2004). PCs express α3 subunits during development and α1 at maturation.
Adapted from Atchison 2005. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
x

Figure 1.3

Cartoon depiction of the GABAA receptor, a heteropentameric ion channel that
conducts Cl . In vivo it typically forms a unit from two α (from 6 subtypes), two β
(from 3 subtypes), and one additional subunit, including γ and δ. GABAARs have
two GABA binding sites, located at the interface of α and β subunits, which based
on binding studies are also the location of the agonist muscimol and antagonist
bicuculline binding sites. The construction of these interfaces controls GABA
binding and thus the physiology of the receptor. An additional site where
benzodiazepines (BZs) bind and potentiate GABA activation is located between α
and γ subunits, so GABAARs without γ subunit are insensitive to these
modulators. Synaptic scaffolds such as gephyrin bind to certain subtypes of
GABAARs, clustering them inside synapses, whereas others such as those
containing α6 can be found spread around the membrane. The direction of ion
flow of depends on the Cl gradient. An inward gradient is present at most postsynaptic domains and most mature cells and activation produces
hyperpolarization and thus inhibition of activity. However, in developing neurons
there is no extrusion of Cl through potassium-chloride co-transporter (KCC2),
and thus the gradient is outward, so activation causes depolarization. This reversal
of the gradient may also be present in some locations of mature cells, such as
axons. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

Figure 2.1

A A representative pseudocolor fluorescent image of an acutely-isolated
cerebellar slice from PND9 rat showing modified fluo4/ethD-1 live/dead
cytotoxicity assay, post-incubation control. Images are from internal folds of the
developing cerebellum. Images show maximum fluo4 and ethD-1 fluorescence
values by pixel. Approximate pial surface of an inner fold is indicated by a white
line and other boundaries are indicated by dashed lines. EthD-1 (red), an indicator
of dead cells, and Fluo4AM (green), an indicator of relative intracellular calcium
concentration, do not colocalize. Scale bar represents 30 μm. B Representative
image from separate slice from the same cerebellum after one additional h in
ACSF supplemented with 20 μM MeHg. C EthD-1 mean pixel staining (% pixels
± SEM) comparison of layers post-incubation in fluo4 and ethD-1 (n=5), with an
additional 1 h incubation in ACSF (n=3) or with an additional 1 h incubation in
20 μM MeHg ACSF (n=3). Abbreviations: EGL, external granule layer; ML,
molecular layer; IGL, internal granule layer. . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

Figure 2.2

40x grayscale examples of cerebellar slices before (-10) and after (40 min)
treatment with MeHg. Images show projection of maximum fluo4 fluorescence
by pixel. A Fluorescence does not change over the course of imaging in control
slices. B MeHg increased fluorescence in all layers, with the signal from many
cells saturating the detector. Scale bars are 60 μm in all images. Surfaces and
boundaries are indicated as with Figure 1.1. White arrows indicate examples of
cells with extinguished fluorescence by the end of treatment. . . . . . . . . . . . . . . 57

Figure 2.3

Relative fluo4 fluorescence intensity (F/F0) in CGCs by histological layer in
slices from neonatal rat cerebellum exposed to concentrations of MeHg
continuously for 30 min. Control MeHg concentrations did not change
xi

fluorescence significantly. Continuous perfusion of ACSF with 0 μM (B, n=5), 1

μM (?, n=3), 5 μM (>, n=5), 10 μM (Ë, n=6) or 20 μM MeHg (C, n=4) were
compared in A, the EGL, B, the ML, and C, the IGL. Significant differences are
indicated from untreated control (*) and between 20 μM and 10 μM treatments
(#) at the same time point. Intensity values are expressed in arbitrary units (mean
± SEM). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
2

Figure 2.4

Cell density per (100 μm) for CGCs monitored for fluo4 fluorescence intensity,
in each layer (EGL, ML and IGL), for tissue treated with different concentrations
of MeHg. Fluorescent CGCs were more dense in the EGL and IGL layers than in
the ML. At higher MeHg concentrations, fluorescent CGCs in the EGL were
more dense than in IGL, but at lower concentrations, density was higher in the
IGL. Within the EGL, fluorescent cell density increased significantly with greater
concentrations of continuous MeHg treatment, non-overlapping letters are
significantly different comparisons (P<0.05). Paired comparisons were made
between each concentration of MeHg by layer. Significantly different pairs within
each layer are indicated by different letters (P<0.05). n for each treatment is the
same as in Figure 2.3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

Figure 2.5

Fluorescent cell loss “extinguished” counts from slices during treatment with
several concentrations of MeHg over 30 min. Cells extinguished when they lost
significant fluo4 fluorescence between time points. Fluo4 fluorescence loss was
occasionally associated with visible lysis of cells, so presumably all fluorescence
loss is associated with cell death. Comparisons were made between each
concentration of MeHg and untreated control, no comparisons were significantly
different (P>0.05). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

Figure 2.6

Effect of bicuculline and MeHg on relative fluo4 fluorescence intensity (F/F0) of
CGCs by histological layer. Untreated controls (B, n=5) and 20 μM MeHg alone

(C, n=5) are the same as in Figure 2.3. Bicuculline pulses (arrows, 10 μM) with 20
μM MeHg ( , n=4) were compared to bicuculline alone ( , n=3) and MeHg
alone. Significant difference are indicated from bicuculline control (*) and from
MeHg alone (#) at that time point, P<0.05. Intensity values are expressed in
arbitrary units (mean ± SEM). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
2

Figure 2.7

Cell density (per (100 μm) ) for CGCs monitored for fluo4 fluorescence
intensity, in each layer (EGL, ML and IGL) and density for the total tissue, for
tissue treated with two concentrations of MeHg (0 and 20 μM) and two
concentrations of bicuculline (0 and 10 μM). CGCs were more densely distributed
in the EGL and IGL layers than in the ML. Paired comparisons were made
between each concentration of MeHg by layer. Significantly different pairs within
each layer are indicated by different letters (P<0.05). Values are mean ± SEM, for
the same n as described in Figure 2.6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70

Figure 2.8

Percentage of “extinguished” cells in experiments comparing MeHg and
xii

bicuculline treatment. Comparisons were made between 20 and 0 μM MeHg
treatments, between bicuculline alone and 0 μM MeHg, between bicuculline
alone and bicuculline combined with 20 μM MeHg, and between bicuculline
combined with MeHg and 20 μM alone. Significantly different pairs within each
layer are indicated by different letters (P<0.05). Values are mean ± SEM, for the
same n as described in Figure 2.6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Figure 2.9

Effect of muscimol and MeHg on relative fluo4 fluorescence intensity (F/F0) of
CGCs by histological layer. Untreated controls (B, n=5) and 20 μM MeHg alone

(C, n=5) are the same as in figure 3. Muscimol pulses (arrows, 100 μM) with 20
μM MeHg ( , n=4) were compared to muscimol alone ( , n=3) and MeHg alone.
Significant difference are indicated from muscimol control (*) and from MeHg
alone (#) at that time point, P<0.05. Intensity values are expressed in arbitrary
units. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
2

Figure 2.10

Cell density (per (100 μm) ) for CGCs monitored for fluo4 fluorescence
intensity, in each layer (EGL, ML and IGL) and density for the total tissue, for
tissue treated with two concentrations of MeHg (0 and 20 μM) and two
concentrations of muscimol (0 and 100 μM). CGCs were more densely distributed
in the EGL and IGL layers than in the ML. Paired comparisons were made
between each concentration of MeHg by layer. Significantly different pairs within
each layer are indicated by different letters (P<0.05). Values are mean ± SEM
with n the same as described in Figure 2.9. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76

Figure 2.11

Percentage of “extinguished” cells in experiments comparing MeHg and
muscimol treatment. Comparisons were made between 20 and 0 μM MeHg
treatments, between muscimol alone and 0 μM MeHg, between muscimol alone
and muscimol combined with 20 μM MeHg, and between muscimol combined
with MeHg and 20 μM alone. Significantly different pairs within each layer are
indicated by non-overlapping letters (P<0.05). Values are mean ± SEM with n the
same as described in Figure 2.9. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77

Figure 3.1

Representative images of frames of interest and labeled layers from (-/-) mouse
cerebellar slices. Images are from two slices; both slices are from the same (-/-)
mouse. Detailed description of assigning boundaries is discussed in Chapter Two
materials and methods. Layers and boundaries from surface of cerebellum (solid
line) include the external granule layer (EGL), the EGL-ML interface (dotted
line), the molecular layer (ML), the Purkinje layer (dashed line) and the internal
granule layer (IGL). The left images are prior to administration of MeHg, and
right images are the last image taken from respective series. The top slice was
treated with only ACSF (0 µM MeHg, untreated control) for 55 min, the bottom
slice was treated to 20 µM MeHg in ACSF for 55 min. Untreated control had little
change to fluorescence, but 20 µM treatment increased fluorescence ~2-fold in
the EGL in this example, with smaller increases in the ML and IGL. Scale bar is
equal to 50 µm. Images are summary projections of ~30 scans through the depth
of the slice, intensity values are adjusted to the value histogram of the first image.
xiii

Images are cropped to remove blank space around edges from image
registration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Figure 3.2

Time course of effects of MeHg on fluo4 fluorescence in cerebellar slices from
neonatal α6 (+/+) and (-/-) mice. Mean fluo4 fluorescence intensity was compared
between 0 (untreated control, Con) and 20 µM MeHg (MeHg) in (-/-) (circles,
n=5) and (+/+) (squares, n=6) mice over 55 min of treatment. Each of the three
developmental layers are shown separately. MeHg is applied continuously for the
duration of the experiment starting at t=0 (20 µM, black; 0 µM, white). MeHg
treatment causes a rapid increase in fluo4 fluorescence in all layers and
genotypes, with the largest change in the (-/-) EGL. * represents time points that
are significantly greater than pretreatment (P<0.05) and ** represents time points
that are significantly greater than pretreatment as well as greater than the paired,
untreated time point (P<0.05). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

Figure 3.3

Extinguished cell percentages as found in paired mouse cerebellar slices treated
with 0 (ACSF) and 20 µM MeHg (MeHg) from (-/-) (white bars) and (+/+) (black
bars) mice. Extinguished cells are cells that rapidly and significantly lose fluo4
fluorescence intensity as compared to the normal variation during treatment.
These cells are removed from time course intensity traces and tallied. In these
MeHg treatments, only the CGCs in (+/+) IGL significantly extinguish as
compared to their untreated paired slices (*, P<0.05) . . . . . . . . . . . . . . . . . . . . 101

Figure 3.4

Time course of effects of MeHg and muscimol on fluo4 fluorescence in cerebellar
slices from neonatal α6 (-/-) mice. Mean fluo4 fluorescence intensity was
compared between 20 µM MeHg (MeHg, white circles) and 20 μM MeHg with
100 μM muscimol pulses (Muscimol, black circles) in (-/-) mice over 55 min of
treatment (± SEM, n=5). Each of the three developmental layers are shown
separately. MeHg is applied continuously for the duration of the experiment
starting at t=0 and muscimol is pulsed into the feed for 60 s at 9.5, 19.5, 29.5,
39.5 and 49.5 min (arrows). MeHg alone causes a rapid increase in fluo4
fluorescence in all layers in a manner similar to previous paired experiments in (/-) mice. Muscimol pulses reduced this effect immediately and kept fluo4
fluorescence near baseline for the duration of experiments. * represents time
points that are significantly greater than pretreatment (P<0.05). . . . . . . . . . . . 104

Figure 3.5

Time course of effects of MeHg and muscimol on fluo4 fluorescence in cerebellar
slices from neonatal α6 (+/+) mice. Mean fluo4 fluorescence intensity was
compared between 20 µM MeHg (MeHg, white circles) and 20 μM MeHg with
100 μM muscimol pulses (Muscimol, black circles) in (+/+) mice over 55 min of
treatment (± SEM, n=5). Each of the three developmental layers are shown
separately. MeHg is applied continuously for the duration of the experiment
starting at t=0 and muscimol is pulsed into the feed for 60 s at 9.5, 19.5, 29.5,
39.5 and 49.5 min (arrows). MeHg alone causes a rapid increase in fluo4
fluorescence in all layers in a manner similar to previous paired experiments in
(+/+) mice. Muscimol pulses reduced this effect in all layers and kept fluo4
fluorescence near baseline for the duration of experiments in the EGL and ML. *
xiv

represents time points that are significantly greater than pretreatment
(P<0.05). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
Figure 3.6

Time course of effects of MeHg and bicuculline on fluo4 fluorescence in
cerebellar slices from neonatal α6 (-/-) mice. Mean fluo4 fluorescence intensity
was compared between 20 µM MeHg (MeHg, white circles) and 20 μM MeHg
with 10 μM bicuculline pulses (Bicuculline, black circles) in (-/-) mice over 55
min of treatment (± SEM, n=5). Each of the three developmental layers are shown
separately. MeHg is applied continuously for the duration of the experiment
starting at t=0 and bicuculline is pulsed into the feed for 60 s at 9.5, 19.5, 29.5,
39.5 and 49.5 min (arrows). MeHg alone causes a rapid increase in fluo4
fluorescence in all layers in a manner similar to previous paired experiments in (/-) mice. Bicuculline pulses reduced this effect immediately and kept fluo4
fluorescence near baseline for the duration of experiments in the IGL. *
represents time points that are significantly greater than pretreatment
(P<0.05). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108

Figure 3.7

Time course of effects of MeHg and bicuculline on fluo4 fluorescence in
cerebellar slices from neonatal α6 (+/+) mice. Mean fluo4 fluorescence intensity
was compared between 20 µM MeHg (MeHg, white circles) and 20 μM MeHg
with 10 μM bicuculline pulses (Bicuculline, black circles) in (+/+) mice over 55
min of treatment (± SEM, n=6). Each of the three developmental layers are shown
separately. MeHg is applied continuously for the duration of the experiment
starting at t=0 and bicuculline is pulsed into the feed for 60 s at 9.5, 19.5, 29.5,
39.5 and 49.5 min (arrows). MeHg alone causes an increase in fluo4 fluorescence
in all layers in a manner similar to previous paired experiments in (+/+) mice.
Bicuculline did not reduce the effect in any layers, but also did not increase fluo4
fluorescence in the ML compared to pretreatment. * represents time points that
are significantly greater than pretreatment (P<0.05). . . . . . . . . . . . . . . . . . . . . 110

Figure 4.1

Example frames of interest from confocal fluo4 time course imaging. Slices were
imaged through 45 min of continuous perfusion with 20 μM MeHg in 6 month
chronically-treated mice. All images are from the same batch of animals. Images
are registered, rotated and cropped for clarity; scale bar equals 50 μm. A. Slices of
cerebellum are from four 9 month old animals treated chronically for 6 months
with, in order, standard food and water (O), 6.25 ppm MeHg in drinking water
(M) with standard food, 2 ppm isradipine in food (I) with standard water, and the
combination of MeHg in water and isradipine in food (M/I). MeHg application
over 45 min increases relative fluo4 fluorescence in CGCs in the IGL by about 2fold within 45 min, the duration of observation, in all 4 chronic treatment groups.
B. Expanded frame of interest from group O after 45 min of continuous MeHg
perfusion. Image highlights processes in the ML that become apparent with
MeHg treatment (arrows), infrequent PC fluorescence (boxes), and non CGC cells
in the ML (circles). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135

Figure 4.2

Mean relative fluo4 fluorescence intensities as measured from confocal time
course image series of cerebellum slices from 6 month chronically-treated mice.
xv

Slices from untreated control (O, n=11, white circles), MeHg-treated (M, n=11,
black circles), isradipine-treated (I, n=9, white triangles), and combination MeHgand isradipine-treated (M/I, n=11, black triangles) were imaged at 5 min intervals
for 50 min with one pretreatment scan and 10 images with 20 µM MeHg flowing
over the slices. All acute MeHg treatments caused about a 2-fold increase in
relative fluo4 fluorescence in CGCs (± SEM) and reached significance compared
to pretreatment by 20 min of imaging (P<0.05), but were not significantly
different from one another (P>0.05). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
Figure 4.3

Mean relative fluo4 fluorescence intensities of CGCs in slice during 20 μM
MeHg exposure over 45 min from 12 month chronically-treated mice. Traces are
from acute treatments of 0 µM MeHg (untreated controls, white circles) and 20
µM MeHg (black circles) applied in the bath ACSF solution to cerebellum from
15 month old mice chronically treated for 12 months. Paired comparisons were
made between matching time points (*, P<0.05). A. Acute treatments of mice
receiving no chronic MeHg or isradipine (Group O). Acute MeHg treatment
(n=16) significantly increased fluorescence above no MeHg treatment (n=5). B.
Acute treatments of mice receiving chronic 6.25 ppm MeHg in water and no
isradipine (Group M). Acute MeHg treatment (n=16) significantly increased
fluorescence above no MeHg treatment (n=7). C. Acute treatments of mice
receiving no MeHg and 2 ppm isradipine in diet (Group I). Acute MeHg
treatment (n=13) increased fluorescence above no MeHg treatment (n=5), but
only at one intermediate time point. D. Acute treatments of mice receiving both
6.25 ppm MeHg in water and 2 ppm isradipine in diet (Group M/I). Acute MeHg
treatment (n=17) did not significantly increase fluorescence above no MeHg
treatment (n=6) during observation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139

Figure 4.4

Mean relative fura-2 ratio values for acute 25 μM AMPA, 50 μM NMDA + 10
μM glycine and 40 mM KCl applications measured from high-speed (0.5 Hz)
time course image series in cerebellar slices from 9 month old mice. Mice were
treated chronically for 6 months with MeHg or isradipine. Separate lines represent
different chronic treatments, including no MeHg and no isradipine (Group O,
black line, n=13, representative +SEM shown), MeHg and no isradipine (Group
M, red line, n=11), isradipine and no MeHg (Group I, blue line, n=11), and both
MeHg and isradipine (Group M/I, magenta line, n=11). Treatments are in order,
and consisted of treatment for 2 min, with approximately 20 s to reach the bath,
imaging for an additional 8 min and a 2 min wash between imaging. A. AMPA
treatment of slices, all four chronic treatment groups responded to the treatment
(*, P<0.05 compared to baseline). Several stayed above baseline ratio, while
group M/I returned during the imaging period. B. NMDA treatment of slices,
following AMPA. No population-wide change in fluorescence occurred. C. KCl
treatment, following NMDA. Only group O showed a significant, populationwide change in fluorescence ratio during treatment. . . . . . . . . . . . . . . . . . . . . . 144

Figure 4.5

Mean relative fura-2 ratio values for acute 25 μM AMPA, 50 μM NMDA + 10
μM glycine and 40 mM KCl applications measured from high-speed (0.5 Hz)
xvi

time course image series in cerebellar slices from 15 month old mice. Mice were
treated chronically for 12 months with MeHg or isradipine. Separate lines
represent different chronic treatments, including no MeHg and no isradipine
(Group O, black line, n=15 representative +SEM shown), MeHg and no isradipine
(Group M, red line, n=17), isradipine and no MeHg (Group I, blue line, n=14),
and both MeHg and isradipine (Group M/I, magenta line, n=15). Treatments are
in order, and consisted of treatment for 2 min, with approximately 20 s to reach
the bath, imaging for an additional 8 min and a 2 min wash between imaging. A.
AMPA treatment of slices, all four chronic treatment groups responded to the
treatment (*, P<0.05 compared to baseline). All stayed above the baseline ratio
for the duration of the imaging period. B. NMDA treatment of slices, following
AMPA. No population-wide change in fluorescence occurred. C. KCl treatment,
following NMDA. All four chronic treatment groups showed significant
responses to treatment, and all were temporary. . . . . . . . . . . . . . . . . . . . . . . . . 146
Figure 4.6

Mean relative fura-2 ratio values from acute ACSF control, 100 μM muscimol
and 10 μM bicuculline applications as measured from high-speed (0.5 Hz) time
course image series in cerebellar slices from 15 month old mice. Mice were
treated chronically for 12 months with MeHg or isradipine. Separate lines
represent different chronic treatments, including no MeHg and no isradipine
(Group O, black line, n=5, representative +SEM shown), MeHg and no isradipine
(Group M, red line, n=6), isradipine and no MeHg (Group I, blue line, n=4), and
both MeHg and isradipine (Group M/I, magenta line, n=6). Treatments are in
order, and consisted of treatment for 2 min, with approximately 20 s to reach the
bath, imaging for an additional 8 min and a 2 min wash between imaging. A.
ACSF control treatment (no change from pretreatment) of slices. Both chronically
MeHg-treated groups (M and M/I) showed significant responses during this
treatment (*, P<0.05 compared to baseline). B. Muscimol treatment of slices.
Both chronically MeHg-treated groups showed significant increases in
fluorescence, including the largest increase of any treatment in Group M. C.
Bicuculline treatment of slices, showing no significant changes from
baseline. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152

Figure 5.1

Example of two spontaneous events and KCl depolarization from a young (+/+)
mouse cerebellar slice loaded with fluo4-AM. Images are registered, cropped and
contrast enhanced for clarity. Images were captured at 60x magnification; the
scale bar equals 100 μm. Two nearly simultaneous events in the IGL are depicted
in the first two images. The first image is just before the events and the second at
the peak intensity. Arrows indicate the two cells showing the events. This slice
was later treated with high [KCl] ACSF for the purposes of depolarizing the slice.
Peak depolarization is depicted in the third image. The two cells that displayed
spontaneous transients were depolarized by this treatment, as were most cells in
the slice. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167

Figure 5.2

Representative fluo4 intensity traces from cells in three groups. Each trace is
normalized to the mean intensity value of the individual trace, so units are
xvii

relative. The first group is traces from young (PND 4-14) GABAAR α6 (+/+)
mice while the second is traces from young (-/-) mice. The third set includes
traces from older mice (22 months) from a separate line, so they are considered
(+/+) for the purposes of these results. Some qualities of the traces can be seen
from these examples. Note that several spontaneous events are repeated in young
animals and the fluo4 events in aged animals were much more protracted.. . . 168
Figure 5.3

Slice characteristics from total cell and active cell regions of interest (ROI).
Parameters from the automated event counter included cell counts and unique
event ROIs, allowing a count of total cells as well as active cells. Because of the
differences between layers in young (Y) and aged (A) mice, these counts are not
divided into layers or corrected to calculate cell density as has been performed in
previous experiments in this dissertation. No significant differences were found in
comparisons of total cell count. Young α6 (+/+) and (-/-) slices showed 13-33%
active cells while aged slices only showed about 4% compared to total fluorescent
cells, which is significantly fewer (P<0.05). . . . . . . . . . . . . . . . . . . . . . . . . . . . 170

Figure 5.4

Mean fluo4 intensity of events within frames of interest in slices from α6 (+/+)
and (-/-) genotypes and young and old age groups. Event parameters were
calculated relative to the mean value of an ROI over all time points, so values are
relative. Even though it appeared that (+/+) A had lower overall intensity, there
were no significant differences. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171

Figure 5.5

Cumulative distribution of event durations in slices from α6 (+/+) and (-/-)
genotypes and young and old age groups. Events were distributed into 1 sec bins
and plotted cumulatively up to 100% and out to the longest duration of 174 sec.
Because of the relative differences in number of total events, young mice have a
smoother distribution. Statistics were run comparing the largest % difference
between distributions and comparing to the critical difference (D) calculated from
event counts and standard deviation (σ). Slices from young (-/-) mice had
significantly more short duration events than (+/+) mice, while slices from aged
(+/+) mice had significantly more long duration events. . . . . . . . . . . . . . . . . . . 172

Figure 5.6

Mean frequency of events in slices from α6 (+/+) and (-/-) genotypes and young
and old age groups. Frequency was calculated from the total events over total
time, with some total times reduced to just pretreatment time to focus on the
spontaneous events. All slices had at least one significant event. Events occur
about every 20 s in young (+/+) slices, about every 4 s in young (-/-) slices and
about every 150 s in (+/+) aged slices. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174

Figure 6.1

Cartoon depiction of GABAAR and MeHg interactions with [Ca ]i as
highlighted by results of this dissertation. MeHg acts through GABAARs in part
2+
to increase [Ca ]i in all CGCs and all developmental stages. Both muscimol and
2+
bicuculline reduce MeHg effects on [Ca ]i through GABAARs, apparently by
desensitization in the case of muscimol and direct inhibition by bicuculline. Loss
of α6 subunits is disruptive to developing granule cells, possibly through
increased neurotransmission associated with mature CGCs with local synapses or

2+

xviii

2+

possibly through alterations in GABAAR expression. [Ca ]i-dependent cell
death, through enzyme activation, rundown of ATP and aberrant nuclear
signaling, is present, but not significant compared to regular handling. Chronic
MeHg appears to alter responses to muscimol, indicating a role for changes in
GABAAR surface expression, such as through internalization, as well as possible
changes in gene expression of GABAAR subunits (dashed arrows). . . . . . . . . 182

xix

KEY TO SYMBOLS AND ABBREVIATIONS

5-HT3 - 5-hydroxytryptamine (serotonin)
ACSF - artificial cerebrospinal fluid
ALS - Amyotrophic lateral sclerosis
AMPA - 2-amino-3-hydroxy-5-methyl-5-isoxazole propionic acid
ANOVA - analysis of variance
ATP - adenosine 5'-triphosphate
BDNF - brain-derived neurotrophic factor
BZ - benzodiazepine
2+

[Ca ]e - extracellular calcium concentration
2+

[Ca ]i - intracellular calcium concentration
2+

[Ca ]m - mitochondrial calcium concentration
Ca

2+

- calcium ion

CaM - calmodulin
CaMK - calmodulin kinase
cAMP - cyclic adenosine monophosphate
CF - climbing fiber
CGC - cerebellar granule cell
-

Cl - chloride ion
CNQX - 6-cyano-7-nitroquinoxaline-2,3-dione
CNS - central nervous system
dNTP - deoxynucleotide triphosphate
DMSO - dimethyl sulfoxide
xx

EDTA - ethylene-diamine-tetra-acetic acid
EGL - external granule (or germinal) layer
EPA - Environmental Protection Agency
ER - endoplasmic reticulum
erk - extracellular related kinase
eth-D1 - ethidium homodimer
Fluo 4 NW - fluo 4 no wash
Fura-2 AM - fura-2 acetoxymethylester
GABA - γ-aminobutyric acid
GABAA or GABAAR - GABA type A receptor
GAD - glutamic acid decarboxylase
GAT - GABA transporter
GCaMP - Green fluorescent protein-CaM protein
h - hour(s)
Hg

2+

- mercury ion
-1

Hz - hertz (s )
IGF-1 - insulin-like growth factor 1
IGL - internal granule layer
IP3 - inositol-1,4,5-triphosphate
iPSP - inhibitory post-synaptic potential
+

K - potassium ion
KCC2 - potassium-chloride co-transporter
LTD - long term depression
LTP - long term potentiation
xxi

M3 - muscarinic receptor type 3
mapk - mitogen associated protein kinase
MEK1 - mapk/erk kinase 1
MeHg - methylmercury
min - minute(s)
ML - molecular layer
ml - milliliter
mo - month(s)
mRNA - messenger RNA
μ - micro
mPTP - mitochondrial transition pore
nACHr - nicotinic acetylcholine receptor
NCX - sodium-calcium exchange transporter
NKCC1 - sodium-potassium-chloride co-transporter 1
NMDA - N-methyl-D-aspartate
NSF - n-ethylmaleimide sensitive factor
PC - Purkinje cell
PCB - poly-chlorinated biphenyl
PCR - polymerase chain reaction
PCMB - 4-chloromercuribenzoic acid
PF - parallel fiber
PKC - protein kinase C
PLC - phospholipase C
PMA - phenylmercury acetate
xxii

PMCA - plasma membrane Ca

2+

ATPase

PND - post-natal day
RfD - reference dose
ROI - region of interest
ROS - reactive oxygen species
SEM - standard error of the mean
2+

SERCA - smooth endoplasmic reticulum Ca

ATPase

sIPSC - spontaneous inhibitory post-synaptic current
SNARE - soluble NSF attachment receptor
TPEN - N,N,N’,N’-tertrakis-(2-pyridylmethyl)ethylediamine
VGCC - voltage-gated Ca
2+

Zn

2+

channel

- zinc ion

xxiii

CHAPTER ONE

INTRODUCTION

1

A) Background
a. Methylmercury (MeHg)
i. MeHg in the environment
+

Methylmercury (MeHg , henceforth abbreviated MeHg) is an organic heavy metal ion
formed from the addition of a single methyl side group to inorganic elemental mercury (Hg or
2+

Hg ). Inorganic mercury in the environment can be the result of many natural and
anthropogenic sources, including volcanic eruptions, erosion, mining, coal and oil burning, and
is typically deposited in bodies of water through precipitation in rain and runoff over land. In
water, inorganic mercury is toxic and irritable to many single-celled organisms, and is converted
to its methylated form. While this MeHg form is less toxic to some organisms, it is more
lipophilic and more difficult for multicellular organisms to excrete, leading to bioaccumulation
up the marine food chain (US EPA, 1997). Fish that consume microorganisms and plankton are
least likely to accumulate MeHg, but predatory fish such as tuna and swordfish are likely
accumulators. Even further, marine mammals at the top of the food chain have the largest
concentrations of all, and may have additional accumulated toxicants, such as dioxins and
polychlorinated biphenyls (PCBs) accumulating in their tissues. Consumption of these mammals
such as whales and seals, is not usually considered because of its rarity, but is important for
some populations (Debes, et al., 2006).
A few other forms of organic mercury exist, mostly compounds produced by advanced
chemistry techniques. These include ethylmercury, which has been an ingredient used in dental
amalgam, and dimethylmercury, a more lipohilic and toxic compound than MeHg that is not
considered an environmental contaminant. There are also several compounds of mercury that
display some of the characteristic chemistry and toxicity of mercury, including 4chloromercuribenzoic acid (PCMB), a protease inhibitor that reacts with protein thiol groups,
2

and phenylmercury acetate (PMA), a discontinued fungicide, disinfectant and paint preservative.
While these toxicants have their own profiles and are of interest in the study of mercury toxicity,
they will only be noted when necessary.
ii. MeHg in human health
Because of the bioaccumulation of MeHg in the marine food chain, it is primarily a
concern of low doses through seafood consumption, but it is necessary to note other exposures
and their importance to understand MeHg exposure and toxicity. A handful of medical reports
indicate that MeHg exposure through occupational handling of mercury compounds and possible
aerosols trace the medical characterization of mercury poisoning to the early half of the 19

th

century (Hunter and Russell, 1954, Magos and Clarkson, 2006). Though the exact mixture of
metals in these cases was not always clear, neurotoxicity was noted alongside liver and renal
failure. In particular, imprecise motor control and gait disturbance, termed ataxia, was a common
sign of neurological damage. Ataxia was correlated with degeneration of the cerebellum, as
determined by postmortem histology as well as early animal experiments. Additional
characteristics associated with MeHg poisoning included paresthesia (peripheral sensory loss),
disarthria (slurred speech), disphagia (difficulty swallowing), constricted vision, and tremors. In
human and mammalian exposures, many symptoms can take several months to manifest,
indicating a latency for MeHg toxicity. Some MeHg is excreted through the hair and is
dependent on concentration, so suspected cases of MeHg poisoning can be extrapolated
backwards to the highest concentration, and with sufficient hair a history of exposure can be
constructed.
The first major mass poisoning traced to MeHg occurred in the 1950s in a region of
Japan around the Minamata Bay (Takeuchi, et al., 1962). Hundreds of people in the area began
to show a severe neurological disorder characterized by motor dysfunction, sensory and visual
3

losses. This disorder was termed Minamata Disease. Through comparisons with occupational
MeHg exposures, it was hypothesized that MeHg was the cause, later confirmed through blood
and hair analysis. The source was eventually traced to industrial waste from a local chemical
company dumping MeHg into the bay, where the toxicant had bioaccumulated in the local fish
stocks over several years. As with occupational exposures, the Minamata exposure caused
preferential loss of cerebellar granule cells (CGCs) (Takeuchi, 1968, Tokuomi, et al., 1982).
Visual degeneration was attributed to damage in the occipital lobe, and peripheral sensory nerve
damage was also noted. This mass exposure was also of note because it was also the first event
that highlighted one of the hallmarks of MeHg poisoning, developmental toxicity. Fetal
Minamata disease, a cerebral palsy-like disorder, was found not long after the adult disorder was
characterized (Harada 1978). It was later determined to be a result of MeHg absorbed and
concentrated in the placenta, a mechanism that was unheard of at the time, but has been shown to
be an attribute of several lipophilic and bioaccumulative toxicants since. No abnormalities in the
exposed mothers of the infants were noted, and there was a delay of about 6 months between
birth and notable abnormalities such as loss of muscle tone, convulsions, mental retardation,
hyperactivity, disarthria, and slowed growth. In 15 years of followup, some symptoms recovered,
particularly in mild cases, but intelligence was not improved and many of the more severelyaffected children died within that time. Many umbilical cords were kept from the Minamata
exposure and analyzed for MeHg concentration when techniques were developed (Akagi, et al.,
1998). This analysis indicated that cord blood in affected infants contained about 216 µg/L, and
it was then estimated that women pregnant with affected children consumed MeHg at about 225
µg/day.
A second mass exposure event occurred in Iraq in the 1970s through a much different
source (Bakir, et al., 1973). Relief shipments of grain sent to the country after a crisis contained
4

seed grain coated in MeHg and ethylmercury, common fungicides at the time. The grain bags
warned about proper handling and planting of the seeds, as the fungicides were meant to wash
off over germination. Through various errors, including delays that brought the seed in after
planting season and siphoning of the aid onto the black market, warnings were not read and the
seed was frequently used in flour instead. Concentrations of the fungicide in this food were much
higher than in the exposures in Minamata, so toxicity was more acute and more widespread
around the country. Through monitoring of the aftermath, it was extrapolated that severe cases of
the exposure included as much as 5 ppm (20 μM) MeHg in blood. As with the Minamata event,
exposed pregnant women with few to no symptoms frequently gave birth to severely affected
children (Amin-Zaki, et al., 1974).
While there has not been any major exposure event of the size or severity of Minamata or
Iraq since, MeHg is still an area of concern, particularly in fish-consuming populations around
the world. Also, due to the developmental susceptibility noted in exposures, special attention has
been paid to neurological function in infants, children and adolescents (Castoldi, et al., 2008).
Two major ongoing studies have monitored MeHg exposure and low dose effects for several
decades. In the Seychelles Islands, located north of Madagascar in the Indian Ocean, locals are
almost entirely dependent on fish diets, and almost the entire population receives regular health
checks that include monitoring MeHg (Davidson, et al., 2008). In the Faroe Islands, located
between Scotland and Iceland in the Arctic Ocean, fish diets are supplemented by marine
mammals which may also accumulate persistent organic pollutants such as dioxin and PCBs
(Debes, et al., 2006). The Seychelles study has found few neurological differences in its cohorts,
possibly due to the beneficial nutrients of fish protein. The Faroe study has found significant
detrimental neurological effects from the combination of pollutants in whale meat, enough to
repeatedly call for the end of whale consumption.
5

As with many environmental toxicants, regulators have formulated a reference dose
(RfD) of 4-5 μg/L in blood and approximately 1 μg/g in hair for MeHg through integration of the
epidemiological research (US EPA, 1997). The RfD is approximately 1/10th of the concentration
found to cause minor developmental delays in the Iraq exposure. Blood and hair measures of
MeHg, which cannot distinguish between types of Hg exposure, are typically used. Blood
concentrations are best for determining MeHg concentration immediately, but hair concentration
offers several advantages. MeHg is concentrated in hair when excreted and hair grows at a
constant rate, so past MeHg exposure can be tracked. Because of the ease of hair monitoring,
several smaller studies have included monitoring indigenous populations in North America, in
the general population of Europe, and miners and indigenous populations in the Amazon that
may be exposed to fumes and runoff from gold mining operations. While more specific
correlation of brain and blood biomarkers for MeHg exposure and neurotoxicity has been shown
in laboratory animals, biomarkers such as muscarinic receptor activity and monoamine oxidase
activity have shown mixed effectiveness in predicting outcomes in humans (Roda, et al., 2012)
so direct measurements remain the standard. The data at the time of the EPA report indicated that
about 10% of the population exceeded the RfD regularly, with as much as 3 fold excess in
regular fish consumers. Because of a relative lack of specific cases involving children and
pregnant women, the RfD is based on the general population. Recent studies of seafood
consumption indicate that in European females of child-bearing age, about 11% of consumers
are still getting levels of MeHg above the RfD while 44% are not getting the recommended
amounts of beneficial fatty acids associated with fish consumption (Strom, et al., 2011). As
many populations of fish do not contain an adequate balance of MeHg versus beneficial
nutrients, future recommendations for fish consumption may need to take this into account.

6

iii. MeHg chemistry and mechanisms
While MeHg and Hg

2+

do not appear much different chemically as ions, the methyl

group addition produces an ion capable of making much longer and weaker bonds with targets
(Tai and Lim, 2006). This results in a number of important chemical differences. First, MeHg
has high affinity to bind sulfhydryl groups, including cysteine (Cys) and methionine (Met)
groups on many proteins. Second, this chemical bond is weak and labile, meaning that though
the ion is bulky, interactions are likely to be temporary. Together this binding pattern makes
MeHg likely to bind and interact with many proteins, and it makes MeHg a particularly difficult
toxicant to characterize. Though research on MeHg has been intense for many decades, the exact
suite of interactions and mechanisms by which it causes toxicity still have not been determined.
Interactions with sulfhydryl groups appear to be critical for the comparative efficiency of
MeHg to pass the blood-brain barrier and placenta, because when bound to L-Cys or L-Met,
MeHg is able to pass through amino acid transporters easily (Aschner and Aschner, 1990,
Kerper, et al., 1992). Unlike MeHg, Hg

2+

is primarily toxic to liver and kidneys, in part due to

inefficient passage through the blood-brain barrier. Despite this major difference, MeHg is still
2+

ten-fold more toxic to cultured CGCs than Hg , meaning that additional differences in entering
cells or interactions with cell signaling pathways must exist (Gasso, et al., 2001). Interaction
with sulfhydryl groups also makes MeHg an easy target for glutathione-based excretion, but its
lability may allow it to escape this cleanup, expecially once outside the blood stream (Choi, et al,
1996).
While the passage of MeHg through the blood-brain barrier allows it to accumulate in the
central nervous system (CNS), MeHg also accumulates in cells. Though MeHg is charged, there
is ample evidence that it is able to enter cells, where it is able to act directly on mitochondria and
cytosolic signaling cascades (Atchison and Hare, 1994, Limke and Atchison, 2002). Proteomic
7

analysis of CGCs indicates that MeHg may be able to alter expression of mitochondrial proteins
and cytoskeletal proteins critical for migration (Vendrell, et al., 2007, Vendrell, et al., 2010). As
with other metals, some of this permeability may come from the simply entering through the
pores of ionotropic receptor channels, such as voltage-gated calcium channels (VGCCs)
2+

(Atchison, 1987). However, altering the concentration of extracellular calcium (Ca ) or using
ions other than Ca

2+

in order to compete with MeHg ions does not alter the effects of the

toxicant, so other mechanisms of MeHg cell permeability must exist (Traxinger and Atchison,
1987). Just as it passed through the blood-brain barrier, MeHg may be able to pass into cells
while bound to amino acids or other transported molecules, or, if incorporated into a small
enough non-polar molecule, MeHg could pass directly through the membrane. Once inside,
MeHg could interact with many signaling cascades, and could only be removed by scavengers
such as glutathione or n-acetylcysteine (Aremu, et al., 2008). Despite evidence that MeHg enters
cells and may even accumulate, many of the most critical actions of MeHg appear to happen at
the surface of neurons.
Once inside the CNS, MeHg is able and likely to encounter many cell surface receptors,
and due to its lability, it can interact with available sulfhydryl groups. In many cases, MeHg is
able to alter both activity of a wide array of neuronal receptors directly and expression of these
receptors through signaling cascades. Because it is present both inside and outside the cell,
MeHg is able to interact with many more sites on these receptors than an extracellular toxicant.
The interaction of MeHg and neuronal receptors is a major area of interest of the Atchison
laboratory. In particular, a number of receptors that are able to alter the intracellular
concentration of Ca

2+

2+

([Ca ]i) are known to be targets of MeHg (Figure 1.1). MeHg alters the

activity of some of these, including VGCCs (Marty and Atchison, 1997, Sirois and Atchison,
2000, Hajela, et al., 2003), and the muscarinic M3 receptor, a metabotropic receptor that
8

regulates Ca
gated Ca

2+

release from internal stores (Limke, et al., 2003, Limke, et al., 2004). Ligand-

2+-

permeable channels, such as N-methyl-D-aspartate (NMDA), 2-amino-3-hydroxy-5-

methyl-isoxazole propanoic acid (AMPA) and kainate receptors, may also be affected by MeHg,
either directly altering their activity or by changing their expression (Juarez, et al., 2005, Basu, et
al., 2007, Johnson, et al., 2011). These are by no means the only receptors or mechanisms by
which MeHg acts. However, the presence of many surface receptor targets of MeHg that are
2+

capable of altering [Ca ]i and its signaling suggest a collective mechanism, changes to Ca

2+

homeostasis, by which MeHg can act. As with many possible mechanisms of MeHg toxicity,
these interactions require further scrutiny to understand under what conditions they contribute or
are affected by MeHg toxicity.
One receptor stands out as needing further consideration. In electrophysiological study
elucidating the receptors in susceptible cells that respond to low concentrations of MeHg, the γaminobutyric acid type A GABAA receptor (GABAAR) was found to be a target (Yuan and
Atchison, 1997, Yuan and Atchison, 2003, Yuan and Atchison, 2007). Further study suggested
2+

that GABAARs could be involved with MeHg toxicity in a Ca -dependent manner (Herden, et
2+

al., 2008). GABAARs fit well with a [Ca ]i model for MeHg toxicity due to several different
factors. First, though they are generally considered inhibitory, GABAAR activation is still
frequently associated with Ca

2+

influx through secondary mechanisms, and activation may

actually be excitatory under some conditions (Owens, et al., 1996, Eilers, et al., 2001, Brodinsky,
et al., 2003). Second, GABAARs are heterogeneously constructed, with many different
combinations of subunits and responses to ligands (Brickley, et al., 1996, Saxena and
MacDonald, 1996). The expression of specific GABAARs frequently matches characteristics of
MeHg toxicity, such as cell and developmental susceptibility. Third, GABAARs are closely

9

2+

Figure 1.1. Schematic representation of neuronal MeHg targets and the central action of [Ca ]i in normal neuronal function and
toxicity.
10

Figure 1.1 (cont’d). This schematic represents an axon terminal, but many of these mechanisms
2+
are present throughout a neuron. MeHg acts on Ca channels (such as VGCCs), metabotropic
receptors that signal through IP3 (such as the M3 muscarinic receptor), and the GABAA receptor
at the surface of neurons (red arrows). MeHg is also able to enter cells and act on proteins such
as the mitochondrial transition pore (mPTP), but the exact mechanisms of entry are unclear
2+
(dashed red arrow). [Ca ]i is normally increased transiently through several mechanisms, such
2+
2+
as Ca -dependent Ca release from ER through ryanodine receptors, but MeHg increasec
2+
2+
[Ca ]i beyond normal levels (black arrows). The action by which MeHg can increase [Ca ]i
through GABAA receptors is unclear, but may be dependent on VGCCs through depolarization
2+
2+
or other mechanisms (dashed black line). Some Ca may be buffered by Ca binding proteins
2+
2+
in the cytosol, but excess Ca is normally taken up by the ER through Ca -ATPases such as
2+
the smooth endoplasmic reticulum Ca ATPase (SERCA), and may be expelled from the
cytoplasm by other ATP-dependent mechanisms, both of which require ATP from the
2+
mitochondria (blue arrows). The normal actions of increased [Ca ]i include regulating vesicle
release, neuronal motility during development, and regulation of gene expression, but the excess
2+
of Ca from MeHg actions can cause excess neurotransmitter release, aberrant gene expression,
and can activate several enzymes involved in apoptosis, such as lipases and calpains (magenta
arrows). For interpretation of the references to color in this and all other figures, the reader is
referred to the electronic version of this dissertation.

11

associated with other receptors and are critical for regulation of neuronal excitability and
neurotransmission (Long, et al., 2009). In whole brain, alterations in neurotransmission can be as
toxic for surrounding cells as direct toxicity. MeHg affects three unique components of CGC
+

neurotransmission, GABAARs, M3 muscarinic receptors, and a K leak current. All three
contribute to an increase in glutamate release (Atchison, 2005). MeHg also inhibits glutamate
uptake (Aschner, et al., 1990, Yin, et al., 2007), so with increased glutamate release and
decreased uptake, MeHg can be considered toxic through glutamate excitotoxicity. Through
other changes to neurotransmitter and modulator release, such as changes to brain-derived
neurotropic factor (BDNF), MeHg is capable of altering neuroprotection throughout the brain
(Onishchenko, et al., 2008).
iv. MeHg toxicity susceptibility and resistance
Despite widespread brain and nerve toxicity displayed in major exposure events, MeHg
shows a progressive neurotoxicity, with certain parts of the brain affected at lower
concentrations than others. The cerebellum is a target, as are dorsal root ganglia in the spinal
cord, and parts of the visual cortex, all of which degrade before regions such as the cerebral
hemispheres. This is most clearly shown in post-mortem histology of both humans and
laboratory animals (Hunter and Russell, 1954, Takeuchi, 1968, Sakamoto, et al., 1998). Because
it is one of the larger brain structures that can be targeted by even small concentrations of MeHg,
much of the toxicological characterization of MeHg has focused on the cerebellum. Within the
cerebellum, cell-specific toxicity is also seen. In both acute and chronic MeHg exposure, there is
a significant loss of CGCs while Purkinje cells (PCs), their synaptic targets, remain relatively
intact (Edwards, et al., 2005). CGCs and PCs have many cellular differences that may account
for the differences in their relative susceptibilities to MeHg toxicity. The most readily apparent
difference is in size, CGCs are some of the smallest neurons in the brain while PCs are among
12

the largest. This size difference leads to a difference in buffering capacity, wherein CGCs are
more likely to respond to a smaller ion flux than the larger PCs, and this difference is driven
further by differences in expression of binding proteins and other typical cytosolic buffers (Gall,
et al., 2005).
Many of the specific effects of MeHg are enhanced in developing animals. Even lowlevel MeHg exposure during pregnancy, ~1ppm as measured by maternal hair, subtly decreases
the size of a newborn’s cerebellum (Cace, et al., 2011). As suggested by the historical record of a
latency period before overt toxicity in human infants, this latency period is reciprocated in
animal studies of perinatal exposure (Sakamoto, et al., 1998). Toxicity in animals includes many
of the same disorders and degeneration as seen in human exposures, including degeneration of
the cerebellum, neocortex, striatum, brainstem and dorsal root ganglia, leading to hind limb
paralysis and unsteady gait. Overt toxicity in this type of application is enhanced by lack of the
critical metabolism and excretion protein metallothionein (Yoshida, et al., 2008). As with many
toxicants, there is evidence that damage to liver and kidney occurs during MeHg exposure in
young animals in addition to its effects on the CNS (Abdalla, et al., 2012). While most of the
research focusing on the developmental effects of MeHg has focused on the time corresponding
to the later susceptible periods of gestation in humans, MeHg may have other effects in early
neuronal development (Xu, et al., 2010, Engel, et al., 2012).
The effects of MeHg on aged animals have not been as widely studied as developmental
effects, but it is important to consider elderly susceptibility and the possibility that MeHg
enhances neuronal degeneration (Weiss, 1990, Weiss, et al., 2002, Weiss, 2010). Enhanced
effects of MeHg might come about by either direct effects due to declining protective
mechanisms during aging or, in the case of developmental or lifelong exposure, prevention of the
buildup of neurons and connections in the brain, leading to a smaller safe threshold above
13

neurodegenerative disorders. As an example of a direct effect that could enhance age-related
decline, MeHg treatment decreases dopamine receptor density in hippocampal neurons (Coccini,
et al., 2011). Dopamine signaling is important for motor control in the hippocampus and
normally declines in aging, and this additional alteration by MeHg could enhance the risk of
neurological disorders associated with dopamine, such as Parkinson’s Disease. MeHg exposure
also accelerates motor decline in a mouse model of familial Amyotrophic Lateral Sclerosis
(ALS), suggesting that genetic susceptibilities for decline can be exacerbated by the toxicant
(Johnson, et al., 2009). One study suggests that acute MeHg exposure does not enhance the
typically monitored motor effects in normal populations of aged mice compared to younger adult
exposure (Bellum, et al., 2012). But in zebrafish, it appears that in older, already deteriorating
populations, acute exposure to MeHg is more toxic than exposure in young adults (Xu, et al.,
2012). Compared to acute exposures, it is clear that chronic or perinatal exposure to MeHg
causes motor deficits that are at least persistent into old age, and that some new deficits appear in
geriatric animals, including deficits in visual and spatial discrimination and increased weight
gain in geriatric rodents (Paletz, et al., 2007, Vitalone, et al., 2010).
Along with susceptibility, ability to resist MeHg toxicity may also elucidate mechanisms
responsible for toxicity. MeHg-tolerant organisms indicate the importance of metabolism and
clearance in protection from MeHg toxicity. C. elegans appears to show a resistance to the
effects of MeHg on neurons, which may be due in part to induced expression of glutathione, heat
shock proteins and metallothioneins, which can export or inactivate MeHg (Helmcke and
Aschner, 2010). This induction leads to hormesis, wherein a small dose of MeHg applied induces
a strong resistance to further applications. Several generations of flies can be selected to produce
MeHg tolerant strains by using MeHg-supplemented food. When compared to non-tolerant
populations with and without MeHg-containing food, several metabolic genes and one immune
14

gene were highlighted (Mahapatra, et al., 2010). Resistance appears in these studies to include
antioxidant and inflammatory regulatory pathways.
b. Calcium
i. Calcium signaling and regulation
Just as MeHg has broad effects in the CNS, one of the mechanisms by which it can act,
2+

changes to [Ca ]i and other Ca

2+

dynamics, is critical for a multitude of neuronal mechanisms.

Because of these wide-ranging interactions, Ca

2+

signaling will be related to other mechanisms

of MeHg toxicity, primarily control of electrochemical gradients, cell death and neurotransmitter
release, along with cell, developmental and aging susceptibilities that may be related.
2+

In neurons, [Ca ]i is tightly regulated, with concentrations usually in the 100-200 nM
range, while external Ca

2+

2+

([Ca ]e) in the 1-2 mM range, representing a 10,000 fold difference

and establishing a strong chemical gradient, even though most of the ion is bound (Kass and
2+

Orrenius, 1999). While cytosolic [Ca ]i is maintained at these low concentrations, neurons also
maintain stores of Ca

2+

in organelles, both in the endoplasmic reticulum (ER) and in

mitochondria. The ER maintains a dynamic store of Ca

2+

2+

at concentrations near [Ca ]e, with

2+

uptake controlled by the smooth ER Ca -ATPase pump (SERCA) and release regulated by
inositol-phosphate 3 (IP3) receptors. Typically mitochondria maintain concentrations of Ca

2+

2+

([Ca ]m) similar to the cytosol, but can rapidly take up the excess during stimulation, at about
0.5 µM/s (Rizzuto, et al., 1994). Ca

2+

levels in neurons and glia are also critical for metabolism,

regulating glycolytic activity and ATP metabolism at rest or in active states (Pancani, et al.,
2+

2011). This baseline configuration of cytosolic and organellar [Ca ] is considered homeostasis,
and any regulatory mechanism that maintains it is homeostatic. Many stimuli can induce changes
in one or more of these concentrations, however most are temporary. Influx of Ca

15

2+

comes

primarily from two sources, neurotransmitter activation of receptors and membrane
depolarization. Both neurotransmitter activation and membrane depolarization can propagate
2+

along the surface of neurons through IP3-dependent release from ER and Ca -dependent Ca
influx from membrane receptors (Gleichmann and Mattson, 2011). Ca

2+

2+

is directly and
2+

continually removed from cytoplasm by SERCA and the plasma membrane Ca -ATPase
2+

(PMCA), which both require ATP. Large concentrations of [Ca ]i may also be extruded from
+

neurons by the Na /Ca

2+

exchanger (NCX), which indirectly requires ATP to reestablish the

sodium gradient. Frequent activity and influx of Ca

2+

2+

or low ATP increases [Ca ]i, which is
2+

taken up by mitochondria and signals a need for increased ATP production. Increased [Ca ]m is
2+

a signal of dysfunction, both as it reflects a high [Ca ]i and as increased ATP production can
lead to increased production of reactive oxygen species (ROS).
As seen in the uptake of Ca

2+

by mitochondria, Ca

2+

can serve as both a signal for

mechanisms to attenuate stress, as with the increase in ATP production, and as a stressor,
2+

inducing ROS production. In a similar manner Ca

signals both cell survival and cell death

(Cerella, et al., 2010). If a neuronal insult, including toxicant exposure or excessive activation, is
2+

sufficient to induce rapid, strong increases of [Ca ]i, necrosis can occur. However, most
2+

stresses only induce apoptosis. Ca -dependent apoptosis is characterized by activation of
2+

several Ca -sensitive enzymes, including m-calpains, which break down many cytosolic
proteins (Chan and Mattson, 1999, Demarchi and Schneider, 2007), and lipoxygenases and
phospholipases, which break down lipids in the membrane and release inflammatory mediators
(Exton, 1990, Tang, et al., 2007). Neurons rely on a close coupling of ER and mitochondria to
2+

maintain low [Ca ]i (Csordás, et al., 1999, Shoshan-Barmatz, et al., 2004). While Ca

2+

is

normally taken up by ER through SERCA, if this is overwhelmed or ATP is depleted, neurons
2+

must rely on mitochondrial uptake (Spät, et al., 2008, Ruiz, et al., 2009). Excess [Ca ]i taken up
16

by mitochondria can depolarize and overwhelm the organelle, leading to opening of the
mitochondrial permeability transition pore (mPTP), which can release critical mitochondrial
proteins and degenerate the mitochondria (Dahlem, et al., 2004, Denlaud, et al., 2008,
Baumgartner, et al., 2009). Since ATP is already depleted, loss of mitochondria exacerbates the
energy failure.
ii. Calcium and neurotransmission
2+

All neurotransmitter release requires Ca
Ca

2+

signaling and current models place a variety of

channels in close proximity to vesicle pools (Chad and Eckert, 1984, Fogelson and Zucker,

1985, Westenbroek et al., 1995, Schneggenburger, et al., 2012). Typical neurotransmitter release
occurs when a depolarizing action potential reaches the presynaptic membrane. This
depolarization activates VGCCs in the membrane, inducing a large Ca
2+

al., 1985, Sabatini and Regehr, 1995,). This influx of Ca

2+

influx (Augustine, et

is able to mechanically trigger

neurotransmitter release directly (Stanley, 1993, Atlas, et al., 2001). This is because Ca

2+

influx

in close proximity to vesicles activates synaptotagmin I in vesicular soluble n-ethylmaleimide
sensitive fusion (NSF) attachment receptor (SNARE) complexes and causes rapid (µs-ms)
vesicle fusion and release (Geppert, et al., 1994, Xu, et al., 2007, Chapman, 2008).
Neurotransmitter release is regulated in part through differences in VGCC expression at
presynaptic and postsynaptic sites in different types of synapses (Jones and Heinemann, 1987,
Doroshenko, et al., 1997, Catterall and Few, 2008). These differences in expression are also
important for neuronal plasticity, the stored response level of a neuron based on previous
activity, including long-term depression (LTD) and long-term potentiation (LTP) (Pelkey, et al.,
2006, Wankerl, et al., 2010). Defects in Cav2.1, the major presynaptic VGCC in the CNS,
underlie familial migraine and two forms of ataxia (Ophoff, et al., 1996, Zhuchenko, et al.,
1997). These ataxias are characterized by excitotoxic loss of CGCs similar to losses seen in
17

MeHg poisoning, suggesting excitotoxic alterations to neurotransmission may be a major factor
in MeHg neurotoxicity.
iii. Calcium in development and aging
Ca

2+

signaling is critical for some early neuronal developmental steps (Leclerc, et al.,

2012), but is critical for neuronal migration and axon elongation (Zheng and Poo, 2012). Ca

2+

dependence for neuronal migration was first described in CGCs, and is the most characterized
(Komuro and Rakic, 1992, Komuro and Rakic, 1993, Komuro and Kamada, 2005). Ca
2+

of these cells indicates repeated, somatic increases and decreases of [Ca ]i, Ca

2+

2+

imaging

oscillations,

2+

with forward progress during high [Ca ]i and rest at low concentration. These transients are
mediated through NMDA receptors and VGCCs (Rossi and Slater, 1993, Komuro and
Yacubova, 2003), which in intact brain are activated by a depolarizing current through
GABAARs and ambient GABA (Brodinsky, et al., 2003). Ca

2+

signaling within the cells during

migration is dependent on several pathways, including phospholipase C (PLC), protein kinase C
(PKC), cyclic AMP (cAMP), and calmodulin kinases (CaMKI and CaMKII) (Kumada and
Komuro, 2004). Many of these signals produce locomotion through interactions with the
cytoskeleton (Rakic et al., 1994). When CGCs in organotypic slice culture are treated with
2+

caffeine or thimerosal, a mercury-containing compound, both increase neuronal [Ca ]i
continuously, but the effects are dependent on stage of migration, and both are able to reverse
2+

migration at some stages that normally have low [Ca ]i (Kumada and Kumoro, 2004). Most
inhibitors of Ca

2+

2+

oscillations, and many agents that simply disrupt [Ca ]i, including MeHg and

nicotine, decrease both Ca

2+

oscillations and migration, while modulators that increase spike

frequency, such as cAMP activation and insulin-like growth factor 1 (IGF-1) signaling, restore
migration, both in vitro and in vivo (Mancini, et al., 2009, Fahrion, et al., 2012). Directional cues
may also be encoded in Ca

2+

gradients across migrating neurons. Ethanol is able to interrupt
18

these cues, possibly through actions on GABAARs, leading to disruption in turning of CGCs
during migration and ectopic expression of the neurons (Kumada et al., 2010). Complete block
of Ca

2+

oscillations, as demonstrated in many of these studies, is a signal to neurons to cease

migration, both in normal migration and when agents are added, and mature neurons are rarely
2+

found with regular changes to [Ca ]i compared to migration.
Changes to Ca

2+

dynamics, have been hypothesized as a risk factor for many

neurodegenerative diseases, including Alzheimer’s and Parkinson’s, but are also a hallmark of
aging neurons (Landfield and Pitler, 1984, Toescu and Vreugdenhil, 2010). Aged neurons
display increases in VGCCs, increased Ca

2+

release from stores, and changes to Ca

2+

buffering

(Thibault and Landsfield, 1996, Ouanounou, et al., 1999, Nicholls, 2004, Gant, et al., 2006).
2+

Several types of neuron also have significant increases in baseline [Ca ]i in aged brains
compared to younger brains, including CGCs and some hippocampal neurons (Thibault, et al.,
2001, Xiong, et al., 2002). Perhaps more importantly, some presynaptic terminals in aged
2+

animals have increased [Ca ]i, indicating a mechanism for increased neurotransmitter release
2+

(Tonkikh, et al., 2006, Tonkikh, et al., 2009). As [Ca ]i is tied to cell metabolism and ATP
production, ATP reserves are typically lower in aging neurons (Mattson, et al., 2009). Even a
2+

slight increase in baseline [Ca ]i attributable to basic variations and polymorphisms could
contribute to declines by requiring more ATP use to recover from each event (Stutzmann, 2007).
2+

Because of this decreased ATP pool and decreased buffering of [Ca ]i, aging brains have been
hypothesized to be chronically depolarized, leading to degeneration of synapses and nerve
terminals, and contributing to age- and disease-related declines in function (Marczynski, 1998).
2+

While [Ca ]i increases in normal brain aging contribute to some of these declines in synaptic
function, remodeling and plasticity, actual cell death and neurodegeneration requires additional
2+

insults, such as genetic disorders, excessive stimulation or injury, that increase [Ca ]i further
19

2+

(Brewer, et al., 2007, Sun, et al., 2008). Further increases of baseline [Ca ]i in these models
lead to activation of many of the lipases, kinases and phosphatases associated with cell death.
iv. MeHg and calcium
MeHg directly modulates several neuronal membrane receptors and channels, including
VGCCs (Sirois and Atchison, 1996). MeHg also activates the M3 muscarinic receptor, a
metabotropic receptor that, via IP3 signaling, is able to rapidly release ER Ca

2+

stores (Hare and

Atchison, 1995, Limke, et al., 2004). Acute MeHg exposure can also open the mPTP, releasing
additional stores (Limke and Atchison, 2002).Through these mechanisms, MeHg induces
2+

sustained and irreversible dysregulation of [Ca ]i. This is seen in Ca
different types of cells, with an initial phase of Ca

2+

2+

imaging of many

entry into the cytosol from internal stores,

followed by a second phase of entry from extracellular sources (Hare, et al., 1993, Marty and
Atchison, 1998, Edwards, et al., 2005). The second phase can be blocked by extracellular Ca

2+

chelation, but the first phase is irreversible (Hare and Atchison, 1995, Marty and Atchison,
2+

1997). In separately cultured CGCs and PCs, MeHg causes increases in [Ca ]i in both cell
types, but the effect was delayed in PCs compared to CGCs at the same concentrations. In
2+

cerebellar slices from rats, MeHg increases [Ca ]i in CGCs as well as diffusely in many intact
processes, while PCs in slice do not respond visibly (Yuan and Atchison, 2007). This increase in
2+

[Ca ]i corresponded with increased spontaneous vesicular release.
A surge in neurotransmitter release is another mechanism characteristic of acute MeHg
toxicity in neurons and synaptosomes (Atchison, 1986, Minnema, et al., 1989, Kalisch and Racz,
1996, ). Mitochondrial uptake of Ca

2+

as well as Ca

2+

blockers are able to reduce this surge

(Levesque and Atchison, 1988, Levesque, et al., 1992, Denny and Atchison, 1996). Since CGCs
2+

are the most susceptible neurons to dysregulation of [Ca ]i by MeHg and also release glutamate
as a neurotransmitter, it is possible they contribute to enhanced glutamate in the cerebellum.
20

Glutamate is toxic to CGCs through NMDA receptor activation, which feeds back into increased
2+

[Ca ]i (Eimerl and Schramm, 1991). MeHg inhibits glutamate uptake in astrocytes, the main
scavenger of the neurotransmitter, further enhancing the concentration of this excitatory
neurotransmitter (Aschner, et al., 1990, Moretto, et al., 2005). MeHg also enhances release of
amino acid signaling molecules from astrocytes, potentially contributing to other toxic signaling
cascades throughout the brain (Aschner, et al.,1995).
2+

Irreversibly high increases in [Ca ]i can occur after many toxic insults, including
+

glutamate, MeHg and repeated high K depolarizations, and these increases precede other cell
death markers, such as cell membrane permeation (Tymianski, et al., 1993, Ding, et al., 2011).
2+

These [Ca ]i increases are capable of inducing acute necrosis of some neurons immediately, but
it is more notable as inducing delayed apoptosis in many more exposed neurons (Castoldi, et al.,
2+

2000). Apoptosis is typically [Ca ]i-dependent, and Ca
MeHg neurotoxicity by blocking some Ca

2+

2+

blockers are able to protect against

influx in both in vitro and in vivo models

(Sakamoto, et al., 1996, Edwards, et al., 2005). While protective, these blockers are only
temporarily effective, if MeHg is not removed, block only delays toxicity.
2+

Zinc ions (Zn ), may be an additional factor in MeHg toxicity. Zn
MeHg treatment, as indicated by changes in the Ca

2+

2+

is released by

insensitive wavelength of the Ca

2+

fluorophore Fura-2, and from chelating released divalent cations (Denny, et al., 1993, Denny and
Atchison, 1994). Several possible sources of Zn

2+

release from MeHg exposure exist, including

release from proteins due to MeHg interference with thiol groups (Tao and Lim, 2006, Gibon, et
al., 2010). For example, scaffolding proteins Shank2 and Shank3 are involved in the formation
of postsynaptic densities (PSDs). Removal or release of Zn

2+

degenerates PSDs (Grabrucker, et

al., 2011). Though a direct effect has not been characterized, MeHg degenerates similar PSDs
2+

(Castoldi, et al., 2000). Once released, intracellular Zn
21

is itself toxic to neurons, in part through

disruption of mitochondrial ATP production and possibly by contributing to the depolarization of
mitochondria and opening of the mPTP (Manev, et al., 1997, Sensi, et al., 2000, Jiang, et al.,
2001, Isaev, et al., 2012). Extracellular Zn

2+

may be important as well. Much of the Zn

2+

in the

brain is found in vesicles, but compared to the rest of the brain, it is unclear if this is an
important pool available in the cerebellum (Wall, 2005). Golgi cell interneurons are the only cell
in the cerebellum with the appropriate transporters to load Zn
2+

2002). Golgi synaptic release of Zn

2+

into vesicles (Wang, et al.,

may modulate CGCs through inhibition of NMDA and

GABAARs (Schmid, et al., 1999, Erreger and Traynelis, 2008, Amico-Ruvio, et al., 2011).
2+

Extracellular Zn

may actually be protective against MeHg, by inducing signaling cascades in

astrocytes that produce metallothioneins, a scavenger of MeHg (Aschner, et al., 1998).
c. Cerebellum
i. Cerebellar circuitry
The cerebellum has relatively few different types of cells and a fairly uniform structure.
The basic circuit of the cerebellum was famously described in the 1960s (Eccles, et al., 1967). In
the cerebellar cortex, the only two typical neurons are the small, numerous CGCs and the large
PCs, which are known for their large and heavily branched dendrites. While the cerebellar circuit
is considered a simple circuit by many, complex combinations of spike activity and silent
intervals suggest that more complicated activity may be prevalent in cerebellar learning (De
Zeeuw, et al., 2011). CGC axon to PC dendrite synapses, which are glutamatergic, excitatory
2+

synapses, are important in procedural skills and memory, particularly through the Ca dependent synaptic plasticity associated with LTP and LTD. Mutations regulating both are
detrimental to skill acquisition (Ito, 2008, Schonewille, et al., 2010). In addition to parallel
fibers, climbing fibers enter the cerebellum from neuronal circuits in the spinal cord, sensory and
motor cortexes, vestibular system (balance), red nucleus (motor coordination), and superior
22

colliculus (optical and sensory coordination). These climbing fibers form glutamatergic synapses
and act on PC dendrites as well. CGCs receive excitatory glutamatergic synapses from mossy
fibers originating from outside the cerebellum, while also receiving feedback GABAergic
inhibition from Golgi cell interneurons, circuits that have been characterized much further from
the original hypotheses (Ito, 2006). Additional internal circuits, such as regulation by Basket and
Stellate interneurons, Lugaro cell regulation of Golgi cells, and Unipolar Brush cell synapses
with CGCs in some regions of the cerebellum can also regulate cerebellar circuitry, but climbing
fibers and mossy fibers remain the major inputs to the cerebellum, while PC axons are the only
output. Both parallel fibers and climbing fibers act on many different PCs, parallel fibers synapse
on scores of PCs in parallel, while climbing fibers may branch to about 10 PCs with many
synapses on each. These multiple connections between populations of CGCs and PCs lead to
groups of the neurons acting in concert, so even though PC axons are the only output of the
cerebellum, the concerted output to the deep cerebellar nuclei and the system that provided the
input is strong.
ii. Cerebellum growth and neuron development
Cerebellar development starts in the neural tube of the embryo, with the definition of
midbrain and hindbrain, and proceeds with the formation of the rhombic lip and roof plate,
where accumulating neuronal precursors migrate and build the structure of the cerebellar cortex
(Hashimoto and Hibi, 2012). Migrating PCs radiate into the developing cerebellum and form a
thick layer near the surface of the developing cerebellum (the pial surface). At the same time,
CGC precursors migrate tangentially along the surface and form a transient layer, called the
external granule or germinal layer (EGL), where precursors divide and form immature CGCs
(Figure 1.2). As the cerebellum builds up with this influx of cells, PCs spread until they form a
monolayer of cells. Spatial constraints require that this layer folds the cerebellum into its
23

Figure 1.2. Schematic of CGC migration and maturation in the cerebellum. CGCs migrate from
the base of the developing cerebellum through several stages (gray arrows) during the third
trimester in humans and the first two postnatal weeks in rodents, both of which are considered to
be the most susceptible periods for MeHg toxicity in each group. CGC precursors (orange)
migrate along the surface of the cerebellum, accumulate in a temporary layer termed the external
granule (or germinal) layer (EGL) and divide. Division terminates for these CGC precursors, and
the resulting CGCs then migrate down Bergmann glial processes (green) through the molecular
layer (ML) where PCs are developing extensive dendrites (blue). Migration stops, and the CGCs
put out axons that travel back through the ML and produce parallel fibers which synapse onto PC
2+
dendrites. All stages of migration and maturation are dependent on finely-tuned Ca oscillations
and regulation by expression of surface receptors such as GABAARs (Komuro and Rakic, 1998).
CGCs express mostly α2 subunit-containing GABAARs during development, while expressing
mostly α6 in maturation (Takayama and Inoue, 2004). PCs express α3 subunits during
development and α1 at maturation. Adapted from Atchison 2005.
24

characteristic shape. The development of PC dendrites builds up a layer between the PCs and
EGL, the molecular layer (ML). As this occurs, CGCs that have finished dividing switch from
tangential migration to a radial migration inward, through the ML and past the PCs to a new
layer between the PCs and deep cerebellar white matter, the inner granule layer (IGL) (Figure
1.2). From the IGL, CGCs send out an axon into the ML, which bifurcates into an axonal
structure termed a parallel fiber (PF). Because all PFs are roughly parallel, they extend toward
the right and left hemispheres. PFs produce synapses on many PC dendrites along their path, and
because CGCs greatly outnumber PCs, PCs receive many PF synapses on their complex
dendritic trees.
Cerebellar growth in humans accelerates during the second half of gestation during
migration of cerebellar neurons and construction of cerebellar circuits (Scott, et al., 2012). The
cerebellar structure is fragile during this time; premature birth greatly increases the likelihood of
cerebellar damage and motor deficits in infants (Volpe, 2009). Human cerebellar development
continues after birth, including a nearly two-fold increase in volume during the period between
3-13 mo, outpacing whole brain growth (Choe, et al., 2012). Some additional CGC migration
appears to occur at least until age 4, and possibly longer. During early growth spurts, cerebellar
neuronal precursors are rapidly dividing and multiplying. Defects in cerebellar precursors
typically lead to less growth in the cerebellum, leading to cerebellar hypoplasia or even aplasia,
as seen in Joubert syndrome (Doherty, 2009). These are also periods of intense synaptic
formation and remodeling (Castejon, et al., 2004). Defects in developmental migration can occur
in various genetic disorders and some forms of autism (Métin, et al., 2008, Laure-Kamionowska
and Maslinska, 2011). Normally, neurons that either do not reach their destination or do not stop
die by apoptosis due to a lack of synaptic input. However, in these genetic disorders, clusters of
CGCs can survive in white matter or PCs scatter in many parts of the cerebellar layers.
25

Migration and growth of the cerebellum is clearly a fragile period, and injury is possible through
mechanisms such as premature birth, so toxic insults are potentially devastating. Even low-level
MeHg exposure during pregnancy, ~1 ppm as measured by maternal hair, subtly decreases the
size of a newborn’s cerebellum (Cace, et al., 2011). While injection of MeHg in mice after
weaning does induce motor defects, treatment during gestation and pre-weaning causes more
severe deficits, even if treatment is ended at weaning (Huang, et al., 2011). MeHg treatment is
2+

also able to delay CGC migration through a Ca -dependent mechanism (Mancini, et al., 2009,
Fahrion, et al., 2012). While developmental exposures are clearly susceptible to MeHg, MeHg is
able to interfere directly at mature parallel fiber and climbing fiber synapses, blocking them
completely through both presynaptic and postsynaptic mechanisms after an initial increase in
excitation (Yuan and Atchison, 1999).
d. GABAA receptors
i. GABAA receptor structure and function
GABAARs are members of the family of Cys-loop receptors, which include nicotinic
acetylcholine receptors (nAChRs), 5-HT3 serotonin receptors, and glycine receptors (Olsen and
Tobin, 1990, Olsen, et al., 1991, Tsang, et al., 2007, Tsetlin, et al., 2010). All Cys-loop receptors
are pentameric membrane-spanning receptors with a central ion channel pore. GABAARs are
generally formed from two α subunits (α1-6), two β (β1-3) subunits and one of several available
additional subunits (γ1 or 2, δ, ε, θ, or π), arranged with alternation of the α and β subunits, with
the additional subunit located between an α and β (Jacob, et al., 2008) (Figure 1.3). The α and β
subunits within a receptor do not even have to be the same numbered type (e.g. α1α6β2β3γ2)
(Sigel and Baur, 2000). The various combinations of subunits determine many characteristics of
the receptor, including available binding sites, ligand affinity and the electrophysiological
characteristics of the channel (Saxena and Macdonald, 1996, Dhulst, et al., 2009). Almost any
26

combination can be expressed in vitro, but in many of the expression systems there is no
preference for construction, producing non-physiological receptors such as those with only α
subunits. Many combinations still potentially exist in vivo, along with additional modifications
such as slice variants and post-translational modifications allow neurons to tailor the exact
purpose of GABAARs even down to specialized regional combinations in the membrane (Lo, et
al., 2010). However, many of these potential constructions are not common, and many neurons
have typical variants associated with them.
As with many neuronal receptors and channels, cytosolic domains are where most of the
interactions with other proteins occur, as well as domains that can be modified by
phosphorylation. Phosphorylation of GABAAR subunits may be important for surface expression
through stabilization. In PCs continually stimulated by the glutamatergic climbing fibers,
GABAARs are dephosphorylated, leading to dissociation from GABAAR associated protein
(GABARAP) scaffold and decreased surface expression (Qian, et al., 2011). This relationship
between phosphorylation and expression has been shown with the α4 subunit in other parts of the
brain (Abramian et al., 2010). This change also increases activity of receptors with the α4
subunit by slowing run down. Several additional scaffolding proteins and interacting factors
appear to be critical for forming GABAAR complexes with other membrane and cytosolic
components, such as post-synaptic domains. Gephyrin, while generally associated with glycine
receptors, also forms scaffolds with GABAAR subunits α1, α2, and α3 (Tretter, et al., 2012).
Removal of the synaptic GABAAR scaffold gephyrin reduces surface expression of both
synaptic and extrasynaptic GABAARs, but it is unclear if this is due to a reduction in innervation
and GABA spillover or another mechanism (Marchionni, et al., 2009). Further characterization
of modifications of GABAARs as well as interactions with scaffolding proteins could elucidate

27

Figure 1.3. Cartoon depiction of the GABAA receptor, a heteropentameric ion channel that
conducts Cl . In vivo it typically forms a unit from two α (from 6 subtypes), two β (from 3
subtypes), and one additional subunit, including γ and δ. GABAARs have two GABA
binding sites, located at the interface of α and β subunits, which based on binding studies are
also the location of the agonist muscimol and antagonist bicuculline binding sites. The
construction of these interfaces controls GABA binding and thus the physiology of the
receptor. An additional site where benzodiazepines (BZs) bind and potentiate GABA
activation is located between α and γ subunits, so GABAARs without γ subunit are
insensitive to these modulators. Synaptic scaffolds such as gephyrin bind to certain subtypes
of GABAARs, clustering them inside synapses, whereas others such as those containing α6
can be found spread around the membrane. The direction of ion flow of depends on the Cl
gradient. An inward gradient is present at most post-synaptic domains and most mature cells
and activation produces hyperpolarization and thus inhibition of activity. However, in
developing neurons there is no extrusion of Cl through potassium-chloride co-transporter
(KCC2), and thus the gradient is outward, so activation causes depolarization. This reversal
of the gradient may also be present in some locations of mature cells, such as axons.
28

functional roles of the receptors, such as cross-talk in post-synaptic domains or functional
pairing in presynaptic domains.
Unlike some other neuronal receptors, GABAARs can function both inside and outside of
synapses. Even finer localization of GABAARs may be important in determining function,
including synaptic and extrasynaptic receptors on dendrites, soma, axon and presynaptic sites.
Localization of GABAARs is generally determined by α subunit composition, where α1 is almost
always synaptic, whereas α6 and α4 are usually extrasynaptic (Nusser, et al., 1995, Olsen and
Sieghart, 2009). Chimeras of these receptors with switched intracellular domains also switch
localization (Wu, et al., 2012). Receptors with α2, α3 and α5 have not been as thoroughly
characterized, although it appears that α2 and α3 can be found synaptically or extrasynaptically,
while α5 is mostly extrasynaptic, even though it clusters similar to synaptic receptors. In many
synapses, GABAARs may also be found on the presynaptic surface, and are critical in
modulating neurotransmitter release (Engelman and MacDermott, 2004).
Several alterations to neuronal receptors, including certain GABAAR subunits, can alter
the expression of other combinations. Knockout of the α6 subunit of GABAARs selectively
degrades the δ subunit without changing its mRNA expression, indicating that other receptor
subunits are not able to construct surface receptors with δ in vivo (Jones, et al, 1997). In
knockout mice, deletion of the GABAAR α6 subunit in mice results in ~50% GABAAR decrease
overall in the cerebellum and the complete removal of surface δ subunits (Nusser, et al., 1999).
AMPA and kainate receptors, additional ligand-gated Ca

2+

channels in the CNS, can alter the

2+

expression of various GABA subunits through [Ca ]i -dependent mechanisms (Payne, et al.,
2008). The Stargazer mouse, which has mutations of the AMPA receptor that regulates its
trafficking as well as mutations to a VGCC subunit, also decreases GABAAR α6 expression in
CGCs, but increases α1 subunits expressed in synapses (Letts, et al., 1998, Payne, et al., 2007).
29

Unlike the α6 knockout, this mouse expresses considerable neurological dysfunction, and the
compensation of α1 may contribute to them. Another VGCC mutant mouse, tottering, has
deficits in neurotransmitter release and alterations to GABAAR expression, particularly the
GABAAR α6 subunit (Tehrani and Barnes, 1995, Kaja, et al., 2007). These mutations and
deficits indicate a strong link between VGCC and GABAAR subunit expression.
ii. GABAA receptor signaling
-

GABAARs in neurons, like glycine receptors, open to Cl with activation from their
major ligand, in this case GABA (Lehoullier and Ticku, 1989). GABA is synthesized in neurons
from glutamate by the enzymes glutamic acid decarboxylase 65 and 67 (GAD65 and GAD67).
These enzymes are frequently used as a marker for GABAergic synapses. GABAergic synapses
release GABA and produce inhibitory post-synaptic potentials (iPSPs) in their target neurons.
GABAergic transmission is the primary inhibitory mechanism in the CNS. iPSPs are capable of
interrupting action potentials and generally lowering excitability of target neurons. As with many
other synapses, GABA is recovered from the synapse by high affinity GABA transporters (GAT)
in nerve terminals where it can be reloaded into vesicles, or in surrounding glia, where it can be
metabolized. Some synapses appear to have GABAARs arranged around synapses to detect
overflow if too much GABA is released and escapes the synapse. GABAARs can be also found
in synapses with other neurotransmitters, and GABA can be released from the same vesicles as
other neurotransmitters or independent vesicle pools (Shrivastava, et al., 2011). When released in
this manner, GABA is typically able to reduce the efficacy of the post-synaptic response through
inhibitory action on post-synaptic GABAARs, indicating it takes part in fine tuning of response.
In addition to the standard inhibition characterized by phasic, synaptic inhibition,
GABAARs can take part in a continual form of inhibition, termed tonic inhibition. When first
described in maturing CGCs with no GABAergic synapse input, this form of inhibition was
30

found to account for as much as 99% of the inhibition of the cells (Brickley, et al., 1996). This
tonic current is a result of extrasynaptic GABAARs containing α6 and δ subunits, which have
unusually high affinity for GABA, low inactivation and little dissociation (Bianchi, et al., 2002).
Expression of the α6 subunit in cells that do not normally express them produces extrasynaptic
receptors capable of inducing tonic inhibition and inhibition of spontaneous and miniature iPSCs
at synaptic sites (Wisden, et al., 2002). Tonic inhibition appears to be able to propagate through
the cerebellum through large populations of Golgi cells releasing GABA both through synapses
and into the extracellular environment. When large populations of CGCs are inhibited tonically
in this manner, the collective glutamatergic release from parallel fibers onto PC dendrites is
reduced, and the output of the cerebellum is reduced (De Schutter, 2002). Tonic and phasic
inhibition appear to be independent of one another, In the thalamus, sustained activity of phasic
inhibition as a form of LTD learning is able to alter individual cells, but not change response to
extrasynaptic GABAAR activation (Bright and Brickley, 2008). Phasic inhibition may thus be
important for development of individual cells and memory, but tonic inhibition may be related to
collective changes in larger populations of neurons in the brain.
Interestingly, the first suggested mechanism for tonic inhibition, that the extrasynaptic
GABAARs were responsive primarily to GABA spilled from nearby synapses, has been
overturned. Even with low ambient GABA, extrasynaptic GABAARs were found to be
desensitized and unable to respond to spillover GABA from GABAergic glomeruli synapses
(Bright, et al., 2011, Brickley and Mody, 2011). This appears to indicate that the normal mode of
GABAAR-mediated tonic inhibition requires only a very few of the receptors to be open, but for
them to be open continuously, acting as hyperpolarizing channels and molecular shunts.
-

In CGC axons, it appears there may be some locations with high local [Cl ]i, leading to a
reversal of the gradient and depolarization of the axon with GABA application simulating a
31

synaptic release (Pugh and Jahr, 2011). Despite the combination of depolarization and a resulting
Ca

2+

influx, these stimuli were rarely capable of initiating an action potential and releasing
2+

neurotransmitter. However, the increase in [Ca ]i did propagate along the axon all the way to
the soma, increasing spontaneous release.
iii. GABAA receptors and calcium
GABAARs have a complex set of secondary signals, many of which involve changes to
2+

2+

[Ca ]i. In cultures of CGCs from post-natal day (PND) 6-8 rats, GABA increases [Ca ]i
through an L-type VGCC and depolarizes the neurons (Brodinsky, et al., 2003). Long-term
culture with GABA induces neuritogenesis which is blocked by Ca

2+

channel blockers as well as

mitogen activated protein kinase (mapk) / extracellular-related kinase (erk) kinase 1 (MEK1) and
calmodulin kinase II (CaMKII) blockers, suggesting several mechanisms of continuous GABA
signaling. Phosphorylation of CaMKII and erk1 and erk2 also change with this GABA
2+

application. GABA activation in PCs induces a similar transient increase in [Ca ]i that is more
pronounced in developing PCs. These transients are abolished by both bicuculline and VGCC
block (Eilers, et al., 2001). Interestingly, positive allosteric modulators of GABAARs such as
alcohol and pentobarbital inhibit the action of L-type VGCCs (Earl and Tietz, 2011). This is also
the case in some neocortical nerve terminals, where GABAAR activation and Ca

2+

influx

inhibits L-type VGCCs and neurotransmitter release (Long, et al., 2009). This makes clear that
2+

the eventual effect of [Ca ]i increases caused by GABA activation are dependent on the type of
VGCC present in the nerve terminal. In cerebellar interneurons, it appears that GABAAR
2+

activation can induce [Ca ]i increases without any concurrent depolarization or action of
VGCCs, possibly through signals triggered by changes in osmotic pressure (Chavas, et al.,
2004).

32

In CGC axons, muscimol activation of GABAARs is able to stimulate release of cytosolic
2+

glutamate without any changes to membrane potential or [Ca ]i, indicating even more
unexplored secondary signals (Raiteri, et al., 2001). Many of the effects of GABAAR activation
may be “sub-threshold”, acting in a manner that does not immediately alter the resting potential
nor directly inducing depolarizations through Ca

2+

influx (MacDermott et al., 1999). It appears
-

that due to mechanisms such as VGCC activation, equilibration with HCO3 , and other
secondary signaling events, that non-synaptic GABAAR activation may have a net excitatory
effect, rather than the presumed inhibitory effect (Kullman, et al., 2005). It will be critical in
future studies to determine the conditions under which GABA activation is excitatory versus
-

inhibitory, including the location on the cells, the existing localized Cl and Ca

2+

concentrations,

and the ligand used.
iv. GABAA receptor pharmacology
GABAARs were originally mistaken for two receptors, one that responded to GABA and
one that responded to benzodiazepines (BZs) such as diazepam and chlordiazepoxide, and was
termed the BZ receptor (Lehoullier and Ticku, 1989). This may have been in part due to the fact
that some GABAARs do not respond to BZs. This is due to the arrangement of ligand binding
sites on the receptors, with the two GABA binding sites located between α and β subunits, while
the BZ site sits between the α and γ subunit (Jacob, et al., 2008). When there is no γ subunit, as
there is in extrasynaptic δ-containing receptors, there is no BZ site and so no BZ response. BZs
are allosteric modulators of GABAARs, changing the receptor’s affinity and open dynamics.
When bound by BZs, GABAARs open more frequently in response to GABA, but BZ binding
alone is not enough to activate the receptors. BZs, through enhancing GABA signals, inhibit
neurotransmission and neuronal activity throughout the brain, generally leading to anxiolytic and
anesthetic results in mammals. Some of the effects of alcohol may be similar to BZs, due to
33

sharing a similar binding site on the GABAAR, but this mechanism has not been well
characterized yet due to the wide ranging effects of alcohol in the brain (Brickley and Mody,
2011).
Several agonists and antagonists of GABAARs have been used for many years in
electrophysiological and biochemical recordings of GABAAR activity. Muscimol is derived
from many types of Amanita mushrooms, and is the major psychoactive component of the
mushrooms. It is highly selective for GABAARs and binds the same sites as GABA, activating
the receptors without altering the ambient GABA concentrations. Muscimol may have some
subunit preference for extrasynaptic receptors due to a higher binding affinity, much like GABA
does, but this preference has not been fully characterized (Chandra, et al., 2010). The agonist
propofol acts through a binding site on the β subunit distinct from the GABA and muscimol site,
and loreclazone and etomidate demonstrates preference for β2- and β3-containing receptors
(Sanna, et al., 1995, Sanna, et al., 1996, Smith, et al., 2004, Drexler, et al., 2009). Bicuculline is
derived from a plant alkaloid and induces seizures. It is a competitive antagonist for the GABA
binding site and out-competes muscimol binding at effective concentrations. Bicuculline does
not exhibit subunit specificity, but other antagonists, such as pore blocking competitive
3+

antagonists lanthanum (La ), Zn

2+

and furosemide, show preference for extrasynaptic receptors

(Wafford, et al., 1996, Schmidd, et al., 1999, Wall, 2002).
A diverse array of modulators is available or in development for modulation of
GABAARs, some more effective on extrasynaptic receptors, but few have subunit specific
effects (Belelli, et al., 2009). Compared to more general agonists, antagonists and allosteric
modulators, subunit-specific modulators can produce widely different effects on GABAAR
signaling (Baur, et al., 2005, Khom, et al., 2010, Joksimovic, et al., 2013). One of these more
recent classes of modulators for GABAARs is neurosteroids. As with BZs, many of these
34

molecules produce sedative effects, and neurosteroids are generally considered to be similar
allosteric modulators of GABAARs, though they do not share the same binding sites. Despite
multiple sites present for binding of neurosteroids, only one needs to be occupied for
potentiation of the GABAAR (Bracamontes, et al., 2011).With a homologous structure now
available through ELIC1, additional interaction domains are being found for GABAARs, so
many more modulators may be found soon (Varagic, et al., 2013).
Finally, there are a few modulators of other receptors in the CNS that may have effects of
GABAARs as well. The AMPA receptor antagonist 6-cyano-7-nitroquinoxaline-2,3-dione
(CNQX) appears to increase GABA-induced spontaneous currents in hippocampus and
cerebellum, (Brickley et al., 2001). Not all antagonists of AMPA acted the same as CNQX and
its derivatives, which were also able to increase sIPSC frequency. This change cannot be
attributed to NMDA antagonism, as NMDA specific antagonists did not increase sIPSC
frequency.
v. GABAA receptors during development
One of the most interesting discoveries about GABAARs is that they are not always
hyperpolarizing when activated. In fact, during development and migration in almost every
neuron that expresses GABAARs, including CGCs, GABAAR activation is directly depolarizing,
-

due to a Cl concentration equilibrium that is positive compared to the cell resting potential
(Owens, et al., 1996). As neurons migrate and mature, the electrochemical equilibrium potential
becomes more negative, and the cells instead hyperpolarize with application of GABA.
Expression of GABAARs also changes with migration and maturation. It appears that CGCs
express mostly the α2 subunit in GABAARs during migration in the cerebellum whereas α6
rapidly becomes the dominant receptor type once they reach the IGL (Takayama and Inoue,
2004). In these same treatments, immature PCs appear to express more α3 subunits while mature
35

PCs have mostly α1 receptors. As with GABAAR activation in adult neurons, activation of
2+

immature GABAARs also increases [Ca ]i through a VGCC-dependent mechanisms. This can
2+

be recapitulated in cultured CGCs from PND 6-8 rats, where GABA increases [Ca ]i through an
L-type VGCC and depolarizes the neurons (Brodinsky, et al., 2003). When these CGCs are
cultured in a depolarizing medium, simulating the formation of active synapses common in
maturation, there is a gradual decrease in the depolarizing effect of GABA. Conversely, CGCs
cultured in non-depolarizing medium can remain responsive for several days in vitro. Similarly,
GABA reversal potentials change over PC development alongside CGCs, becoming more
negative as the cells develop to a final equilibrium of -87 mV (Eilers, et al., 2001).
2+

Depolarization and increases in [Ca ]i increase δ subunit expression in CGCs, suggesting that
migration and the IGL environment induces expression of extrasynaptic GABAARs (Gault and
Siegal, 1997).
The critical point determining GABAAR depolarization versus hyperpolarization appears
to come from expression of the Potassium-Chloride Co-transporter type 2 (KCC2). KCC2 acts as
a stop signal to many migrating neurons, as its expression in the membrane is able to reduce the
efficacy of GABA-induced depolarization, which is a driving force of migration via Ca

2+

oscillations (Bortone and Polleaux, 2009). Phosphorylation of KCC2 by protein kinase C (PKC)
appears to be critical for its membrane trafficking and stability, indicating another regulatory
step in maturation (Lee, et al., 2007). Because of its association with specific subtypes of
GABAAR expression and thus inhibition, alterations in KCC2 expression in the mature brain can
be associated with epileptic seizures as well (Mao, et al., 2011). Once established, KCC2 surface
expression can still be altered, as with ischemia, so the relative effects of GABAAR activation
+

-

and K /Cl transport may still change in adult and aging neurons (Galeffi, et al., 2004). Golgi
cells are highly active during the migration of CGCs, and rat CGCs receive inhibitory Golgi
36

synapses in the IGL as early as PND 3 (Farrant and Brickley, 2003). Synaptic GABA could
signal the end of migration for individual cells and promote KCC2 expression.
An argument has been made that depolarizing actions of GABA in developing neurons is
an artifact of slicing and other preparations (Bregetovski and Bernard, 2012), but evidence from
many model organisms and non-invasive techniques strongly refutes this (Ben-Ari, et al., 2012).
Even the original discovery of tonic inhibition indicates that GABA depolarized the developing
rat CGCs before they matured (Brickley, et al., 1996). This is perhaps best demonstrated, though,
in live intact organisms. In early postnatal mice, when neurons can be imaged by two-photon
microscopy while still in the live animal, depolarizing GABA is critical for migration of
interneurons in the cortex, (Inada, et al., 2011).
vi. GABAA receptor disorders and toxicology
Several disorders are associated with alterations to GABAAR function or to general
GABA signaling. Knockouts of several subunits of GABAARs display seizure phenotypes,
typically caused by overactive signaling in discrete brain regions, indicating that GABAAR
signaling is critical for general function (Belelli, et al., 2009). Because of their inhibitory
importance, GABAARs that are mutated in a way that prevents appropriate translation, folding,
trafficking or function are associated with several types of epilepsy, including child absence
epilepsy, Dravet syndrome, generalized epilepsy with febrile seizures plus, juvenile myoclonic
epilepsy and severe myoclonic epilepsy in infancy (Macdonald, et al., 2010). Cerebellar and
hippocampal GABA increases, as brought about by inhibition of GABA uptake through GABA
transporter type 1 (GAT1), cause tremor, anxiety, and ataxia, both in knockout models and with
the antiepileptic GAT1 inhibitor tiagabine (Chiu, et al., 2005). Alternatively, overexpression of
GAT1 through a human mutation that decreases GAT1 degradation in turn decreases ambient
GABA in the cerebellum and tonic inhibition (Egawa, et al., 2012). This in turn leads to a variety
37

of disorders, including developmental delay, lack of balance, speech impairment and epilepsy.
This balance of GABA is critical to drugs that can alter the levels of GABA in the brain, such as
tiagabine, so this must be taken into consideration for treatment strategies such as those for
epilepsy (Houston, et al., 2012).
Several studies indicate direct effects of MeHg on GABAARs. At most relevant
concentrations, MeHg potentiates activation of GABAARs, as indicated by larger amplitudes of
spontaneous and evoked currents in the presence of ambient GABA (Arakawa, et al., 1991).
Higher concentrations instead block GABAARs, indicating a possible mechanism that may
explain aberrant results in some studies. MeHg, as well as Hg

2+

and the sulfhydryl alkylating

agent n-ethylmaleimide, potentiate binding of BZs to GABAARs, suggesting this is the
immediate effect of MeHg on these receptors (Fonfria, et al., 2001). MeHg also shows subunit
specificity in its action, acting on α6-containing receptors at lower concentrations than on α1
receptors, significant changes to GABAARs occur at lower concentrations of MeHg than any
other receptor, including VGCCs (Yuan and Atchison, 2003, Yuan, et al., 2005). These
differences may contribute to different cell viability when comparing cells expressing mostly
GABAAR α6 subunits such as CGCs compared to cells expressing mostly α1, such as PCs or
cortical neurons (Herden, et al., 2008).
Some regulation of GABAARs appears to occur due to internalization of the receptors,
which can occur with both acute and chronic stress such as pregnancy, and appears to be due to
excess activity (Maguire, et al., 2009, Poulter, et al., 2009). While toxic levels of activity
frequently alter the surface expression of GABAARs and decrease GABA binding, this is not due
to changes in membrane potential (Lyons, et al., 2001). Instead, it appears expression changes
2+

are mediated through increases in [Ca ]i and decoupling of GABA binding appears to relate to
VGCC activation. Mink fed environmentally-relevant concentrations of MeHg for three months
38

showed decreases of GABAAR expression up to 94% in cerebellum, along with decreased
metabolic production of GABA and GABA receptor binding in other regions of the brain
relevant to motor disorders (Basu, et al., 2010).
Some insight may also be gained by examining other toxicants that affect GABAARs.
Dieldrin, a pesticide, antagonizes GABAARs as well as the structurally-similar glycine receptor.
Long-term cell exposure to dieldrin antagonizes GABA and NMDA receptors and decreases
surface expression of both, in turn reducing glutamate neurotransmission and excitotoxicity
(Vale, et al., 2003, Babot, et al. 2007). Many of the effects of dieldrin can be directly reversed by
the neurosteroid allopregnanolone, suggesting that neurosteroids are potent and lasting
competitors for actions on GABAARs. When several chemicals and drugs of abuse were tested
in neuronal cell cultures, assays for GABAAR function, GABA uptake and cell membrane
potential were the best indicators of neurotoxicity in humans (Galofre, et al., 2010).
B) Objectives and Rationale
a. Hypothesis and aims
2+

Because of their individual interactions with Ca , we hypothesize that the interaction of
2+

MeHg and GABAARs will have effects on [Ca ]i in CGCs, and that these effects will be
dependent on age and expression of receptor subunits. Previous experiments in the lab have
begun to probe these interactions, through electrophysiology and other approaches (Yuan and
Atchison, 2003, 2007, Mancini, dissertation, 2006). Direct interactions, through comparisons of
cells with MeHg and GABAAR modulators, or through comparisons of cells in different states,
such as developing versus mature cells, have not been characterized thoroughly. Three projects
2+

were developed to characterize the interactions of MeHg and GABAARs as they relate to [Ca ]i
in CGCs. The first, presented in Chapter 2, uses neonatal rats, the model for developmental
toxicity used in this lab and others for many years. Neonatal mice are used in Chapter 3 in order
39

to introduce the additional factor of GABAAR subunit knockout, in this case knockout of the α6
subunit, the expression of which marks another difference between developing and mature
2+

CGCs. Chronic, lifetime exposure is probed in mice in Chapter 4, including several Ca dependent mechanisms in addition to GABAAR modulation. Together these aims characterize
the interactions of GABAARs, MeHg and Ca

2+

in susceptible CGC populations of both

developing and aging model populations.
b. Ca

2+

imaging in cerebellar slices from rats and mice as a model for MeHg toxicity
From the background described, it is clear that the underlying mechanisms of MeHg

toxicity, particularly mechanisms of selective neurotoxicity, remain elusive, and there are many
potential approaches. One approach to studying MeHg toxicity is to model susceptible and
potentially susceptible groups, including cell types and genotypes, against more resistant groups
in the same environment. Thankfully, previous research has developed models and techniques
that make this approach feasible. Slice culture has long been a convenient method for the study
of electrophysiology, and has developed into a method for other forms of live study as well
(Tashiro, et al., 2006). Finely cut slices of neural tissue from animals are able to maintain live,
functional cells for several hours, and keep active synapses and other intercellular interactions
once outside the brain. In cerebellar slices, we can examine and image CGCs, PCs and the other
various cells and components in the cerebellum without the disruption that occurs with
dissociation or similar approaches. Imaging in slice culture also allows for the possibility of
examining in three dimensions, rather than in the relatively planar culture of cells. Computer
modeling of the various components of cerebellar circuits reinforces the importance of working
with a three-dimensional model as can be found in cerebellar slice culture (Berends, et al., 2004).
Advances in these techniques are even capable of maintaining large cell populations for many

40

days in culture media (Mancini and Atchison, 2009), though much of the structure does
eventually degrade.
One drawback of slice culture when compared to cell culture or live animal study is that
the act of slicing by nature destroys and damages cells near the plane of the slice. This in turn
leads to progressive degeneration of the slice over time (Kasischke, et al., 2001). In cerebellar
slices, especially of immature cerebella, slicing is also rapidly toxic to PCs, in part due to a
release of GABA from sliced cells (Rakotomamonjy, et al., 2011). Because we are interested
2+

more in CGCs, and because maintaining depolarized slices and using Ca
other inhibitors could interfere with our ability to measure Ca

2+

channel blockers or

accurately, we are using

standard acute slice methods. These include important steps in slice culture incubation, including
temperature control, rapid progression and precise concentrations of ions in buffering solutions
that mimic cerebrospinal fluid to control osmolarity and reduce the actions of damaging ions
released from cut cells. The initial concentrations of these artificial cerebrospinal fluid (ACSF)
buffers are based on previous experiments in the lab (Yuan and Atchison, 2003, 2007). These
protocols are updated and optimized as needed, particularly for handling tissue from older
animals.
One goal of this dissertation is to examine different populations of cells that are
considered more or less susceptible to MeHg yet reside within the same environment. This
includes differences between CGCs and PCs, but primarily focuses on differences in cells at
different stages of development. Conveniently, windows for the peak of CGC migration are 9-12
PND in rats and 8-11 PND in mice, as indicated by a comparative database
(www.translatingtime.net, Clancy, et al., 2007, Workman, et al., 2013). With appropriate
selection of images of slices from these animals, we can see CGCs in all stages, from dividing in
the EGL, migrating in the ML, and maturing and forming synapses in the IGL. We can make
41

comparisons based on populations of these cells with treatments of MeHg. In addition to
characterizing neonatal slice preparations, we had the opportunity to examine slice preparations
from adult and geriatric mice through an overarching project involving chronic MeHg treatment
in mice and co-treatment with a VGCC antagonist, isradipine.
Another goal of this dissertation is to examine a potential genetic susceptibility to MeHg,
in this case a knockout of the GABAAR subunit α6. Knocking out the α6 subunit causes a 50%
overall reduction in GABAAR expression in the cerebellum, no changes to synaptic α1 subunits,
and no gross changes to phenotype (Nusser, et al., 1999). A permanent knockout of α6, caused
by insertion of a neomycin cassette into the subunit gene, has been produced (Homanics, et al.,
1997). It is of interest in this approach in part because though it has no obvious phenotype and
no obvious difference in response to some agents thought to act on the subunit, we are certain of
a difference in the selectivity of MeHg between extrasynaptic and synaptic GABAARs and we
have evidence it acts preferentially on α6 receptors (Yuan and Atchison, 2003, Yuan, et al.,
2005, Herden, et al., 2008). Recording from these animals would reinforce our data on the
selectivity of MeHg, and characterize differences from other neurotoxicants acting on
GABAARs.
While there are many approaches for imaging and biochemical analysis of cells in slice,
2+

in this dissertation, the approach chosen for imaging is live-cell Ca

fluorescence imaging. This

approach uses live tissue and thus can elucidate changes over time in addition to special
2+

differences. [Ca ]i is clearly important to the toxicity of MeHg, and we have evidence of major
2+

changes to [Ca ]i as shown by Ca

2+

fluorescence imaging in several cell types and preparations,

including neuroblastoma-glioma cultures, synaptosome preparations and CGC cultures (Hare, et
al., 1993, Denny, et al., 1993, Marty and Atchison, 1997, Marty and Atchison, 1998).
2+

Additionally, we have evidence of [Ca ]i increases throughout the cerebellum in slices, but
42

these responses need further characterization, (Yuan and Atchison, 2007). Such characterization
includes determination of what concentrations of MeHg are effective in these slices over various
2+

periods of time as well as whether or not the time course of MeHg disruption of [Ca ]i can be
modulated by agents that affect Ca
This Ca

2+

2+

signaling.

imaging approach allows for exploration of some additional hypotheses as
2+

well. The GABAAR connection to [Ca ]i is unclear, but may be dependent on the same cellular
and developmental differences as MeHg toxicity susceptibility and has some clear connections to
the toxicity itself. Therefore, we seek to characterize this interaction in these slice preparations
through the use of the nonspecific agonist muscimol and antagonist bicuculline.
There are generally two approaches to loading fluorescent Ca

2+

probes into neurons,

bolus injection and bulk loading (Yuste, et al., 2011). While injection appears to produce the best
sensitivity to changes in fluorescence, bulk loading highlights many cells and structures, and is
more in line with our needs. Bulk loading of acetoxymethyl (AM) esters can be uneven
depending on temperature and cell type, and some cells and techniques also produce inactive
fluorescent intermediates, so it is important to have controls and optimize protocols (Cannell and
Cody, 2006). Thankfully, this research group has had many years experience working with bulk
loading of fluorescent probes, and has developed and optimized these techniques already.
Another drawback in Ca

2+

imaging is that intense laser light produces reactive oxygen

species in neurons, so it is critical to reduce light intensity and duration to avoid phototoxicity
(Knight, et al., 2003). Careful attention is paid in this dissertation to techniques that can reduce
phototoxicity, especially where phototoxicity can directly contribute to changes in fluorescence
of both the cells and the fluorophores. Confocal laser light intensity is kept to a minimum, while
the shutter system on the epifluorescence microscope minimizes illumination time. Where

43

possible, control treatments were run with only the illumination and untreated perfusion buffers
in order to establish baseline levels of fluorescence and cell death during experiments.
Finally, with large numbers of cells and complex treatments over time, it becomes time
consuming to identify and track Ca

2+

imaging experiments over time manually, so automated

cell identification and tracking is valuable (Wong, et al., 2010). Techniques used in this
dissertation include both manual and automated approaches, depending on the bulk and
complexity of data used. The automation techniques used may be important for further study, as
the complexities involved in comparisons that must be made to characterize MeHg toxicity are
increasing.

44

CHAPTER TWO

METHYLMERCURY-ASSOCIATED INCREASES IN FLUO4 FLUORESCENCE IN
NEONATAL RAT CEREBELLAR SLICES ARE DEPENDENT ON GRANULE CELL
MIGRATIONAL STAGE AND MODULATED BY GABAA RECEPTORS.

45

A) ABSTRACT
Methylmercury (MeHg) induces permanent motor dysfunction through damage to the
cerebellum in mammals and is of environmental concern due to bioaccumulation in diets and
enhanced developmental susceptibility. MeHg induces neuronal cell death, particularly in
developing cerebellar granule cells (CGCs). The critical mechanisms by which MeHg is toxic
and selective for cell type are unclear. MeHg treatment alters current through many neuronal
2+

channels and disrupts divalent cation homeostasis, particularly of Ca , through multiple paths
leading to toxicity. MeHg could act in CGCs via disruption of the γ-aminobutyric acid receptor
A (GABAAR), a target of MeHg even at low concentrations that can regulate Ca

2+

signaling in

CGCs, particularly during migration. Using confocal microscopy and the fluorophore fluo4, we
studied the disruption of internal divalent cation concentrations in slices of neonatal rat
cerebellum. Slice preparation allows visualization of developing CGCs in multiple stages of their
migration. We treated slices acutely with several concentrations of MeHg to determine Ca

2+

response. In addition, we used the nonselective GABAAR agonist muscimol and antagonist
bicuculline to test their effects on Ca

2+

and interactions with MeHg treatment. Acute MeHg

concentrations increased fluo4 fluorescence within 40 min of treatment up to 260% of baseline
in developing CGCs, but increased values only up to 170% of baseline in mature CGCs, showing
a difference in cell response based on developmental stage. MeHg increased the fluorescence of
processes in the slice as well, but the associated fluorescence was diffuse and could not be
quantified as belonging to a single type of process. Purkinje cells (PCs), which are more resistant
to MeHg toxicity, rarely fluoresced in these slices, and so could not be compared to CGCs either.
Both muscimol and bicuculline reduced the effects of MeHg over 40 min. These results indicate
that immediate disruption of Ca

2+

homeostasis in CGCs is in part through effects on GABAARs.

46

B) INTRODUCTION
Methylmercury (MeHg) is a widespread environmental neurotoxicant known to affect the
cerebellum (Hunter and Russell, 1954, Takeuchi et al., 1962, Bakir et al., 1973). In particular,
MeHg is toxic to cerebellar granule cells (CGCs), the smallest and most numerous neurons in the
brain. These neurons die in both chronic and acute MeHg poisoning, whereas a higher
percentage of neighboring Purkinje cells (PCs) survive (Hunter and Russell, 1954, Sakamoto et
al., 1998, Edwards et al., 2005, Yuan and Atchison, 2007).
CGC precursors divide and migrate from the external granule cell layer (EGL) through
the maturing PC dendrites in the molecular layer (ML) and mature in the internal granule layer
(IGL) (Komuro and Rakic, 1998). Human fetuses exposed to MeHg during CGC migration show
the greatest susceptibility to toxicity (Amin-Zaki et al., 1974). Rats and mice are also susceptible
to MeHg during a postnatal period of CGC migration, suggesting it is not the time before or after
birth that indicates susceptibility, but some mechanism involved in neuronal migration and
survival (Rice and Barone, 2000, Sakamoto, 2004).
MeHg interacts with and binds many proteins and small molecules, but circulating MeHg
initially encounters cell surface proteins. Electrophysiological studies indicate that MeHg
interacts with several types of voltage-gated Ca

2+

channel (VGCC) (Shafer and Atchison, 1991,

Sirois and Atchison, 2000, Peng et al., 2002, Hajela et al., 2003). These VGCCs are involved in
the normal function of neurons, and do not typically produce large or lasting changes in internal
Ca

2+

2+

2+

concentrations ([Ca ]i). Maintenance of low [Ca ]i is critical for cell viability and

signaling, and disruption of intracellular divalent cation homeostasis is a common sign of
cellular MeHg toxicity (Hare, et al., 1993, Marty and Atchison, 1998, Edwards, et al., 2005). In
2+

CGCs, oscillating [Ca ]i is an important migration signal; disruption of Ca

47

2+

oscillation can

accelerate, delay, halt or even reverse CGC migration (Komuro and Kumada, 2005). Continuous
MeHg treatment delays CGC migration in all layers, suggesting that disruption of Ca

2+

oscillations occurs (Mancini et al., 2010).
In addition to Ca

2+

-

channels, MeHg interacts with GABAARs, ligand-gated Cl channels

that act as the principal inhibitory receptors in the mammalian adult brain (Yuan and Atchison,
1997; Fonfria et al., 2001; Yuan and Atchison, 2003). Cell surface function and subunit
composition of GABAAR differs between CGCs and PCs, which are more resistant to MeHg,
including differences in α subunit stoichiometry. These subunits comprise two of the five
subunits of the channel (Nusser et al., 1998, Joyce, 2007). CGCs contain an abundance of
extrasynaptic GABAARs with α6 subunits while PCs contain only synaptic α1 subunits.
GABAARs containing α1 produce fast inhibitory postsynaptic currents, whereas α6-containing
GABAARs produce a hyperpolarizing current with little inactivation, termed tonic current
(Saxena and MacDonald, 1996, Bianchi et al., 2002). Tonic current maintains prolonged
2+

inhibition and a low [Ca ]i (Brickley et al., 1996, Brickley et al., 2001). In isolated cells in
culture, α6-containing CGCs and cortical cells containing only α1 GABAARs displayed similar
current block by MeHg in both cell types (Herden et al., 2008). This suggests either that
additional differences between CGCs and other cells occur or interactions among the receptor
subunits contribute to susceptibility, possibly including other GABAAR subunits associated with
α6 or conditions unique to slice culture.
In addition to expression differences between CGCs and PCs, GABAARs undergo
important changes during development. Before migration, GABAAR activation depolarizes
2+

immature CGCs, opening VGCCs and increasing [Ca ]i (Owens et al., 1996). This is due to the
accumulation of intracellular chloride via activation of the sodium-potassium-chloride

48

transporter NKCC1 (Yamada et al., 2004). During CGC migration, GABAARs switch from
excitatory to inhibitory as expression of the potassium-chloride transporter KCC2 overtakes
NKCC1, reducing intracellular chloride concentrations (Zheng et al., 2003, Takayama and Inoue,
2006). During this period, expression of CGC GABAAR subunits changes, including expression
changes among relatively unstudied and transient receptor subtypes, such as α2 (Takayama and
-

Inoue, 2004). These findings suggest that both changes in the Cl ionic gradient and GABAAR
expression may contribute to the effects of MeHg during cerebellar development.
We sought to test the hypothesis that CGCs at different stages of development react
differentially to MeHg by investigating the acute actions of MeHg on developing CGCs. Slices
of brain maintained in oxygenated artificial cerebrospinal fluid (ACSF) permit rapid and
controlled treatment of cells with toxicant, while retaining much of their structure and function.
Because CGCs at different stages of development appear to have differential sensitivities to
MeHg, we wanted to determine if there is a difference in acutely applied MeHg-dependent
2+

disruption of [Ca ]i. We used the fluorescent indicator fluo4, which labels free divalent cations,
2+

to indicate relative changes in [Ca ]i. In addition, we tested the involvement of GABAARs with
the non-specific GABAAR agonist and antagonist muscimol and bicuculline, respectively. While
the mechanisms of cerebellar toxicity appear similar between acute and chronic MeHg exposure,
the insight gained from these acute studies would require further study in chronic exposures to
be certain of their similarity.

49

C) MATERIALS AND METHODS
a. Chemicals and solutions
Methylmercuric (II) chloride was obtained from Aldrich Chemical (Milwaukee, WI).
Fluo4 (NW, no wash), ethidium homodimer-1 (ethD-1) and probenecid were obtained from
Invitrogen Molecular Probes (Eugene, OR). Muscimol hydrobromide and bicuculline
methobromide were obtained from Sigma Chemical Co. (St. Louis, MO).
Tissue dissections were carried out in a solution containing (in mM): 222.5, sucrose; 2.5,
KCl; 4, MgCl2; 1.25, KH2PO4; 26, NaHCO3; 1, CaCl2; and 25, D-glucose (pH 7.3-7.5 when
+

2+

2+

oxygenated) at room temperature of 23-25 C. These low [Na ], low [Ca ] high [Mg ] slicing
solutions were used to reduce cellular damage during slicing. Slices were kept in this solution for
no longer than 15 min. Further incubations were performed in ACSF containing (in mM): 125,
NaCl; 2.5, KCl; 1, MgCl2; 1.25, KH2PO4; 26, NaHCO3; 2, CaCl2; and 20, D-glucose (pH 7.37.5 when oxygenated). ACSF was supplemented with MeHg from a 10 mM stock solution in
double-distilled H2O for treatments. All working solutions were prepared within 48 h of use and
oxygenated with 95/5% O2/CO2 at room temperature during treatments. 2x stock of fluo4 was
prepared according to the supplier’s instructions and diluted to a final concentration of 4 μM in
ACSF supplemented with 2.5 μM probenecid to facilitate fluorophore loading (Mehlin et al.,
2003). Stock solutions of 100 mM muscimol and 10 mM bicuculline were prepared in doubledistilled H2O and stored at -20C until use. Choice of concentrations of bicuculline and
muscimol were based on previous electrophysiological experiments showing bath application of
100 μM muscimol and 10 μM bicuculline depolarize and block, respectively, cells in various
brain slices (Yuan and Atchison 1997, 2003), and optimizations showing a temporary increase in
2+

[Ca ]i from bath application of 100 μM muscimol in slices.

50

b. Acute slice preparation
All animal procedures are in accordance with NIH guidelines for experimental animal
use and were approved by the Michigan State University Institutional Animal Care and Use
Committee. Timed-pregnant, female Sprague-Dawley rats were obtained from Charles River
Laboratories (Cambridge, MA). Pregnant and nursing dams were fed high-fat diets (Purina 7024,
Land O Lakes Purina, Richmond, IN) to improve litter size. All rats were given doubly-distilled
water to drink ad libitum.
Male and female rat pups (M=24, F=14, postnatal days 8-12) were euthanized by rapid
decapitation. Their brains were rapidly removed. Cerebella were maintained in ice-cold,
oxygenated solution during slicing. Sagittal slices 250 μm thick were cut using a Leica VT100S
vibratome (Leica Microsystems Inc., Bannockburn, IL). After slicing, slices were incubated in
oxygenated ACSF at room temperature for 1 h.
Incubations in ACSF alone were followed by incubations in fluorophores in ACSF, either
for cytotoxicity assays or for time course studies. A live/dead assay using fluo4/ethD-1 was used
to determine if the concentrations of MeHg used in time course studies were lethal to cells in
slices during the imaging period. Slices were incubated in ACSF with stock fluo4 and 4 μM
ethD-1 for 2 h (post-incubation control), for 3 h (+1 h ACSF) or for 3 h supplemented with 20
μM MeHg during the final h (+1 h MeHg). Due to time-constraints in imaging, these were not
matched to the 40min MeHg treatments of the time courses, but were kept to 60min treatment
and the highest concentration used in time course assays, 20 μM MeHg, to represent the harshest
treatment tested. All other treatments incubated slices in ACSF with 1X fluo4 from 2-8 h as
needed. In some cases, multiple slices from the same animal were used on a treatment day, but
no slices from individual animals or animals in the same litter were given the same treatment.

51

c. Confocal microscopy
For each treatment, a cerebellar slice was removed from fluorophore incubation and
anchored in a microscope perfusion chamber with gravity-fed oxygenated ACSF flowing at 2-3
ml/min at room temperature, 23-25C. Slices were visualized under 10x and 40x water
immersion objectives (NA 0.3 and 0.8, respectively) fitted to an upright Leica TSL confocal
scanning microscope (Leica Microsystems Inc., Bannockburn, IL). Magnified regions of the
slices were examined using an argon laser at 488 nm excitation. To reduce the incidence of
photobleaching, laser power was set at less than 30% of maximum. To maximize the number of
cells available for analysis, 40x magnified regions were chosen that contained fluorescent cells
in all cerebellar layers, based on initial visualization at 10x magnification.
Slices incubated for the cytotoxicity assays were sequentially scanned for fluo4 and
ethD-1. Imaging was performed at 488 and 543 nm laser excitation with emission filters for
fluo4 (500-535 nm) and ethD-1 (556-700 nm). For single-label experiments, a wider fluo4
emission could be traced, so an emission filter of 500-700 nm was used. Imaging parameters for
both fluorophores were scaled manually to the maximum and minimum saturation limits for
pixel values before taking scans. Images were taken at a 512x512 pixel resolution representing a
2

(300 μm) field of view. The full thickness of each slice, typically 100-160 μm, was scanned as a
z-series of images representing an image stack. Between-image depth was set at 3-5 μm to
decrease imaging time and minimize photobleaching while still taking multiple images of each
cell. Imaging of cytotoxicity assays included two regions of each slice, taken at 10x and 40x
magnification. For time course experiments, one region of the slice was selected. To maintain
visualization of the same cells through the experiment, framing of the image was manually reset
between time points if the slice shifted visibly.

52

Figure 2.1. A A representative pseudocolor fluorescent image of an acutely-isolated cerebellar
slice from PND9 rat showing modified fluo4/ethD-1 live/dead cytotoxicity assay, postincubation control. Images are from internal folds of the developing cerebellum. Images show
maximum fluo4 and ethD-1 fluorescence values by pixel. Approximate pial surface of an inner
fold is indicated by a white line and other boundaries are indicated by dashed lines. EthD-1 (red),
an indicator of dead cells, and Fluo4AM (green), an indicator of relative intracellular calcium
concentration, do not colocalize. Scale bar represents 30 μm.
53

Figure 2.1 (cont’d). B Representative image from separate slice from the same cerebellum after
one additional h in ACSF supplemented with 20 μM MeHg. C EthD-1 mean pixel staining (%
pixels ± SEM) comparison of layers post-incubation in fluo4 and ethD-1 (n=5), with an
additional 1 h incubation in ACSF (n=3) or with an additional 1 h incubation in 20 μM MeHg
ACSF (n=3). Abbreviations: EGL, external granule layer; ML, molecular layer; IGL, internal
granule layer.

54

2+

Several concentrations of MeHg were tested for effects on [Ca ]i. In these time course
treatments, two pretreatment time points (-10 and -5 min) established a baseline fluo4 signal,
followed by 40 min of continuous perfusion with a single concentration of either 1, 5, 10 or 20
μM MeHg starting at 0 min. These concentrations are higher than those used for isolated cells in
culture due to enhanced non-specific binding of MeHg in tissue. They fall between 1 μM MeHg
concentrations shown to allow cell survival and migration in rat cerebellar slice culture for
several days (Mancini et al., 2009) and blood concentrations calculated in acute human
exposures in Iraq, up to 19.5 μM (Bakir et al., 1973). Untreated control slices were used to
monitor photobleaching with no MeHg treatment over the same time period. For all treatments,
image stacks were collected every 5 min.
GABAAR function and divalent cation regulation was compared in the presence and
absence of MeHg using muscimol and bicuculline, a nonspecific GABAAR agonist and
antagonist, respectively. As in treatments with MeHg alone, images were collected at two
pretreatment points, -10 and -5 min, and images were collected every 5 min thereafter for 40
min. For combination treatments, continuous perfusion with 20 μM MeHg started at t=0 min and
slices were treated with four 60 sec pulses of either 100 μM muscimol or 10 μM bicuculline in
ACSF. These treatments repeated addition of the same agent 30 sec before the 10, 20, 30 and 40
min time points to allow the perfusate to reach the slice in time for imaging. Bicuculline and
muscimol controls were performed in which these agents were pulsed in the absence of MeHg.
The 20 μM concentration of MeHg was used in combination treatments to allow for treatment
2+

comparison to an increase in [Ca ]i in all layers.

55

d. Image analysis and statistics
Cytotoxicity assay images were processed using Leica Confocal Software Lite (ver. 2.61,
Leica Microsystems Heidelberg, GmbH). Histological layers were visualized and categorized
into EGL, ML or IGL based on fluo4 staining of CGCs (Figure 2.1A). CGCs were identified as
spherical cells up to 10 μm in diameter. The EGL, 10 to 30 μm thick, extended from the pial
edges of the slice to the origin of dendritic processes (Figures 2.1, 2.2). The ML, 60 to 80 μm
thick, contained diffuse fluo4-stained PC dendrites and linear glial processes. CGCs along the
EGL/ML line were considered EGL cells. A single layer of PCs divides the ML and IGL, but
since PCs in this treatment do not typically appear to fluoresce with fluo4 treatment, the ML/IGL
boundary was defined as the end of dendrite staining and the start of densely-packed fluorescent
cells. Cells falling on the ML/IGL boundary were excluded from analysis because Purkinje cell
processes and other fluorescent processes frequently overlapped with CGCs there. The IGL
extended 60 to 80 μm from this line before an observable drop in fluo4 staining occurred,
consistent with the edge of cerebellar white matter.
For cytotoxicity assays, regions of interest were made around each visible layer in 40x
magnification image stacks, divided by visible boundaries described above (Figure 2.1).
Histograms were taken for each layer region of interest, with the threshold for positive staining
set at 10% of maximum pixel intensity. From each histogram, the sum of all pixels positive for
ethD-1 was divided by the layer’s total pixel area to produce ethD-1 pixel intensities, expressed
as a percent. Each layer’s pixel intensities from two 40x image stacks per slice were averaged to
produce mean pixel intensities per animal, and then averaged per treatment. With only one
confocal platform to image slices, the +1 h MeHg-treated slices and the +1 h ACSF-treated slices
had to come from different animals, so an n of 1 represents all treatments from the same litter.

56

Figure 2.2. 40x grayscale examples of cerebellar slices before (-10) and after (40 min) treatment
with MeHg. Images show projection of maximum fluo4 fluorescence by pixel. A Fluorescence
does not change over the course of imaging in control slices. B MeHg increased fluorescence in
all layers, with the signal from many cells saturating the detector. Scale bars are 60 μm in all
images. Surfaces and boundaries are indicated as with Figure 2.1. White arrows indicate
examples of cells with extinguished fluorescence by the end of treatment.

57

For time courses, image stacks were processed using ImageJ (Rasband, 2011,
imagej.nih.gov/ij/). Background was subtracted from each image using the subtract background
tool with a 50 pixel rolling-ball radius. Resulting images were divided into time points and
projected to two dimensional images by summing the pixels in the z-dimension. These summed
image series were aligned over time using the “StackReg” plugin (Thévenaz et al., 1998) and
cropped to eliminate unused space, typically removing 10-12 pixels in each dimension. CGCs
were chosen from all three layers, using the elliptical brush tool to select approximately 10 μm
circles around distinguishable CGCs. Mean pixel intensities (Ft) from these CGCs as circular
regions of interest were collected using the “Multi Measure” plugin (Rasband et al., 2006). Cells
were classified into two categories. Cells that visibly lysed, visibly detached from the tissue, or
lost fluo4 fluorescence between time points and did not recover were tallied as extinguished, as
they either lost cell membrane integrity or detached from the tissue between time points.
Significant fluorescence loss typical for extinguishing events was determined by finding
negative change in fluorescence between time points (ΔF) significantly greater than the mean
change of the cells in the same layer (P<0.05). Typical extinguishing events displayed >10%
fluorescence loss, and were checked visually in the original images before discarding. All nonextinguishing mean pixel intensities were tracked for the time course. Pixel intensities were
2

normalized to the average of two pretreatment intensities (F/F0). Cell density (cells/(100 μm) )
was determined from total cell counts, live and extinguished, corrected for layer areas using
additional polygonal regions of interest in ImageJ. Tallies for extinguished cells were compared
to total cells counted in each layer.
For cytotoxicity assays, cell densities and extinguished cell counts, paired mean
comparisons were made in GraphPad Prism (v4.03, GraphPad Software, Inc., La Jolla, CA). For

58

Ca

2+

time course experiments, overall effects and interactions were determined using IBM SPSS

Statistics (SPSS) (IBM, Somers, NY). Assays were analyzed using repeated measures ANOVA
with layers and time as within-subjects factors (Steel and Torrie, 1960). Because of different
numbers of replicates in different groups (n=3 to 6), adjusted means were used for statistical
comparisons. No treatment group included enough separate replicates of males and females to
compare the two groups. Additionally, each treatment group was assessed separately for
significant changes in F/F0 over time and between layers. Post-hoc comparisons were made in
SPSS for time and layer comparisons and GraphPad Prism for treatment comparisons.
Bonferroni corrections were used for multiple treatment comparisons. All effects, interactions
and comparisons were considered significant for P<0.05.

59

Figure 2.3. Relative fluo4 fluorescence intensity (F/F0) in CGCs by histological layer in slices
from neonatal rat cerebellum exposed to concentrations of MeHg continuously for 30 min.
Control MeHg concentrations did not change fluorescence significantly. Continuous perfusion of
ACSF with 0 μM (B, n=5), 1 μM (?, n=3), 5 μM (>, n=5), 10 μM (Ë, n=6) or 20 μM MeHg (C,
n=4) were compared in A, the EGL, B, the ML, and C, the IGL. Significant differences are
indicated from untreated control (*) and between 20 μM and 10 μM treatments (#) at the same
time point. Intensity values are expressed in arbitrary units (mean ± SEM).

60

61

2

Figure 2.4. Cell density per (100 μm) for CGCs monitored for fluo4 fluorescence intensity, in
each layer (EGL, ML and IGL), for tissue treated with different concentrations of MeHg.
Fluorescent CGCs were more dense in the EGL and IGL layers than in the ML. At higher MeHg
concentrations, fluorescent CGCs in the EGL were more dense than in IGL, but at lower
concentrations, density was higher in the IGL.
62

Figure 2.4 (cont’d) Within the EGL, fluorescent cell density increased significantly with greater
concentrations of continuous MeHg treatment, non-overlapping letters are significantly different
comparisons (P<0.05). Paired comparisons were made between each concentration of MeHg by
layer. Significantly different pairs within each layer are indicated by different letters (P<0.05). n
for each treatment is the same as in Figure 2.3.

63

Figure 2.5. Fluorescent cell loss “extinguished” counts from slices during treatment with several
concentrations of MeHg over 30 min. Cells extinguished when they lost significant fluo4
fluorescence between time points. Fluo4 fluorescence loss was occasionally associated with
visible lysis of cells, so presumably all fluorescence loss is associated with cell death.
Comparisons were made between each concentration of MeHg and untreated control, no
comparisons were significantly different (P>0.05).
64

D) RESULTS
a. Visualization of layers and cytotoxicity
Fluo4 labeled mainly small (5-10 μm diameter), round somata matching the size and
shape of CGCs (Figures 2.1, 2.2). Larger cells were not typically labeled visibly with fluo4,
similar to what was seen previously (Yuan and Atchison, 2007). Some processes were visible
and were used to delineate boundaries of the ML (Figure 2.1A). These processes include linear
structures consistent with Bergmann glia and branching structures consistent with PC dendrites.
Occasionally, cells visible at the boundary between the IGL and ML reacted fluorescently during
treatment, usually around the addition of MeHg. These cells may be Bergmann glia or other
interneurons present in this region that can be loaded with application of fluo4 (Hoogland et al.,
2011). They were sufficiently different in shape and size so as not to be confused with CGCs.
Fluo4-labeled CGCs were located in all three layers of the cerebellum. In the EGL, CGCs were
concentrated at the pial surface and inner folds of lobes, but additional CGCs were located at the
boundary between the EGL and ML. Most fluo4-labeled CGCs were situated in the IGL, evenly
distributed throughout the slice thickness (Figure 2.1). Comparatively few fluo4-labeled CGCs
were located in the ML.
EthD-1 did not stain any fluo4-labeled cells, indicating that only viable cells accumulated
the Ca

2+

indicator fluo4 (Figure 2.1A). In the absence of MeHg, EthD-1 intensity was 15.8% in

the IGL, 13.8% in the ML and 16.1% in the EGL at the post-incubation time point matching the
start of time course imaging, so ethidium staining was similar in all three layers (Figure 2.1B).
This presumably reflects cell damage due to slicing and the incubation process. Slices imaged
after one additional h of untreated incubation, corresponding to the end of a time course
treatment, were not different in ethD-1 mean pixel intensity, indicating there is not a significant

65

change in slice viability within the time frame used. In slices incubated for 3 h in fluorophore
plus 20 μM MeHg for 1 h, roughly corresponding to slices incubated in MeHg for a time course
series, there was an increase in ethD-1 staining in the EGL when compared to untreated slices
after 3 h of incubation, but it was not different from ethD-1 staining with one additional h of
fluorophore incubation. Thus MeHg appears to exacerbate cell death induced by slicing in the
EGL.
b. Acute MeHg slice treatment
2+

Relative pixel fluo4 fluorescence intensity (F/F0), a measure of changes to [Ca ]i, was
unaltered over 40 min in slices in the absence MeHg treatment. Changes in mean F/F0 in cells
did not exceed 5% between any two consecutive time points. Thus, none of the observed effects
were due to photobleaching of fluo4 during confocal imaging (Figures 2.2, 2.3).
2+

Effects of MeHg on [Ca ]i in CGCs were dependent upon concentration and differed by
histological layer. In all treatments with MeHg, the greatest effect on fluorescence occurred in
the EGL, and decreased in magnitude in the deeper layers (Figure 2.4). In slices treated with
MeHg, all regions appeared to increase in fluorescence over the treatment period, including the
processes distributed through the ML (Figure 2.2).
After a 40 min exposure to 20 μM MeHg, fluo4 F/F0 was significantly increased in CGCs
in all three layers, with a significantly larger effect in the EGL compared to deeper laminae
(Figures 2.2, 2.3). This MeHg concentration increased mean F/F0 up to 263% of baseline in the
EGL, 186% in the ML and 173% in the IGL. At 10 μM MeHg, CGC F/F0 increased to 192%
baseline in the EGL, 142% in the ML and 164% in the IGL. In the EGL and ML, 20 μM MeHg
increased F/F0 significantly more than 10 μM MeHg. Perfusion with 1 or 5 μM MeHg produced
no changes compared to untreated control in any layer over this interval of recording. Changes in

66

fluorescence were not linear in this time period. The largest changes in fluorescence occurred
before the 20 min time point. These changes included a 43% increase in fluorescence relative to
F0 in the EGL between 10 and 15 min time points when treated with 20 μM MeHg.
Fluorescent changes rely heavily on the number of cells being tracked, and it appeared
that MeHg treatment increased the number of cells that were visible during treatment. This was
reflected in the density of cells that could be counted in layers. In untreated slices, 37 " 2 CGCs
2

per (100 μm) were visualized in the IGL over the full course of treatment, 5 " 1 in the ML and
16 " 2 in the EGL (Figure 2.4). Cells in the IGL were more densely distributed than in the other
two histological layers. There was an increase in the density of CGCs visible in slices treated
with 10 or 20 μM MeHg, most notably in the EGL, where visible CGCs increased to a density
2

of up to 63 " 7 per (100 μm) (Figure 2.4). Many of the cells tracked in these MeHg treatments
were not readily visible at pretreatment time points, but could be tracked due to the registration
of images. While increased cell count in MeHg treatments meant comparative difficulty tracking
of CGCs in low and no-MeHg treatments with low baseline fluorescence, with the careful setup
of the slices there were always enough visible CGCs for tracking in all slices and layers.
Some CGCs appeared to extinguish fluorescence during 50 min of imaging. These CGCs
lost fluorescence at a rate that was an outlier to other cells in their layer and were thus excluded
from fluorescence intensity measurements made up 17"7% of total CGCs (Figure 2.5). These
extinguished cells were occasionally accompanied by visible lysis or detachment from
surrounding tissue, but evidence of these mechanisms is lost between imaging time points for
most of the extinguished cells. Because there was no change in percent of dead cells in control
treatments in the previous cytotoxicity assay, this percent of cell loss indicates a baseline loss of
cells during experiments due to imaging and perfusion.

67

Figure 2.6. Effect of bicuculline and MeHg on relative fluo4 fluorescence intensity (F/F0) of
CGCs by histological layer. Untreated controls (B, n=5) and 20 μM MeHg alone (C, n=5) are the
same as in Figure 3. Bicuculline pulses (arrows, 10 μM) with 20 μM MeHg ( , n=4) were
compared to bicuculline alone ( , n=3) and MeHg alone. Significant difference are indicated
from bicuculline control (*) and from MeHg alone (#) at that time point, P<0.05. Intensity values
are expressed in arbitrary units (mean ± SEM).

68

69

2

Figure 2.7. Cell density (per (100 μm) ) for CGCs monitored for fluo4 fluorescence intensity, in
each layer (EGL, ML and IGL) and density for the total tissue, for tissue treated with two
concentrations of MeHg (0 and 20 μM) and two concentrations of bicuculline (0 and 10 μM).
CGCs were more densely distributed in the EGL and IGL layers than in the ML. Paired
comparisons were made between each concentration of MeHg by layer. Significantly different
pairs within each layer are indicated by different letters (P<0.05). Values are mean ± SEM, for
the same n as described in Figure 2.6.
70

Figure 2.8. Percentage of “extinguished” cells in experiments comparing MeHg and bicuculline
treatment. Comparisons were made between 20 and 0 μM MeHg treatments, between bicuculline
alone and 0 μM MeHg, between bicuculline alone and bicuculline combined with 20 μM MeHg,
and between bicuculline combined with MeHg and 20 μM alone. Significantly different pairs
within each layer are indicated by different letters (P<0.05). Values are mean ± SEM, for the
same n as described in Figure 2.6.
71

c. Muscimol and bicuculline treatment with MeHg
Both the GABAAR agonist muscimol (100 μM) and the antagonist bicuculline (10 μM)
partially decreased the increase in fluo4 fluorescence caused by MeHg on CGCs over time, with
similar effects from both agents. Repeated pulses of muscimol or bicuculline alone did not
significantly alter F/F0 in any layer (Figures 2.6, 2.9). Muscimol pulses combined with MeHg
treatment increased F/F0 up to 208% of baseline in the EGL, 148% of baseline in the ML and
141% of baseline in the IGL, but this effect size was significantly less than that caused by MeHg
treatment alone in each layer by the end of treatment (Figure 2.6). Bicuculline pulses combined
with MeHg increased F/F0 to 211% of baseline in the EGL, 149% of baseline in the ML, and
157% of baseline in the IGL; these were significantly less than MeHg alone in the EGL and ML
(Figure 2.9). In the IGL, where the effect of MeHg alone was smaller than in the other layers,
Bicuculline was not significantly effective in reducing the increase in F/F0 caused MeHg, but
delayed the onset of significant fluorescence increase from 15 to 35 min of MeHg treatment,
similar to its effect in other layers.
As was seen with the MeHg treatments alone, MeHg treatments in combination with
either muscimol or bicuculline also had larger fluorescent cell density, particularly in the EGL,
but neither agent on its own increased these counts (Figures 2.7, 2.10). Bicuculline in
combination with MeHg was indistinguishable from MeHg alone in terms of fluorescent cell
density. Muscimol in combination with MeHg increased the density of trackable fluorescent
cells in the EGL when compared to controls, but decreased the density when compared to MeHg
alone. In addition the combination increased fluorescent cell density in the IGL enough to be
different from untreated tissue, but not from either MeHg or muscimol treatment alone.
Bicuculline did not change extinguished cell counts from controls (Figure 2.8). While muscimol

72

alone did not change extinguished cell counts, pulses of muscimol in combination with 20 μM
MeHg increased extinguished cell counts compared to both control and muscimol alone, but only
in the ML and IGL layers (Figure 2.11).

73

Figure 2.9. Effect of muscimol and MeHg on relative fluo4 fluorescence intensity (F/F0) of
CGCs by histological layer. Untreated controls (B, n=5) and 20 μM MeHg alone (C, n=5) are the
same as in Figure 3. Muscimol pulses (arrows, 100 μM) with 20 μM MeHg ( , n=4) were
compared to muscimol alone ( , n=3) and MeHg alone. Significant difference are indicated
from muscimol control (*) and from MeHg alone (#) at that time point, P<0.05. Intensity values
are expressed in arbitrary units.

74

75

2

Figure 2.10. Cell density (per (100 μm) ) for CGCs monitored for fluo4 fluorescence intensity,
in each layer (EGL, ML and IGL) and density for the total tissue, for tissue treated with two
concentrations of MeHg (0 and 20 μM) and two concentrations of muscimol (0 and 100 μM).
CGCs were more densely distributed in the EGL and IGL layers than in the ML. Paired
comparisons were made between each concentration of MeHg by layer. Significantly different
pairs within each layer are indicated by different letters (P<0.05). Values are mean ± SEM with n
the same as described in Figure 2.9.
76

Figure 2.11. Percentage of “extinguished” cells in experiments comparing MeHg and muscimol
treatment. Comparisons were made between 20 and 0 μM MeHg treatments, between muscimol
alone and 0 μM MeHg, between muscimol alone and muscimol combined with 20 μM MeHg,
and between muscimol combined with MeHg and 20 μM alone. Significantly different pairs
within each layer are indicated by non-overlapping letters (P<0.05). Values are mean ± SEM
with n the same as described in Figure 2.9.
77

E) DISCUSSION
The results show that 1) MeHg application triggers rapid and sustained increases in CGC
2+

[Ca ]i, 2) these changes are variable based on the developmental context of the CGCs in slices,
2+

and 3) the disruption of [Ca ]i is regulated in part by activity of GABAARs. The concentrations
of MeHg used did not decrease cell viability during the time of exposure compared to slicing
2+

alone, as indicated by the cytotoxicity assay, though the large increases in [Ca ]i could affect
future viability. Slice culture allowed for conditions similar to cell culture and a basic
concentration/response assessment for this tissue and treatment. Both 10 and 20 μM MeHg
caused intense increases in fluorescence while 5 and 1 μM did not in the time course studied.
Importantly, we showed differences in response between CGCs at different developmental stages
2+

of migration. The basic disruption of [Ca ]i shown here is consistent with previous
measurements showing these increases are an initial mechanism of MeHg toxicity in CGCs
(Marty and Atchison, 1998, Limke et al., 2003, Edwards et al., 2005). We further showed that
2+

modulators of GABAARs are capable of altering increases in [Ca ]i rapidly as well, effectively
decreasing and delaying the MeHg-induced increases.
The effects of MeHg on CGCs during migration can clearly be attenuated by GABAAR
modulation, indicating that the receptor can be affected in any stage of development, and
supporting a role for GABAARs in immediate mechanisms of MeHg toxicity. GABAAR
modulation by either agent does not abolish the effects of MeHg, indicating there are likely other
mechanisms occurring simultaneously. This is not surprising as other effects, such as those on
VGCCs and muscarinic receptors, are also likely to occur rapidly (Sirois and Atchison, 2000;
Limke et al., 2004). Moreover, muscimol appears to be acting similarly to bicuculline over the
2+

longer time courses, in that both reduce the effect of MeHg on [Ca ]i.

78

There are multiple possible mechanisms by which we could see similar actions of
muscimol and bicuculline. If both are acting as competitive antagonists to MeHg, then both
would appear to decrease the effect of MeHg, but the two agents act at slightly different sites and
the muscimol site activates even if MeHg is bound (Fonfria et al., 2001). Those binding studies
assumed an incubation with the agents first. In this study MeHg is applied first, so competition
might be different. The reduction in effect by both an agonist and an antagonist could also be due
to the nature of most agonists, in that the initial pulse increases conductance transiently, but is
followed by lingering desensitization, so we only observe the effects of overall antagonism.
Whatever the mechanism, there are a few subtle differences in this study between bicuculline
and muscimol. First, the combination of bicuculline and MeHg was indistinguishable from
MeHg alone in causing changes in visible density of fluorescent cells, whereas the combination
with muscimol decreased this density in the EGL but increased it in the IGL. Second, the
combination of MeHg and muscimol pulses increased cytotoxicity as measured by extinguished
2+

cells in this study where bicuculline did not. If muscimol transiently increases [Ca ]i that is not
detectable by these imaging parameters, it is possible that it contributes to MeHg cytotoxicity
2+

despite lowering the observed overall [Ca ]i in most cells.
There are not big enough differences in effects between the different layers with
muscimol or bicuculline treatments as compared to MeHg to suggest that one type of agent is
more effective in one layer than another. As the surface expression of receptor subunits such as
α6 changes over development, the agents either function similarly regardless of the expression
profiles of the CGCs, or specific reactions are lost in the variability of the populations. It would
be possible to treat slices with more specific GABAAR modulators, such as those targeting the

79

α6 subunit, but the variability of the receptors and the population of CGCs would complicate
further results.
2+

A tightly regulated amount of free [Ca ]i, is critical to driving migration and maturation
of CGCs (Komuro and Rakic, 1992, Kumada and Komuro, 2004). As with all neurons, a
2+

sustained increase or inability to buffer [Ca ]i can lead to activation of cell death pathways,
including caspase activation, cytoskeletal breakdown and opening of the mitochondrial transition
pore (Fonfria et al., 2002, Limke and Atchison, 2002, Limke et al., 2003, Roda et al., 2008).
2+

Critical cycling of increased [Ca ]i in CGC migration is another potential susceptibility to
2+

MeHg. Continuous increases in [Ca ]i from the addition of caffeine or thimerosal, another
mercury-based compound, disrupt migration of CGCs (Kumada and Komuro 2004, Komuro and
2+

Kumada 2005). Normal physiological changes in [Ca ]i, as would be seen using
electrophysiological current recordings are small, transient, and occur too rapidly to be captured
2+

by this confocal microscopy setup. Even a temporary increase in [Ca ]i could have profound
effects on cell survival.
2+

2+

In these MeHg treatments, the [Ca ]i levels in CGCs increased dramatically, but [Ca ]i
in PCs and other cells present in the slices did not label to any significant degree. This difference
2+

in cell body labeling reflects cell susceptibility and underscores the importance of [Ca ]i in
2+

studying MeHg toxicity. Since the largest [Ca ]i changes in CGCs for many treatments were at
the end of recording, the concentration may continue to rise and may eventually fall before cell
death, as has been seen in previous studies (Yukun and Atchison 2007). This may occur even for
those levels of MeHg which did not reach significance in these experiments. Extending the
treatment and monitoring might allow such increases to be seen, but could be complicated by the
slow export of fluo4 or the continued extinguishing of fluorescence by dead or dying cells.

80

Clinical manifestations of MeHg poisoning and the susceptibility of human fetuses point
strongly to the migration of CGCs in the cerebellum as being a particularly sensitive target for
MeHg toxicity. We categorized CGCs into those that were still dividing, those that were
migrating and those that had reached their destination based on their location within the
cerebellum. In MeHg treatments, fluorescence changes were distinct in the different layers; the
more mature CGCs in the ML and IGL reacted less to MeHg than did migrating CGCs as
indicated by changes in their F/F0. Regardless of the mechanisms, this indicates that immature
2+

CGCs experience a larger sustained increase in [Ca ]i as an early exposure event than more
mature cells.
Even if early free Ca

2+

is sequestered, lasting disruption of cell signaling is possible.

Precise signaling is particularly important for the migration of CGCs; fewer CGCs reaching their
target layer and maturing means fewer circuits made. The CGC and PC circuit is the major
internal cerebellar neuronal circuit; since many CGCs normally synapse on PCs, their signal can
be integrated over a population of the cells. If the population of CGCs is reduced, there is both a
larger load placed on any individual CGC as well as poorer integration of signal. This can have
lasting consequences long after the initial exposure, including motor coordination and motor
memory deficits related to the overall function of the cerebellum. Since MeHg is persistent in
brain tissue, if it is present in the brain during development it may continue to be toxic beyond
the most susceptible developmental period.
While some exploration of the mechanisms of acute MeHg toxicity focuses on various
VGCCs or reactive oxygen species production, the results presented here indicate additional
2+

acute effects via GABAARs on [Ca ]i. GABA is the major inhibitory neurotransmitter in the
central nervous system, and has several roles in cerebellar circuitry, so evidence of the

81

involvement of GABAARs in the toxicity of MeHg must be taken into account in the study of the
toxicant. The results presented here support initial electrophysiological findings (Yuan and
Atchison, 2003, 2007), binding studies (Corda, et al., 1981, Fonfría, et al., 2001), and other
toxicological studies (Basu, et al., 2010) in indicating a role here for GABAARs and might
explain known cellular and developmental susceptibilities to MeHg. Though this work focuses
on only acute effects over less than 1 h of MeHg treatment, GABAARs respond to stressors with
changes in surface expression, indicating that acute effects could trigger longer-term effects as
well. Whether the effects of MeHg on GABAARs are acute or chronic, the sensitive migration
and maturation of CGCs is affected, contributing to the susceptibility of these cells at a crucial
point in development.

82

CHAPTER THREE

ACUTE METHYLMERCURY TREATMENT INCREASES FLUO4 FLUORESCENCE
IN NEONATAL MOUSE CEREBELLAR SLICES AND IS DEPENDENT ON GABAAR
α6 SUBUNIT EXPRESSION, MIGRATIONAL STAGE AND GABAAR MODULATION

83

A) ABSTRACT

Developing CGC migration is a primary target of MeHg neurotoxicity.. GABAAR subunit
composition may contribute to cell-specific and developmental susceptibilities to MeHg. We
examined acute effects of MeHg on cerebellar slices from mice lacking the α6 GABAAR
subunit. This subunit exists exclusively in mature CGCs and regulates their tonic inhibition. It
coexists in CGCs with the abundant α1 subunit, which is not involved in tonic inhibition, nor
expressed only in CGCs. Changes in fluo4 fluorescence induced by MeHg were compared in
acutely prepared slices from PND 8-12 knockout (-/-) and wild type (+/+) mice using fluorescent
confocal microscopy. Migrating CGCs were imaged in developmental layers (EGL and ML),
along with post-migratory CGCs in the IGL. MeHg (20 μM) increased CGC fluorescence over
time in the EGL and ML. In (-/-) cerebellar slices, MeHg increased fluorescence significantly at
15 min and continued up to 201% over baseline in the EGL at 55 min, the end of treatment.
Significant fluorescent changes were delayed and decreased in the IGL, reaching significance at
25 min and continuing up to 145% over baseline. Though MeHg treatment appeared to increase
fluorescence in (+/+) slices, the effect was only significant in the EGL within 55 min. Pulses of
the GABAAR agonist muscimol (100 μM) to tissue perfused with MeHg reduced fluorescence
changes by 75% compared to MeHg alone in the IGL of (-/-) mice and delayed the onset of
significant increases. The GABAAR antagonist bicuculline (10 µM) also reduced and delayed
the effect of MeHg. While the α6 subunit is not expressed fully in CGCs until maturation, there
still is a pronounced effect of MeHg on [Ca2+]i in immature CGCs of (-/-) mice. Thus α6
expression may establish inhibition in developing CGCs, perhaps by controlling

84

2+

neurotransmission or expression of another subunit which contributes to regulation of [Ca ]i in
migrating CGCs.

85

B) INTRODUCTION

MeHg is a neurotoxicant that targets the cerebellum, and more specifically the CGCs in
2+

the cerebellar cortex. In Chapter Two, we demonstrated that MeHg disruption of [Ca ]i was
dependent on CGC development, and that alterations of GABAARs by the agonist muscimol and
2+

antagonist bicuculline reduced this disruption in [Ca ]i. Further characterization was needed,
particularly to model gene-environment interactions in the toxicology of MeHg.

There are several potentially important differences in gene expression that could
2+

contribute to the susceptibility of CGCs to MeHg. Many of these differences relate to [Ca ]i
signaling and have been determined by comparison to the relatively resistant PCs in the
cerebellum onto which CGCs synapse. Differences in expression of intracellular Ca

2+

binding

proteins such as calretinin, calbindin and parvalbumin lead to differences in function and
excitability (Gall, et al., 2003, Bearzatto, et al., 2006, Bornschein, et al., 2013). Similarly, CGCs
and PCs have differences in Ca

2+

channels and receptors, such as VGCCs, AMPA receptors and

others (Chung, et al., 2000, Huang, et al., 2007, Shevtsova and Leitch, 2012).

One receptor expression difference is perhaps most notable of all and that is the
difference in GABAARs. Mature CGCs express mostly GABAARs containing α6 subunits,
whereas PCs express α1 exclusively (Nusser, et al., 1995). The two receptors are functionally
distinct, with α1-containing receptors functioning as typical inhibitory synaptic receptors while
α6-containing receptors reside extrasynaptically and respond to ambient GABA. GABAAR gene
interactions are particularly important in development, during which CGCs go through several
stages of expression before maturation. GABAAR subunit expression goes through several stages

86

as well during development in CGCs. The α6 subunit is not expressed until maturation in CGCs,
and is dependent on the change in chloride gradient (Ortinski, et al., 2004, Takayama and Inoue,
2004, Bortone and Polleux, 2009). Interestingly, differences in subunit expression appear to lead
to differences in response to MeHg, both functionally and in cell viability (Yuan and Atchison,
2003, 2007; Herden, et al., 2008) These individual differences indicate larger functional
differences based on a single variable in gene expression, so it makes GABAARs a strong
candidate to explore through gene knockout.

The GABAAR α6 knockout (-/-) mouse was developed alongside other subunit knockouts
as a potential model for epilepsy, but was found to not have a distinct epilepsy phenotype; in fact
it had no apparent phenotype apart from alterations to pharmacological stimuli associated with
the α6 subunit (Homanics, et al., 1997, Makela, et al., 1997, Nusser, et al., 1999). With no
defects in normal function, it appears that GABAAR activity is compensated, most likely by a
change in potassium current (Brickley, et al., 2001). This potassium “leak” current is able to
maintain CGCs in a less excited state, similar to the hyperpolarized state of normal CGCs, but is
relatively depolarized compared to normal CGCs, and may require more ATP production to
maintain. Motor coordination in these mice can be disrupted treatment with diazepam, which
indicates an important role for remaining α1 and γ subunits, even if there is no indication that
their expression is changed (Korpi, et al., 1999). Despite the lack of an epileptic or otherwise
dysfunctional phenotype, this knockout could provide answers as to whether GABAAR α6
expression contributes to MeHg toxicity or if it is protective.
2+

The slice model for cerebellum has proven valuable in determining the [Ca ]i changes
induced by MeHg application in developing rat, and has already indicated some involvement of

87

GABAARs, so it is the approach used here with developing mice. GABAAR α6 (-/-) and (+/+)
mice are used from the same colony to determine the effect of genotype, and the GABAAR
agonist muscimol and antagonist bicuculline are used once again to determine the action of
GABAAR signaling within the context of MeHg treatment. These experiments will determine
effects of GABAAR expression and signaling to the toxicity of MeHg in developing mice, and in
combination with other sections of this dissertation, will contribute to an overall characterization
2+

of GABAARs, [Ca ]i, and MeHg in multiple susceptible developmental stages.

88

C) MATERIALS AND METHODS

a. Chemicals and solutions

Materials and solutions for slicing of cerebella and Ca

2+

imaging with fluo4 are

described in Chapter Two. Unless otherwise stated, methods for isolating and imaging cells in
slice were identical for mouse brains as for those used for rat brains.

b. Animal care and genotyping of transgenic animals

Animal care was performed according to NIH guidelines for experimental use and was
approved in advance by the Michigan State University Institutional Animal Care and Use
tm1Geh

Committee. GABAAR α6 (-/-) mice (B6;129-Gabra6

/J, strain #002710) and the

background strain (+/+) (B6:129/J) were originally obtained from Jackson Laboratories (Bar
Harbor, ME). Mice were maintained in a breeding colony at MSU, and fed, housed and enriched
normally according to MSU IACUC guidelines. Breeding pairs were co-housed serially over
multiple litters when needed. Multiple breeding strategies were tried at various times depending
on needed genotypes, including (-/-) x (-/-) for producing all (-/-) litters, (+/-) x (+/-) for
producing maximally mixed litters, and (-/-) x (+/-) or (-/-) x (+/+) for producing needed
breeders for successive generations. Every three generations, male (-/-) mice were back-crossed
with (+/+) background strain females obtained directly from Jackson Laboratories.

Genotyping was performed according to the developer of the strain (Homanics, et al.,
1997), with some changes to accommodate the use of neonates. Genotyping was performed
between PND5 and PND21, depending on need for experiments, as mouse pups of this age have

89

less innervation of the tail. Pups were temporarily, individually removed from the box and sterile
2% (w/v) lidocaine hydrochloride jelly (Akorn, Inc., Lake Forest, IL) was applied to the tip of
the tail to anesthetize the area. Tails were disinfected with a 1:100 dilution of Nolvasan (Zoetis,
Florham Park, NJ). Less than 2 mm of the end of the tail was removed and immediately placed in
200 μl of lysis buffer solution containing 10 mM Tris, 50 mM NaCl, 25 mM EDTA , 0.5% SDS
and 5 µM Proteinase K (Sigma-Aldrich, Milwaukee, WI) in 1 ml tubes, labeled and kept on ice
while all tissue was collected. Mouse pups younger than PND18 were coded using colored nontoxic permanent markers on skin of the stomach, feet and base of tail, and color was reapplied as
needed until PND18. Pups PND18 or older were coded by ear punch. Immediately after tissue
collection and coding, the remaining tail was dipped in styptic powder (Kwik-Stop, ARC
Laboratories, Atlanta, GA) to stop bleeding, and the pup was returned to the litter.

Genotyping tissue was kept in lysis buffer overnight in a 50°C dry heating block.
Dissolved tissue was precipitated in a solution of 3 M potassium acetate and 2 M acetic acid,
vortexed and centrifuged (16 g, 10 min) to form a pellet and supernatant containing DNA.
Supernatant was transferred to matching labeled tubes containing 200 μl isopropyl alcohol (F.T.
Baker) and mixed once by inversion. These tubes were centrifuged for 15 min (16 g) to form
another pellet containing the DNA. The supernatant was removed carefully and replaced with a
wash of -20°C 100% ethanol. These tubes were centrifuged for 25 min (16 g). Supernatant wash
was removed and tubes were allowed to dry. Pellets were dissolved in the tubes using 50 μl of
diethylpyrocarbonate (DEPC)-treated water to eliminate RNase activity (Life Technologies).

Isolated DNA was amplified by polymerase chain reaction (PCR). PCR reagents were
mixed according to manufacturers specifications, such as with RedTaq polymerase (Sigma-

90

Aldrich), including concentrations of primers, deoxyribonucleotide triphosphate (dNTP)
monomers and aliquots of DNA and PCR buffer. Forward and reverse primers for the transgenic
neomycin insert and GABAAR α6 subunit gene were obtained according to Jackson lab
protocols (neomycin generic primers, 5’-CTTGGGTGGAGAGGCTATTC-3’ and 5’AGGTGAGATGACAGGAGATC-3’, GABAAR α6 subunit primers, 5’GACCACCTTAAGCATCAG-3’ and 5’-TTTGACTTTGCCACAGGG-3’) and diluted to 1/50
of the original concentration before mixing. Additional MgCl2 was added to the PCR mixture to
bring it to a final concentration of 25 mM, even though many PCR buffers contained a lower
concentration already. PCR was run on a GeneAmpPCR 9700 96-well PCR machine (PerkinElmer, Waltham, MA) using settings according to the PCR buffer manufacturers specifications
for denaturation, amplification, and annealing steps. An initial 1 min denaturation step at 94°C
was added if not included.

Amplified bands of neomycin and GABAAR α6 subunit were resolved by gel
electrophoresis. PCR product was dyed with <1 μl xylene cyanol dye and placed in wells of a
1.5% agarose Tris-acetic acid-EDTA (TAE) gel. Neomycin, indicating presence of at least one
allele containing the neomycin cassette interrupting GABAAR α6, resolved at 280 bp, while
uninterrupted GABAAR α6 resolved at 240 bp. Heterozygote (+/-) appeared as two bands.
Genotypes for initial experiments, those comparing (-/-) and (+/+) littermates for reaction to
MeHg treatment in slices, were blinded from ABB by others in the laboratory, by labeling the
final results as (+/-), “group 1” or “group 2”, with no indication as to whether group 1 or 2 was (/-). The initial coding of groups was maintained through the blinding period. (+/-) animals were
not coded as they were not used in initial experiments, but were a natural result of (+/-) crosses.

91

Group coding results were decoded after initial graphical results were obtained, and were not
performed on further experiments.

c. Acute slice preparation

Slices were prepared as in Chapter Two, with the following changes. Male and female
mouse pups (M= 12, F= 14, PND 8-12) were euthanized by rapid decapitation to provide brains.
After slicing in ice cold slicing solution, slices were incubated in ACSF warmed to 37°C for the
first 30 min, followed by 30 min at room temperature (23-25°C) before fluorophore incubation
and for the remainder of incubations. As both rat and mouse cerebellum appeared to have
sufficient rates of uptake of fluorophores for imaging, concentrations used and periods of
incubation were identical. If possible based on genotyping, matching treatments were performed
on one (-/-) and one (+/+) mouse from the same litter within 24 h.

d. Confocal microscopy

Slice imaging was performed as described in Chapter Two, with some changes for
efficiency. As cells in mouse cerebellum looked identical to cells from rat cerebellum and had
similar fluorescence, laser power and acquisition was kept the same. Fluo4 fluorescence was
always collected at 500-535 nm, the setting used previously for fluo4/ethD-1 experiments in rats.
Between-image depth was set to exactly 3 μm for all image stacks to maintain resolution in the
Z-dimension. Imaging timing was automated and extended, so that two pretreatment imaging
scans were taken and two sets of 5 images collected automatically during treatment at 5 min
intervals unless interrupted by obvious X, Y or Z shifts in the slice. This allowed for a longer
treatment time, extending to 50 min of treatment compared to 40 min shown in Chapter Two.

92

Only four treatments of the type used in Chapter Two were continued in these
experiments. Two to three of these treatments were applied, in random order on the same day, to
slices from the same brain, so maintain closer consistency between slices. Continuous 20 μM
MeHg and no MeHg controls were compared in both (-/-) and (+/+) cerebellar slices.
Additionally, 100 μM muscimol pulses and 10 μM bicuculline pulses were also studied in
combination with 20 μM MeHg, and compared to continuous MeHg alone, in both (-/-) and (+/+)
slices. Because of the extended treatment time, one additional muscimol or bicuculline pulse was
used, as compared to rat treatments, at the 50 min time point.

Image analysis and statistics

All image processing was performed in ImageJ (v 1.45s, imagej.nih.gov/ij/). A scale
factor of 300 μm / 512 pixels was applied to all images imported from Leica Confocal Software.
CGCs and histological layers were otherwise identified exactly as in Leica software, as mouse
CGCs and layers were highly similar to those in rat cerebella. Image stacks, imported as separate
image series, were concatenated to form a single image series. Backgrounds were subtracted and
image stacks summed and registered to maintain cell position as described in Chapter Two.
Histological layers were drawn with the ImageJ polygon brush tool and cells were highlighted
using the elliptical brush set as a 10 μm circle. Multiple regions of interest, either the histological
layers or multiple cells, were tracked using the ROI manager tool. Cells with extinguished
fluorescence were tallied and excluded as in Chapter Two.

Overall statistical comparisons were performed in SPSS statistics (v 20.0.0, IBM,
Armonk, NY) while paired comparisons were made in GraphPad Prism (v 4.03, La Jolla, CA).

93

For extinguished cell counts, within-subject t-tests were performed in Prism between MeHgtreated and untreated slices, and separated by layer and genotype. For time courses, the factors
considered for ANOVA analysis were genotype ((-/-) versus (+/+)), treatment, with two
treatments in each analysis (control/MeHg, MeHg/muscimol combination or MeHg/bicuculline
combination), layers (EGL, ML and IGL), and time (13 time points). Because of the paired
nature of the experiments, the treatment measure was considered a within subjects repeated
measure, as was time. Paired comparisons were made between MeHg-treated and untreated
slices from the same animal, as well as between muscimol or bicuculline combination treatments
and their MeHg-only counterparts. Paired comparisons were also made between genotypes
within each layer and treatment. Each genotype*treatment*layer mean was compared to
pretreatment time points by repeated measures ANOVA to determine change over time. All
comparisons were considered significant for P<0.05, with Bonferroni corrections as applied in
Prism for multiple comparisons.

94

Figure 3.1. Representative images of frames of interest and labeled layers from (-/-) mouse
cerebellar slices. Images are from two slices; both slices are from the same (-/-) mouse.
Detailed description of assigning boundaries is discussed in Chapter Two materials and
methods. Layers and boundaries from surface of cerebellum (solid line) include the external
granule layer (EGL), the EGL-ML interface (dotted line), the molecular layer (ML), the
Purkinje layer (dashed line) and the internal granule layer (IGL). The left images are prior to
administration of MeHg, and right images are the last image taken from respective series. The
top slice was treated with only ACSF (0 µM MeHg, untreated control) for 55 min, the bottom
slice was treated to 20 µM MeHg in ACSF for 55 min.

95

Figure 3.1 (cont’d). Untreated control had little change to fluorescence, but 20 µM treatment
increased fluorescence ~2-fold in the EGL in this example, with smaller increases in the ML and
IGL. Scale bar is equal to 50 µm. Images are summary projections of ~30 scans through the
depth of the slice, intensity values are adjusted to the value histogram of the first image. Images
are cropped to remove blank space around edges from image registration.

96

D) RESULTS

a. Visualization of layers

As with the acutely prepared rat cerebellar slices collected in Chapter Two, in mouse
cerebellar slices fluo4 labeled mostly round somata matching the characteristics of CGCs (Figure
3.1). Fluo4-fluorescent soma were from 5-10 μm in diameter, and the remainder of the
fluorescent material in the images was diffuse staining in an area matching the description of the
ML, based on distance from the pial surface. PCs were not frequently visibly fluorescent lining
the boundary between fluorescent CGCs in the IGL and the diffuse ML staining, but this lack of
PC fluorescence has been noted before in Chapter Two and in previous experiments (Yuan and
Atchison, 2007).

Many of the characteristics of neonatal mouse cerebellar slices appeared similar to rat
slices. The EGL was 10 to 40 μm thick and contained a high density of CGC precursors at the
pial surface and lining the boundary with the ML, while the ML was 60-100 μm thick and
contained relatively few CGCs. CGCs packed the IGL densely in all dimensions. At the slice
thickness used, very few fluorescent CGCs overlapped within the same space in projections
along the Z axis. In all treatments, the visible area contained 38 ± 4 CGCs in IGL, 11 ± 1 in the
ML, and 14 ± 2 in the EGL, fluorescent CGC counts did not differ between MeHg treatment and
untreated controls (P>0.05). Also, fluorescent cell counts did not differ between (+/+) and (-/-)
genotypes.

97

Figure 3.2. Time course of effects of MeHg on fluo4 fluorescence in cerebellar slices from
neonatal α6 (+/+) and (-/-) mice. Mean fluo4 fluorescence intensity was compared between 0
(untreated control, Con) and 20 µM MeHg (MeHg) in (-/-) (circles, n=5) and (+/+) (squares,
n=6) mice over 55 min of treatment. Each of the three developmental layers are shown
separately. MeHg is applied continuously for the duration of the experiment starting at t=0 (20
µM, black; 0 µM, white).
98

Figure 3.2 (cont’d). MeHg treatment causes a rapid increase in fluo4 fluorescence in all layers
and genotypes, with the largest change in the (-/-) EGL. * represents time points that are
significantly greater than pretreatment (P<0.05) and ** represents time points that are
significantly greater than pretreatment as well as greater than the paired, untreated time point
(P<0.05).

99

b. Acute MeHg paired treatment in (+/+) and (-/-) mice

As a measure of photobleaching, relative fluo4 fluorescence (F/F0) based on mean pixel
intensities of selected cells did not change from pretreatment time points when slices from both
genotypes were treated with ACSF alone (untreated control) (Figure 3.2). Relative fluorescence
in these slices, when corrected as described for extinguishing cells, did not change by more than
5% between adjacent time points, indicating no positive or negative change in F/F0 from
imaging at the settings used. Overall changes in untreated controls over 55 min were at most
down 10 ± 7 % in the EGL of (-/-) slices and up 4 ± 1 % in the IGL of (+/+) slices.

Overall, MeHg caused an increase in F/F0 over time, but was dependent on both
genotype and layer (Figure 3.2). In (-/-) mouse cerebellar slices, MeHg treatment increased F/F0
up to 201% of baseline in the EGL, 141% in the ML and 145% in the IGL. When comparing to
initial untreated time points, these results were significant in the EGL from 15 min on, in the
ML from 25 min on and in the IGL from 35 min on. When comparing MeHg treatment to timematched points from untreated slices from the same animals, MeHg induced significantly greater
fluorescence from 35 min on in the EGL, and 50 min on in the ML. This comparison was not
significant in the IGL, despite the differences from pretreatment time points. In (+/+) littermates,
MeHg treatment increased F/F0 significantly compared to pretreatment only in the EGL, where it
increased the F/F0 up to 150% baseline, and was significantly elevated from time matched
untreated control at 50 min on.

As with rat treatments, the change in F/F0 was not linear with MeHg treatment. Large

100

Figure 3.3. Extinguished cell percentages as found in paired mouse cerebellar slices treated with
0 (ACSF) and 20 µM MeHg (MeHg) from (-/-) (white bars) and (+/+) (black bars) mice.
Extinguished cells are cells that rapidly and significantly lose fluo4 fluorescence intensity as
compared to the normal variation during treatment. These cells are removed from time course
intensity traces and tallied. In these MeHg treatments, only the CGCs in (+/+) IGL significantly
extinguish as compared to their untreated paired slices (*, P<0.05)
101

increases in fluorescence occurred between 5 and 20 min time points, but continued to steadily
increase in all MeHg treatments and layers except the (+/+) IGL, where it began to decrease
towards the end of treatment, after 35 min.

All treatment groups had some CGCs that showed extinguishing fluorescence, a rapid
decline of F/F0 associated with loss of cell integrity and viability that was significantly greater
than expected by the variance of the cells (Figure 3.3). This loss of cells was generally less than
20% of CGCs found. While there was a general increase in this loss in the different cell layers
with MeHg treatment, the change was not significant except in the IGL of (+/+) slices. Despite
this interaction specific to the (+/+) IGL, there was no significant difference between collective
genotypes or layers.

c. Muscimol and bicuculline paired treatments with MeHg

As with results in Chapter Two in rats, pulses of both muscimol (100 μM) and
bicuculline (10 μM) decreased the effects of MeHg on fluo4 fluorescence, but in this case effects
were heavily dependent on layer. These experiments were performed on a different set of
animals than the ACSF and MeHg treatment pairings, with a different set of 20 μM MeHg
treatment used as a baseline. The effect of MeHg in the two experiments was not significantly
different. In these paired (-/-) mice, MeHg treatment caused a significant elevation in F/F0
compared to pretreatment as of 10 min in the EGL, ML and IGL layers and remained elevated.
In the corresponding (+/+) mice, MeHg treatment reached F/F0 significance at 30 min in the
EGL, 25 min in the IGL, where F/F0 remained elevated, while in the ML, the only significant

102

point was 35 min. Neither muscimol nor bicuculline treatments significantly affected fluorescent
cell count or extinguishing cell counts (data not shown).

Muscimol pulses decreased the response to MeHg treatment in all layers and both
genotypes (Figures 3.4, 3.5). In all cases but one, muscimol treatment maintained F/F0 close to
baseline, similar to ACSF treatment alone. In IGL of (+/+) mice, F/F0 does eventually increase
significantly, but only at the 50 min time point. In direct comparisons between the muscimol and
MeHg treatments at corresponding time points, there is no significant difference.

As with muscimol, bicuculline dramatically reduced the effect of MeHg, this reduction
was dependent on layer and genotype. In (-/-) slices, bicuculline eliminated the effect of MeHg
in the IGL, and reduced the effect in the EGL and ML to the point where significant effects were
delayed until 35 and 40 min respectively. In (+/+) slices, bicuculline eliminated the effect of
MeHg in the ML, but had no significant effect in the EGL and IGL, where it was only able to
delay significant changes in F/F0 by 5 min.

103

Figure 3.4. Time course of effects of MeHg and muscimol on fluo4 fluorescence in cerebellar
slices from neonatal α6 (-/-) mice. Mean fluo4 fluorescence intensity was compared between 20
µM MeHg (MeHg, white circles) and 20 μM MeHg with 100 μM muscimol pulses (Muscimol,
black circles) in (-/-) mice over 55 min of treatment (± SEM, n=5). Each of the three
developmental layers are shown separately. MeHg is applied continuously for the duration of the
experiment starting at t=0 and muscimol is pulsed into the feed for 60 s at 9.5, 19.5, 29.5, 39.5
and 49.5 min (arrows).
104

Figure 3.4 (cont’d). Time course of effects of MeHg and muscimol on fluo4 fluorescence in
cerebellar slices from neonatal α6 (-/-) mice. Mean fluo4 fluorescence intensity was compared
between 20 µM MeHg (MeHg, white circles) and 20 μM MeHg with 100 μM muscimol pulses
(Muscimol, black circles) in (-/-) mice over 55 min of treatment (± SEM, n=5). Each of the three
developmental layers are shown separately. MeHg is applied continuously for the duration of the
experiment starting at t=0 and muscimol is pulsed into the feed for 60 s at 9.5, 19.5, 29.5, 39.5
and 49.5 min (arrows).

105

Figure 3.5. Time course of effects of MeHg and muscimol on fluo4 fluorescence in cerebellar
slices from neonatal α6 (+/+) mice. Mean fluo4 fluorescence intensity was compared between 20
µM MeHg (MeHg, white circles) and 20 μM MeHg with 100 μM muscimol pulses (Muscimol,
black circles) in (+/+) mice over 55 min of treatment (± SEM, n=5). Each of the three
developmental layers are shown separately. MeHg is applied continuously for the duration of the
experiment starting at t=0 and muscimol is pulsed into the feed for 60 s at 9.5, 19.5, 29.5, 39.5
and 49.5 min (arrows).
106

Figure 3.5 (cont’d) MeHg alone causes a rapid increase in fluo4 fluorescence in all layers in a
manner similar to previous paired experiments in (+/+) mice. Muscimol pulses reduced this
effect in all layers and kept fluo4 fluorescence near baseline for the duration of experiments in
the EGL and ML. * represents time points that are significantly greater than pretreatment
(P<0.05).

107

Figure 3.6. Time course of effects of MeHg and bicuculline on fluo4 fluorescence in cerebellar
slices from neonatal α6 (-/-) mice. Mean fluo4 fluorescence intensity was compared between 20
µM MeHg (MeHg, white circles) and 20 μM MeHg with 10 μM bicuculline pulses (Bicuculline,
black circles) in (-/-) mice over 55 min of treatment (± SEM, n=5). Each of the three
developmental layers are shown separately. MeHg is applied continuously for the duration of the
experiment starting at t=0 and bicuculline is pulsed into the feed for 60 s at 9.5, 19.5, 29.5, 39.5
and 49.5 min (arrows).
108

Figure 3.6 (cont’d). MeHg alone causes a rapid increase in fluo4 fluorescence in all layers in a
manner similar to previous paired experiments in (-/-) mice. Bicuculline pulses reduced this
effect immediately and kept fluo4 fluorescence near baseline for the duration of experiments in
the IGL. * represents time points that are significantly greater than pretreatment (P<0.05).

109

Figure 3.7. Time course of effects of MeHg and bicuculline on fluo4 fluorescence in cerebellar
slices from neonatal α6 (+/+) mice. Mean fluo4 fluorescence intensity was compared between 20
µM MeHg (MeHg, white circles) and 20 μM MeHg with 10 μM bicuculline pulses (Bicuculline,
black circles) in (+/+) mice over 55 min of treatment (± SEM, n=6). Each of the three
developmental layers are shown separately. MeHg is applied continuously for the duration of the
experiment starting at t=0 and bicuculline is pulsed into the feed for 60 s at 9.5, 19.5, 29.5, 39.5
and 49.5 min (arrows).
110

Figure 3.7 (cont’d). MeHg alone causes an increase in fluo4 fluorescence in all layers in a
manner similar to previous paired experiments in (+/+) mice. Bicuculline did not reduce the
effect in any layers, but also did not increase fluo4 fluorescence in the ML compared to
pretreatment. * represents time points that are significantly greater than pretreatment (P<0.05).

111

E) DISCUSSION
2+

Results here further support a role for GABAARs in MeHg-induced increases in [Ca ]i
in developing CGCs. This occurs both through the differences apparent with the knockout of one
of the receptor subunits and through the actions of a GABAAR agonist and antagonist. MeHg
2+

causes rapid increases in [Ca ]i that reach significance in these slices at concentrations that
match estimated acute blood concentrations in the Iraq exposure event for ataxia (Bakir, et al.,
1974). Permanent genetic knockout of the GABAAR α6 subunit, the major subunit expressed in
2+

adult CGCs, enhanced [Ca ]i responses in these slices. Intriguingly, this occurred in the layers
that should not express the subunit normally. Muscimol and bicuculline each decrease this CGC
response to MeHg, consistent with rat neonatal slices, with some important differences between
their responses dependent on development of the CGCs and genotype. This approach works with
minimal changes to the protocol from rat slice preparation, and allows for extended imaging of
the slices. Some direct comparisons between treatments are difficult to make with the large
variability of relative fluorescent responses in these mouse CGC populations, but by pairing
treatments on slices from the same brain many important comparisons could still be made.

The effects of 20 µM MeHg on relative fluo4 fluorescence intensity in mouse cerebellar
slices, while perhaps not as large as those seen in rat cerebellar slices for the same concentration,
was still enough to cause significant increases within 50 min. Changes in intensity were greater
in the EGL than other layers, consistent with previous findings, while in the ML of (+/+) mice
MeHg barely registered any changes. With further time points analyzed than in previous
2+

experiments, it is still clear that the changes in [Ca ]i are maintained, as long as CGCs are
divided into those that are alive and those that have lost integrity and thus fluorescence, which

112

after 65 min of continuous recording under the microscope only appears to be about 20% of
visible CGCs. Live cell monitoring could be extended with this protocol in mind, but it is not
clear how many more cells could be lost without losing the integrity of the overall signal.
Extinguished cells have not been correlated conclusively with cell death in these experiments,
but like the rat CGCs in Chapter Two, extinguished fluorescent was occasionally recorded with
visible lysis of CGCs.

Notably, MeHg effects were genotype-dependent. In all three layers of (-/-) mice, MeHg
almost immediately increased fluo4 fluorescence in the whole population. Though variability
was enough to make the responses statistically similar, in (-/-) cerebellum MeHg caused an
additional 30% increase in fluo4 fluorescence in CGCs in the EGL compared to (+/+). This
clearly indicates a role for the α6 subunit in protecting CGCs at all stages of development. The
actual mechanisms for this are not clear from these experiments. The inhibitory action of α6containing GABAAR activation could conceivably contribute to protection in the mature CGCs
of the IGL and possibly those CGCs migrating in the ML. In those layers, knockout of α6
produces less inhibition and thus less protection, but this does not explain the larger effect size in
the EGL, where CGCs have been shown not to express the subunit yet. Expression profiles of
normal developing CGCs clearly show a change in α6 subunit expression at the cell surface as
well as other subunit expression over the course of CGC migration (Gault and Siegel, 1997,
Takayama and Inoue, 2004). This switch to α6 subunit is prominent in the IGL, not the EGL or
ML . If α6 expression alone is responsible for protection of CGCs and neither (+/+) nor (-/-)
mice express detectable α6 in CGCs migrating through the EGL and ML, then the response to
MeHg should not be expected to be different in these populations, yet they are in the EGL.

113

Absent some role for undetectable minor amounts of α6 expression in migrating and
maturing normal CGCs, the enhanced fluorescence intensity increase in developing CGCs
visualized here indicates an indirect mechanism for the loss of protection in (-/-) CGCs. Other
neurons and cells in the cerebellum do not express α6, so they presumably would not be directly
affected by the change in genotype either. The most direct connection between mature CGCs
expressing α6 in the IGL and the developing CGCs in the EGL is through the parallel fiber axons
projecting from mature CGCs through the ML. These fibers would be found near the pial surface
where immature CGCs are present. It is possible that some signaling goes on between mature
CGC axons and the immature CGCs at the surface. This is an intriguing proposal, in that it also
suggests that toxic signaling, such as excessive release of glutamate from MeHg-exposed axon
2+

terminals, could contribute just as much to the [Ca ]i increases in immature CGCs as direct
actions. This glutamate neurotransmission-dependent mechanism could be explored further in
this model. If maturing CGCs are able to release neurotransmitter or otherwise signal immature
CGCs, it is possible altered signaling from mature (-/-) CGCs would result in loss of protective
mechanisms or enhanced toxicity pathways. Conceivably, some difference in normal signaling
could alter receptor expression patterns of the immature CGCs, such as VGCC or GABAAR
subunit expression, and thus their reaction to new signals. Alterations in migrating cells’ reaction
to signals and movement is critical for migration, such as GABA signaling from interneurons
(Farrant and Brickley, 2003). MeHg-induced effects on CGC migration also suggest changes in
gene expression and signal response could be involved here (Mancini, et al., 2011, Fahrion, et
al., 2012). Even though gene expression changes are relatively long-term changes and would not
be noticed in acute MeHg application, the long-term status of a knockout could prime the cells
for MeHg susceptibility. A surface expression approach could be performed in these cells to

114

determine changes in expression of receptors such as GABAARs and VGCCs. Gene expression
profiling would require selection of populations from the different stages of development, which
would be difficult but not impossible with currently available approaches.

Muscimol and bicuculline application with MeHg further indicate subtle differences in
the genotype. As with application in rat cerebellum slices, both the agonist and the antagonist of
2+

GABAARs cause overall decreases in the [Ca ]i changes induced by MeHg. Comparisons
between the treatments, layers and genotypes are complicated by several interactions. With the
additional time points, it appears that the reductions in MeHg response from these GABAAR
agents may be delays instead, both muscimol and bicuculline produced effects that eventually
reached the same significance above baseline as their MeHg only pairings. This is an expected
2+

result; MeHg increases [Ca ]i through many different mechanisms. Hence, blocking one
pathway can at best delay changes. While many of the genotype, layer and treatment effects are
complicated by interactions, it appears that muscimol and bicuculline treatment may be more
effective in (-/-) mice than in (+/+), in that muscimol is only ineffective in IGL of (+/+) and
bicuculline decreases or delays the response to MeHg in all three layers in (-/-). Previous
characterization of GABAAR α6 (-/-) mice indicates a 50% decrease in GABAARs at the cell
surface with no compensation from other subunit types (Jones, et al., 1997, Nusser, et al., 1999).
Our results indicate GABAAR activation is more effective in (-/-) slices. Muscimol and
bicuculline could be more effective because of the elimination of a large population of their
targets. In GABAAR α6 (-/-) mice, increases in the effects of BZs, which normally target α1
receptors with γ subunits, have been attributed to elimination of the populations of α6γ receptors
(Korpi, et al., 1999). This also indicates that better knowledge of the surface expression and

115

subunit composition of GABAARs would be useful for further exploration of moldulator effects.
With some further knowledge of the GABAAR subunit expression profiles present in slices,
agents tailored to development stage-specific targets could be used, rather than the agonist and
antagonist used here. Subunit-specific GABAAR modulators, including neurosteroids, are being
characterized that could provide approaches to further explore these differences within this
model (Mennerick, et al., 2005, Simeone, et al., 2006, Khom, et al., 2010). It is possible that
both changes to GABAAR expression or function in immature CGCs as well as some change to
the excitotoxic potential of mature CGCs could be important here in combination, so it is worth
considering both in further study.

These experiments further characterized the populations of neurons most susceptible to
MeHg toxicity. They include mature, relatively resistant CGCs and susceptible, developing
CGCs in the same frame, along with PCs and other cells that did not visibly respond to MeHg.
These experiments also characterized GABAAR involvement in the mechanisms of MeHg
toxicity through the knockout of an important subunit and the application of GABAAR agonist
2+

and antagonist. These characterizations support results of previous live cell Ca

imaging and

electrophysiology experiments in showing that GABAAR involvement has a variable effect on
2+

MeHg toxicity that occurs through changes in [Ca ]i (Yuan and Atchison, 2007, Herden, et al.,
2008). Further characterization is possible and needed to determine how much MeHg toxicity is
direct on CGCs versus a larger, circuit-dependent disruption as may be the case with glutamate
excitotoxicity or expression changes between (+/+) and (-/-) mice.

116

CHAPTER FOUR

COMPARATIVE Ca

2+

EFFECTS OF CHRONIC EXPOSURE OF ADULT MICE TO
MeHg OR ISRADIPINE

117

A) ABSTRACT

Studies on the effects of MeHg on animals have focused on developmental exposures,
with few examining the effects of chronic adult exposures, despite the potential of lifetime
exposure for exacerbating genetic conditions or the declines of aging. As part of a
comprehensive study of molecular, behavioral, and pathological effects of chronic MeHg
exposure in adult BalbC mice, as well as the effects of a L-type VGCC blocker isradipine, we
2+

imaged [Ca ]i dynamics in cerebellar slices. We sought to determine if chronic MeHg or
isradipine treatment altered Ca2+-dependent mechanisms, and whether they interact chronically,
as acutely isradipine and other VGCC blockers reduce the effects of MeHg. Two imaging
approaches were used, confocal imaging of fluo4-loaded slices to examine acute application of
MeHg and test for changes to sensitivity, and high-speed epifluorescence microscopy of fura-2loaded slices to test for changes in Ca

2+

signaling pathways and receptors through applications

of AMPA, NMDA and depolarization by high [KCl] ACSF. In slices from 6 month mice, MeHg
2+

application caused a rapid and significant (P<0.05) increase in [Ca ]i, up to 2-fold compared to
baseline. No differences were noted between chronic treatment groups. In slices from the same
animals, AMPA caused significant, greater than 2-fold increases in relative fura-2 signal, and
2+

thus [Ca ]i in chronic control mice. Other chronic treatment groups had similar responses at this
age, though the combination MeHg and isradipine treatment returned to baseline quickly after
2+

the pulse. NMDA was not able to alter [Ca ]i in CGCs as a population at 6 months, and KCl
2+

treatment was only able to increase [Ca ]i significantly and temporarily in the control group. At
12 months, slices responded similarly to MeHg, but due to the addition of a paired, acute ACSF
(0 MeHg) treatment, there was a subtle, but significant, decrease in the effect in chronic

118

isradipine-treated groups. In slices from this age group, AMPA induced larger and more
2+

permanent increases in [Ca ]i and all groups had significant increases from KCl treatment as
well, but NMDA responses were small and variable among cells, so there was no concerted
change. Additional treatments were added at 12 months, including no change in perfusion from
ACSF wash, which surprisingly still caused individual and population-wide depolarization.
2+

These additional treatments also included muscimol, which increased [Ca ]i, but only in mice
treated chronically with MeHg, and bicuculline, which did not induce any population-wide
2+

changes. Together these results indicate widespread changes in Ca -dependent mechanisms
induced by chronic MeHg and isradipine treatment in mice, and will be integrated with many
other results from this project.

119

B) INTRODUCTION

As established in previous chapters, MeHg is a neurotoxicant of special concern to
developing cerebellum. What has not been emphasized so far is that lifetime exposure to MeHg
and the possible interaction of aging with the toxicant are also concerns. Because of the long
latency periods of MeHg toxicity and because of its low clearance from the body, it is clear that
even if people are able to avoid MeHg during development and childhood, many populations
maintain long, low concentrations of MeHg in their CNS as adults. The actions of MeHg have
mostly been studied in high exposure conditions, such as after the Japan and Iraq exposures, not
chronic exposures.

While developmental exposure is of the utmost concern for neurotoxicology, MeHg
exposure occurs at low levels over a lifetime, as indicated by the frequent metrics of hair and
blood total Hg concentration used in epidemiological and experimental studies (Myers, et al.,
2009, Yaginuma-Sakirai, et al., 2012, Bellanger, et al., 2013). As with several other similar lowconsumption, persistent, ubiquitous neurotoxicants, MeHg has been hypothesized as a risk factor
for neurodegenerative diseases such as Parkinson’s Disease or psychiatric disorders (Uchino, et
al., 1995, Weiss, et al., 2002, Petersen, et al., 2008, Weiss, 2010, Yorifuji, et al., 2011). It is
unclear if the increased risk is due to lifetime exposure or exposure during another susceptible
period, such as development (Fox et al., 2012). Even leaving out a potential for developmental
sources for neurodegenerative disorders in aging, aging brains may be susceptible on their own.
As discussed in Chapter One, the susceptibility of aging brains may be due to an increased level
2+

of baseline [Ca ]i in aging neurons which is functionally relevant to normal declines in aging,

120

but also poses a susceptible set of mechanisms on which MeHg can act (Toescu and
Vreugdenhil, 2010).
2+

Since we have already seen that MeHg rapidly increases [Ca ]i in susceptible neonatal
neurons, it was clear that aged animals represented another susceptible population in which the
effects of acute MeHg exposure might be seen. Experimental results using models treated over a
lifetime have been scarce, particularly postnatal exposures, and have shown divergent results.
Some indicate few or no neurological results (Yasutake, et al., 1997). Some indicate moderate
increases in dysfunction in geriatric rodents compared to adults (Paletz, et al., 2007, Vitalone, et
al., 2010). Finally, some indicate robust differences in neurological decline, particularly if
lifetime exposure includes some perinatal exposure (Stern, et al., 2001, Weiss et al., 2005, Heath,
et al., 2010). These results emphasize the need for comprehensive study of lifetime exposure to
MeHg, including adult-only exposure regimens, to determine if adult and aging brains are
themselves vulnerable to chronic MeHg.

A chronic study also provides the opportunity to examine potentially protective
interventions. Ca

2+

blockers reduce and delay the effects of MeHg both in vitro and in vivo

(Sakamoto, et al., 1996, Sirois and Atchison, 2000, Peng, et al., 2002). One Ca

2+

blocker,

isradipine, is a selective antagonist for Cav1.3 (L-type) VGCCs and is typically used chronically
to treat hypertension and angina, but also has protective effects against stroke-induced
neurodegeneration in the brain (Rüegg and Nelson, 1989, Sauter and Rudin, 1990, 1991;
Campbell, et al., 1997). Changes to L-type VGCC expression in the hippocampus are associated
with aging as well as neurodegenerative diseases such as Alzheimer’s Disease (Coon, et al.,
1999, Veng, et al., 2003, Brewer, et al., 2007). In the cerebellum, L-type VGCCs are associated

121

with PCs, and could alter expression during aging as well (Araki, et al., 1993, Chung, et al.,
2001). Isradipine has not been examined for interactions with MeHg in chronic treatments, but
could be protective in several brain regions.

As part of a comprehensive, highly-collaborative project to study the cellular basis for
MeHg neurotoxicity and isradipine intervention in adult and chronic exposures, we used our
2+

established slice imaging protocols to characterize [Ca ]i responses in chronically-treated mice.
This includes confocal imaging of acute MeHg exposure, both to establish a baseline to compare
to neonatal susceptible populations and to determine if there are sufficient chronic effects on the
mechanisms to produce a different response entirely, positive or negative. Treatments also
2+

include high-speed imaging of responses to various [Ca ]i stimuli to test other baselines of
aging and effects of chronic treatments. High-speed imaging was chosen for many of these
treatments because of the better temporal resolution for effects and the ability to use UV-based
2+

ratiometric fluorophores. Unlike MeHg, with its continual increases in [Ca ]i over hours, the
effects of most agents occur on the order of seconds, so better temporal resolution was needed.
2+

AMPA facilitates a fast [Ca ]i response through the AMPA receptors, a typical receptor
2+

important for Ca

2+

influx for signaling or for reestablishing [Ca ] in organelles. AMPA

receptors are also highly regulated, typically by internalization when overly active, so chronic
treatments could conceivably regulate their expression. NMDA activates the NMDA receptor, an
ionotropic and voltage-dependent Ca

2+

channel that is normally responsive to glutamate. NMDA

receptors could be involved in MeHg toxicity through glutamate release, and could also be
involved with changes in aging (Glazner and Mattson, 2000, Faro, et al., 2002, Fang, et al., 2005,
Petroni, et al., 2011). Interestingly, chronic MeHg exposure is also associated with NMDA

122

expression (Basu, et al., 2007). NMDA responses are generally slower, in part due to voltage
sensitive block of the pore by a Mg

2+

cells. That depolarization induces Ca

ion. High [KCl] solutions are able to depolarize excitable

2+

influx through VGCCs, so this treatment can be used to

characterize both excitability of cells and the function of their VGCCs. The effects of GABAAR
modulation were suggested for these experiments as well, but due to constraints on the use of
slices, these treatments were only able to be performed in the 12 month chronically-treated
groups. Muscimol and bicuculline were used in experiments that also tested baseline activity of
these slices through treatment with the ACSF that is used throughout incubations and washes.

123

C) MATERIALS AND METHODS

a. Chemicals and solutions

Chemicals and solutions described in Chapter Two were also used in these experiments
as formulated previously, including ACSF, MeHgCl, sucrose slicing solution, and fluo4. Some
handling of these solutions were changed and those changes are described fully here, along with
alternate solutions and additions. For incubation and treatment solutions, Fura-2 AM, TPEN,
AMPA, NMDA, pluronic acid, muscimol and bicuculline methobromide were obtained from
Sigma-Aldrich (Milwaukee, WI). Probenecid was obtained from Molecular Probes (Eugene,
OR). All incubation solution chemicals were aliquoted in double-distilled H2O or DMSO
according to manufacturer instructions and stored at -20 °C.

Sucrose slicing solution was made as described previously. Stock (5X) solutions of
slicing solution minus MgCl2 and CaCl2 were kept at 4 °C, and then diluted to 1X with MgCl2
and CaCl2 added within 48 h of use. About 25% of this prepared solution was frozen in cubes at
least 8 h at -20 °C, while the remainder was stored at 4 °C. 8 cubes, approximately 200 ml, were
crushed in a blender just before animal dissection, and supplemented with 200 ml 4 °C solution.
This solution was immediately oxygenated by bubbling a 95%/5% O2/CO2 gas mixture through
2+

it. All oxygenated solutions were bubbled with this gas mixture. ACSF and Mg -free ACSF
were kept at room temperature (23-25 °C) for at least an hour prior to dissection on the day of
treatments to avoid temperature changes during imaging.

124

In order for the NMDA receptor to respond to NMDA application alone, slices for
2+

microfluorimetry needed to incubate in Mg -free ACSF. This ASCF was made simply by not
adding MgCl2 when mixing the solution, the difference in osmolarity and pH was not significant.
10 ml solutions each of 25 μM AMPA, 50 μM NMDA plus 10μM glycine, 100 μM muscimol
2+

and 10 μM bicuculline, were made in Mg -free ACSF on the day of each relevant experiment.
Additionally, 40 mM KCl ACSF (containing 87.5 mM NaCl; 40 mM KCl; 1.25 mM KH2PO4;
26 mM NaHCO3; 25 mM D-glucose; 1 mM MgCl2; and 2 mM CaCl2), was prepared weekly.
b. Animal care and treatments

All animal use and care was based on NIH laboratory animal use guidelines and was
approved by the MSU Institutional Animal Care and Use Committee. Male BalbC mice were
obtained from Harlan (Frederick, MD) at approximately three weeks of age, housed in pairs with
standard feed (Teklad 18% Protein Global Rodent Diet) obtained from Harlan, water and
accommodations and given hard plastic houses and nesting material for enrichment. Isradipine
was obtained from VWR (Chicago, IL) and sent directly to Harlan. Isradipine was mixed into
standard feed at 2 ppm to make a custom diet, and an inert yellow dye was added to the
isradipine diet to distinguish it from the control diet. Sixteen batches of eight mice each were
obtained in a staggered configuration to allow for weekly sets of several different experimental
approaches. These mice were acclimated until three months of age, when they were divided into
four treatment pairs and coded by ear punch. The first pair received standard feed and water. The
second pair received standard feed and water supplemented with 6.25 ppm MeHg. The third pair
received isradipine-supplemented feed and standard water. The fourth pair received both the
MeHg-supplemented water and the isradipine-supplemented feed. One animal from each pair

125

was chosen randomly for 6 month collection, while the other was retained for 12 month
collection. Animals were coded by numbers and colors, to allow for treatment blinding from
those working on the tissue and behavioral tests.

All animals were monitored daily for any health problems. Animals were weighed and
checked for hind-limb crossing caused by weakness and paralysis, three times a week (M, W, F).
Additionally, for one week of training and acclimation and again three times every other week
(T, W, Th), animals were given balance testing on a variable-speed rotating cylinder (IITC, Inc.,
Woodland Hills, CA) to test cerebellar function. All bedding, waste and food was handled as
potential hazardous waste due to the presence of MeHg.

c. Tissue collection

Each week one animal from each of the four treatment groups was selected for dissection
and tissue collection. Because of the time required for incubations and experiments performed on
the same day, no more than 2 animals were used on a given day. One mouse was removed and
sealed in a transparent glass container with 2 ml of 99% isofluorane (Abbott laboratories,
Chicago, IL) absorbed into a cotton swab as anesthetic until righting reflex was lost and
breathing slowed. The mouse was removed and secured to a cold dissecting platform, with
additional isofluorane applied over the nose to maintain unconsciousness. Secured mice were
checked for blink reflex, without which they were determined sufficiently unconscious for
cardiac perfusion. The chest cavity was opened and the heart pierced in the left ventricle with a
27 gauge needle attached to a 50 ml syringe loaded with blended frozen slicing solution. The
right atrium was cut, and resulting blood collected from the chest cavity by pipette, placed in a

126

labeled tube and frozen in liquid nitrogen to be removed to -80 °C for storage. The entire 50 ml
blended slicing solution was pumped through the left ventricle to replace blood throughout the
body. This was immediately followed by removal of the head. Cardiac perfusion and
decapitation were considered to be time of death.

Various tissues, including spinal cord, liver, kidneys, and EDL muscle were taken as
needed for other studies and are not included in this section. The head was removed to a bath of
blended slicing solution, and allowed at least 1 min to cool. The head was removed from the
bath, the skin was removed and the skull removed as described in Chapter Two, with special
attention added to removing the forebrain and brainstem without damage. Forebrain was given to
another researcher for other studies. Brainstem and trimmed excess hemisphere tissue from
cerebellum was frozen in liquid nitrogen and stored at -80°C. Sagittal slices of 180 μm thickness
were taken from the cerebellum in a bath of blended slicing solution. This thickness was chosen
because it is a size preferable for electrophysiological studies, which were performed in addition
to the imaging studies described here. Slices were immediately removed to a mesh surface in a
container of oxygenated ACSF at 37°C. These slices were incubated at this warm temperature
for 30 min from the last addition of a slice, and then the container was removed to room
temperature for further incubations. The time between animal death and heated slice incubations
was kept minimal, and generally did not exceed 10 min.

Immediately upon completion of 37°C incubation, one slice was chosen for either highspeed or confocal imaging. When possible, imaging platform use was staggered in order to
complete more treatments before starting a new animal. Because time between slicing and use

127

could potentially lead to degradation, slicing times and start times for imaging were tracked to
compare to markers of slice quality taken from images and notes.

d. High-speed imaging

A slice was transferred to an opaque 35mm petri dish containing a solution of 2 mL
ACSF supplemented with 5 µM fura-2 AM, 2.5 mM probenecid, 5 µM TPEN, and 0.002% (w/v)
pluronic acid. Pluronic acid and probenecid were used in order to facilitate uptake and inhibit
expulsion of fluorophore, respectively. TPEN is a selective intracellular Zn

2+

chelator which

2+

eliminates any Zn -dependent components of fluorescence. This solution was prepared during
37 °C incubation. This incubation bath was transported to a 5% CO2, 37 °C incubator for 45
min. This is the only step in which ACSF is not oxygenated, to reduce the production of
insensitive fura-2 intermediates.

During incubation, the slice imaging chamber on an inverted Nikon Eclipse TE 2000-U
2+

microscope was perfused with oxygenated Mg -free ACSF with a tissue anchor in the chamber.
This was to prepare the chamber and soak the anchor so there was less movement during
treatment. After incubation, the slice was placed in the chamber and the anchor was laid on top
2+

to secure it. Fresh Mg -free ACSF was perfused across the chamber via a perfusion pump
running at about 2 mL/min. The slice was allowed at least 10 min to wash. During this time a
frame of interest was visualized using the microscope. The frame was selected by finding a
region with a reasonable number (>20) of CGCs, recognizable by size and shape, fluorescent and
in focus at this excitation line, and no obvious tears or other damage. Light from a Polychrome V
light source (Till Photonics) was shone through a custom fura-2 dichroic mirror (Chroma,

128

71000A Fura2 C59742) and a 40x oil immersion objective (Nikon S Fluor 40x/1.3 oil DIC
H/N2) onto the slice from below. A special low fluorescence oil was used for contact between
the objective and chamber (FF, Cargill, Cedar Grove, NJ). For frame acquisition, a 380 nm
excitation line was used.

Frame of interest acquisition was handled through software when possible to limit
continuous illumination of the slice. Image acquisition was performed using TillVision control
software (v 4.0). Live acquisition of the frame of interest was tested using live acquisition grab
settings, set to 380 nm and 500 ms intervals. This briefly switches on the Polychrome V light
source, and opens the high-speed electron-multiplying CCD camera (Ixon EM+, Andor, South
Windsor, CT) to reduce illumination on the cells compared to eyepiece illumination. Fura-2 is a
ratiometric fluorophore, with the value of emission at 500-540 nm based on the excitation at 340
2+

2+

2+

nm for Ca -bound fluorophore, and 380 nm for Ca -free fluorophore. As [Ca ]i increases, the
value from 340 nm increases and the value from 380 nm decreases. For the start of experiments,
illumination time was adjusted at 380 nm to about 75% of maximum, to allow for decreases at
this wavelength with stimulus. This illumination time was checked at 340 nm as well, which
generally started at a much lower fluorescence emission and could be adjusted further. Initial
frame snapshots were saved for both of these wavelengths.

Illumination time was adjusted in a template protocol in TillVision. Briefly, this template
kept the settings between treatments the same, apart from this adjustment to illumination time,
and automated the collection of images. This protocol specified three treatment periods, each
consisting of a set of 300 images taken at 2 s intervals (10 min total) with each time point

129

imaging twice with a rapid switch between 340 and 380 nm illuminations. A user-terminated
pause was set between each treatment period to allow for extra wash times or other adjustments.

Once the initial 10 min wash and the frame of interest and illumination was set, treatment
was started. Concentrations of AMPA, NMDA and KCl were selected based on previous Ca

2+

recordings from brainstem slices (Johnson, et al., 2011). These concentrations were optimized in
cerebellar slices from up to 18 month mice before the experiments performed here. The
2+

perfusion buffer was switched to 25 µM AMPA in Mg -free ACSF for 2 min and the start of
imaging. Once two minutes passed with the AMPA stimulus treatment, perfusion was switched
2+

back to Mg -free ACSF for the remaining 8 min of imaging as well as at least 2 min more of
wash. The second treatment was performed exactly as the first, but used 50 µM NMDA + 10 µM
glycine as treatment. The third treatment in this series was performed exactly as the first two, but
used 40mM ACSF.

In the 12 month treated animals, an additional set of slices from the same animals as the
first was taken, and an additional series was added. The concentrations of muscimol and
bicuculline used were based on concentrations that induced or blocked depolarization in
cerebellar slices, respectively (Yuan and Atchison, 1997, 2003). In this treatment series, the first
2+

treatment consisted of continuing Mg -free ACSF (i.e. no treatment), the second treatment
consisted of 100 µM muscimol, and the third 10 µM bicuculline.

Image series were exported from TillVision as tagged image files (.tif). In Image J (v
1.45s, imagej.nih.gov/ij/). Background was subtracted by a 50 pixel rolling ball subtraction and
images were registered to lessen movement between time frames. Because of the time this

130

algorithm takes to process these steps in 300 images, background subtraction and registration
were performed automatically by batch processing. Images from 6 month-old animals were
processed manually using the “region of interest (ROI) manager” tool collecting circles with 1418 pixel diameters. Using the “multi” tool in the ROI manager, mean intensity values for both
340 and 380 nm excitations from these ROIs were exported in spreadsheets. In contrast, because
of an increase in number of treatments and total experiments, 12 month images were processed
by an automated method. In this automated approach, registered image stacks for 380 nm were
opened for further processing. From the first 380 nm image, a granulometric filter (circle, 7 pixel
radius, 2 pixel step) was run to remove non-cellular fluorescence and highlight circular cells
between 7 and 9 pixels, representing CGCs (Prodanov, et al., 2006). This filter tags points
(“granules”) in the image that represent uniformly fluorescent circles of the selected size. It
produces a new image with the selected size circles added to these points with a value of 1,
several of these circles can overlap around the center of mass of a cell. More clustered points
represent more confidence a cell has been found of the correct size. CGCs typically have 5 or
more of these points in the center of mass, so a binary filter was added wherein all areas in this
filter with >5 points were considered positive (1) and all <5 points negative (0). Since some
CGCs touched or were even collected in clusters, binary watershed filtering was applied to this
image to insert borders between uniformly sized sections of clustered cells. The “analyze
particles” tool was used to define the cells in this image and export them to the “ROI manager”
tool. Once these cells were defined in ROI manager, the defined cells were applied to the
original image stack. Just as with manual tracking of ROIs, the “multi” tool was used to
determine mean intensity values from these ROIs and means were exported in spreadsheets. The
locations of all ROIs used were saved for reference.

131

Spreadsheets were imported into Excel® (Microsoft, Redmond, WA), from which
ratiometric values (R) were produced from the individual 340 and 380 nm traces. Of these traces,
only those that started with a 380 nm value greater than 340 nm were used, as a reversed 340 to
380 ratio indicates a damaged cell. Additionally, some traces displayed a sudden extinguishing
of 340 and 380 fluorescence. This was indicated by a significantly outlying negative change in
fluorescence at both wavelengths. Since this loss of fluorescence represents a burst or detached
cell similar to the loss of fluo4 fluorescence described in Chapters Two and Three, these cells
were excluded from time course analyses. Counts of reversed ratio cells and counts of
extinguished cells were compared to other potential markers of slice quality, such as time from
slicing by linear regression and chronic treatment group by ANOVA.

e. Confocal microscopy

A slice was transferred to an opaque chamber consisting of a mesh bed in a well of a 12well culture plate containing oxygenated 1x fluo4NW + probenecid as described in Chapter 3,
with the addition of 5 µM TPEN to eliminate the contribution of Zn

2+

to fluorescence in these

experiments. This incubation lasted 40 min at room temperature (25-27°C). While this
incubation ran, the imaging chamber on the microscope used in Chapter Three was prepared,
with ACSF flowing at about 1.5 mL/min via a gravity feed.

Imaging on confocal was performed as described in Chapter Three, with a few changes
described here. Briefly, only one pretreatment image scan was taken, the remainder images were
taken exactly as described in Chapter Three, with two blocks of five automated imaging scans

132

with a break in between to check for slice movement. Emission range was reset to settings
described in Chapter Two for fluo4 alone.

Due to time constraints, only 20 µM MeHg treatment was performed on the confocal
imaging platform for 6 month treated mice. An additional control, 0 µM MeHg in ACSF, was
deemed necessary partway through 12 month experiments, and performed on the last 5-7 mice
from each treatment group. These concentrations are based on the experiments conducted in
Chapters Two and Three in which MeHg treatment caused a several fold, significant increase in
fluo4 fluorescence within 50 minutes in cerebellar slices.

Image series were imported into ImageJ and concatenated into raw image series. A 50pixel rolling ball radius was used to remove background from these series. These images were
grouped and converted into sum projections. CGCs were manually drawn as 15 pixel diameter
circles and tracked using the ROI manager tool. Once ROIs were defined, the ROI manager
multi tool was used to produce a spreadsheet of the means over 11 time points.

f. Statistics

Spreadsheets were imported into Excel® to check traces. Extinguished cells were
determined by determining negative outliers for changes in fluorescence. These cells were tallied
and removed from the population of traces. Remaining traces were collected and averaged for
treatment groups. Count, standard deviations and standard errors were collected. Means for each
animal were exported to SPSS (IBM, Somers, NY) and overall statistics were calculated by
multifactorial, repeated measures ANOVA. All traces were normalized to the initial time point.

133

For high-speed imaging, no statistical software appears to be able to handle all the factors
used (300 time points, 4 chronic treatments, 6 acute treatments, 2 age groups) when determining
overall statistics, so only every tenth time point was used for calculating these statistics. Withinsubject repeated measures included time (30 time points) and acute treatment (3 levels in order,
as part of either AMPA/NMDA/KCl or ACSF/muscimol/bicuculline). Between subjects factors
included chronic MeHg treatment (control or 6.25 ppm), chronic isradipine treatment (control or
2 ppm), and treatment series (AMPA/NMDA/KCl or ACSF/muscimol/bicuculline). To reduce
error caused by differences in initial R, statistics were performed with traces that were
normalized to their initial value (R/R0). Based on significant effects and interactions, normalized
treatment traces were compared to the second time point used in statistics (20 s) to determine if a
significant change in fluorescence occurred and when.

For confocal imaging, within-subject factors included time (11 time points) and acute
treatment (20 µM MeHg or control). Between subject factors included chronic MeHg treatment
(control or 6.25 ppm) and chronic isradipine treatment (control or 2 ppm). Based on significant
effects and interactions, traces were compared to the initial time point to determine if a
significant change in fluorescence occurred and when. Paired time points, focusing at the end of
treatment, were compared between each treatment group.

134

Control (O)

Fig 4.1 A

Pre-treatment

MeHg 45 min

ML

MeHg + Isr (M/I)

2 ppm Isradipine (I) 6.25 ppm MeHg (M)

IGL
ML

IGL
ML

IGL
ML

IGL

Figure 4.1. Example frames of interest from confocal fluo4 time course imaging.

135

Figure 4.1 (cont’d). Slices were imaged through 45 min of continuous perfusion with 20 μM
MeHg in 6 month chronically-treated mice. All images are from the same batch of animals.
Images are registered, rotated and cropped for clarity; scale bar equals 50 μm. A. Slices of
cerebellum are from four 9 month old animals treated chronically for 6 months with, in order,
standard food and water (O), 6.25 ppm MeHg in drinking water (M) with standard food, 2 ppm
isradipine in food (I) with standard water, and the combination of MeHg in water and isradipine
in food (M/I). MeHg application over 45 min increases relative fluo4 fluorescence in CGCs in
the IGL by about 2-fold within 45 min, the duration of observation, in all 4 chronic treatment
groups. B. Expanded frame of interest from group O after 45 min of continuous MeHg perfusion.
Image highlights processes in the ML that become apparent with MeHg treatment (arrows),
infrequent PC fluorescence (boxes), and non CGC cells in the ML (circles).

136

RESULTS

a. Confocal Ca

2+

imaging

Acute treatments of 20 µM MeHg were applied to slices from four sets of chronically
2+

treated animals in order to assess functional differences in Ca -dependent mechanisms that
could come about by long-term exposure to MeHg and isradipine. Differences between CGCs
could come about from changes in receptor expression, trafficking or recycling, and other long2+

term differences in signaling, such as changes in baseline [Ca ]i. Animals collected at 6 month
time points only received acute MeHg treatments, so could only be compared to pretreatment
baselines, while 12 month time had two treatments where possible, one slice received MeHg
while another received only ACSF. Unlike in neonatal rodents as described in previous chapters,
CGCs were located only in the IGL (Figure 4.1 B). Many processes in the ML were visible, but
diffusely stained and difficult to distinguish (Figure 4.1 B). PCs rarely fluoresced as efficiently
as CGCs, but could be seen in some slices from all age and treatment groups. Some additional
interneurons or glia were evident in the ML or other positions, but could be distinguished from
CGCs by shape (larger than CGCs, not always circular) and position.

Acute MeHg treatment caused rapid increases in relative fluo4 fluorescence in CGCs
tracked in both 6 month and twelve month treatment groups and in all chronic treatments used.
In 6 month control animals (0 MeHg, 0 isradipine, designated O), acute MeHg increased relative
fluo4 fluorescence, up to 203 ± 13% of baseline (Figures 4.1, 4.2). In 6-month chronically
MeHg-treated mice (6.25 ppm MeHg, 0 isradipine, designated M), acute MeHg application

137

Figure 4.2. Mean relative fluo4 fluorescence intensities as measured from confocal time course
image series of cerebellum slices from 6 month chronically-treated mice. Slices from untreated
control (O, n=11, white circles), MeHg-treated (M, n=11, black circles), isradipine-treated (I,
n=9, white triangles), and combination MeHg- and isradipine-treated (M/I, n=11, black triangles)
were imaged at 5 min intervals for 50 min with one pretreatment scan and 10 images with 20 µM
MeHg flowing over the slices. All acute MeHg treatments caused about a 2-fold increase in
relative fluo4 fluorescence in CGCs (± SEM) and reached significance compared to pretreatment
by 20 min of imaging (P<0.05), but were not significantly different from one another (P>0.05).

138

Figure 4.3. Mean relative fluo4 fluorescence intensities of CGCs in slice during 20 μM MeHg
exposure over 45 min from 12 month chronically-treated mice. Traces are from acute treatments
of 0 µM MeHg (untreated controls, white circles) and 20 µM MeHg (black circles) applied in the
bath ACSF solution to cerebellum from 15 month old mice chronically treated for 12 months.
Paired comparisons were made between matching time points (*, P<0.05). A. Acute treatments
of mice receiving no chronic MeHg or isradipine (Group O). Acute MeHg treatment (n=16)
significantly increased fluorescence above no MeHg treatment (n=5). B. Acute treatments of
mice receiving chronic 6.25 ppm MeHg in water and no isradipine (Group M). Acute MeHg
treatment (n=16) significantly increased fluorescence above no MeHg treatment (n=7). C. Acute
treatments of mice receiving no MeHg and 2 ppm isradipine in diet (Group I). Acute MeHg
treatment (n=13) increased fluorescence above no MeHg treatment (n=5), but only at one
intermediate time point. D. Acute treatments of mice receiving both 6.25 ppm MeHg in water
and 2 ppm isradipine in diet (Group M/I). Acute MeHg treatment (n=17) did not significantly
increase fluorescence above no MeHg treatment (n=6) during observation.

139

Figure 4.3 (cont’d)

140

Figure 4.3 (cont’d)

141

increased relative fluorescence of 194 ± 14% of baseline. In 6-month chronically isradipinetreated mice (0 MeHg, 2 ppm isradipine, designated I), acute MeHg treatment increased
fluorescence 179 ± 15% over baseline. Finally in 6-month combination treatments (6.25 ppm
MeHg, 2 ppm isradipine, designated M/I), acute MeHg caused an increase of 192 ± 13%. All
four groups had significantly elevated fluorescence compared to baseline by the end of the
experiment. While the two isradipine-treated groups appear to have a smaller increase in
fluorescence than the other two groups, this difference was not significant.

Slices from mice treated for 12 months (the last 4-6 from each treatment group) were able
to be used for an additional treatment from each subject. This was an untreated control which
was continually perfused with ACSF for the imaging period instead of MeHg in ACSF. These
treatments indicated that, in all treatment groups, perfusion of ACSF and imaging did not in
itself change the relative fluorescence intensities of slices. After 12 months of chronic
treatments, acute MeHg treatment increased relative fluorescence 190 ± 12% in group O (Figure
4.3A), 172 ± 11% in group M (B), 174 ± 13 % in group I (C), and 148 ± 12% in group M/I (D).
While all these groups displayed smaller increases in fluorescence than their 6 month
counterparts, none were sufficiently smaller to indicate significant differences. Where the
comparisons do indicate subtle but significant differences between treatments is in time-matched
comparisons with the control treatments. Here, slices from the same brain in groups O and M
treated with acute MeHg were significantly more intense than ACSF control from 20 minutes
onward, while group I was only different at 30 min and group M/I was no different at any time
point. This indicates that chronic isradipine treatment subtly reduces the reaction of slices to
acute MeHg, regardless of the actions of chronic MeHg.

142

b. High-speed Ca

2+

imaging
2+

Slices incubated with fura-2 were given several stimuli meant to evoke Ca

changes

while imaged at high speed to determine detailed time courses of these changes and any
differences that might be apparent. Much like the acute MeHg treatment on the confocal
platform, changes to the underlying signaling pathways important for each stimuli would be seen
under one or more of these treatments. In the 6 month treatment groups, slices were exposed to
three stimuli. AMPA, the ligand of fast ligand-gated cation (Ca2+, Na+ and K+) channel, was
applied to slices first. AMPA binding to AMPA receptors causes Ca
+

2+

influx from the

+

extracellular space as well as passing Na and K . The receptor rapidly internalizes or is
trafficked to the membrane under various stimuli. NMDA and glycine were applied second in
2+

order to activate NMDA receptors, another non-selective cation channel that can pass Ca . This
2+

step normally requires a voltage change to the membrane to remove Mg
Mg

2+

from the pore, but

was removed from the treatment solutions to facilitate the ion’s removal. Third, slices in
+

this series were forced to depolarize through application of a high [K ] (40 mM) solution.

The populations of cells chosen for tracking reacted differently to each of these
treatments. Even though the use of fura-2 allows for the calculation of precise nanomolar
2+

2+

concentrations of [Ca ]i when [Zn ]i is bound by TPEN, as it is in these treatments, the
measurement of maximum fluorescence and minimum fluorescence has not been performed in
these cells, so results are provided as relative to the starting ratio of cells.

143

Figure 4.4. Mean relative fura-2 ratio values for acute 25 μM AMPA, 50 μM NMDA + 10 μM
glycine and 40 mM KCl applications measured from high-speed (0.5 Hz) time course image
series in cerebellar slices from 9 month old mice. Mice were treated chronically for 6 months
with MeHg or isradipine. Separate lines represent different chronic treatments, including no
MeHg and no isradipine (Group O, black line, n=13, representative +SEM shown), MeHg and no
isradipine (Group M, red line, n=11), isradipine and no MeHg (Group I, blue line, n=11), and
both MeHg and isradipine (Group M/I, magenta line, n=11). Treatments are in order, and
consisted of treatment for 2 min, with approximately 20 s to reach the bath, imaging for an
additional 8 min and a 2 min wash between imaging. A. AMPA treatment of slices, all four
chronic treatment groups responded to the treatment (*, P<0.05 compared to baseline). Several
stayed above baseline ratio, while group M/I returned during the imaging period. B. NMDA
treatment of slices, following AMPA. No population-wide change in fluorescence occurred. C.
KCl treatment, following NMDA. Only group O showed a significant, population-wide change
in fluorescence ratio during treatment.

144

Figure 4.4 (cont’d)

145

Figure 4.5. Mean relative fura-2 ratio values for acute 25 μM AMPA, 50 μM NMDA + 10 μM
glycine and 40 mM KCl applications measured from high-speed (0.5 Hz) time course image
series in cerebellar slices from 15 month old mice. Mice were treated chronically for 12 months
with MeHg or isradipine. Separate lines represent different chronic treatments, including no
MeHg and no isradipine (Group O, black line, n=15 representative +SEM shown), MeHg and no
isradipine (Group M, red line, n=17), isradipine and no MeHg (Group I, blue line, n=14), and
both MeHg and isradipine (Group M/I, magenta line, n=15). Treatments are in order, and
consisted of treatment for 2 min, with approximately 20 s to reach the bath, imaging for an
additional 8 min and a 2 min wash between imaging. A. AMPA treatment of slices, all four
chronic treatment groups responded to the treatment (*, P<0.05 compared to baseline). All
stayed above the baseline ratio for the duration of the imaging period. B. NMDA treatment of
slices, following AMPA. No population-wide change in fluorescence occurred. C. KCl
treatment, following NMDA. All four chronic treatment groups showed significant responses to
treatment, and all were temporary.

146

Figure 4.5 (cont’d)

147

AMPA treatment caused the largest and most rapid changes in fluorescence (Figure 4.4A),
increasing the relative fluorescence ratio value, R, by up to 226% of baseline in group O. This
occurs rapidly, over about 120 s once the agonist reaches the slice. AMPA treatment in the
chronic MeHg treatment group M appears almost identical, with a rapid rise and apparent
plateau. While not significantly different at any time points from O and M, and despite high
variability, the isradipine-treated groups (I and M/I) are closer to baseline at the end of AMPA
treatment than the other treatments. Rather than a single phase change as occurred with groups O
2+

and M, these groups appear to have multiple phases of [Ca ]i increase. Not only that, but where
group I shows both phases early, before 200 s, followed by a slow decrease, the traces in M/I are
much more protracted. All four treatment groups reach significant increases in their ratios, but
again some differences are apparent. Groups O, M and I all reach significance at about 180 s and
continue to stay above baseline. Group M/I reaches significance at about 180 s, but due to
variability loses such significance by 280 s.

While NMDA treatment, the second in the series, did cause some individual CGCs in
slices to react similarly to AMPA treatment, as a whole the CGCs picked out did not react
significantly to NMDA, even if they clearly reacted to the previous ligand (Figure 4.4B). Peaks
in fluorescence ratio only reached 148% of baseline in group I, while they reached 128% in O,
124% in M and 103% in M/I. R was not different from baseline in any of these treatments, nor
were the comparisons between them significant.

KCl depolarized many CGCs across all chronic treatment groups, however due to a high
2+

percentage of nonreactive CGCs, overall the population changes in [Ca ]i are more subtle
(Figure 4.4C). Group O saw a significant increase to 126% of baseline fluorescence as a

148

population at about 140-180 s, but this was a transient increase. Fluorescence increased steadily
as a population in groups I and M, but did not reach significance compared to initial values.
Group M/I decreased fluorescence over time, though not significantly from initial time points.

Slices from mice treated at 12 months showed many similar responses to those of the 6
month treatment group. AMPA caused the largest increases, up to 272% in group O, 252% in
group M, 273% in group I and 254% in group M/I (Figure 4.5A). While the different chronic
treatments did not differ from one another, all were above baseline for most of the imaging
acquisition, Group O at about 80 s, Group M at about 140 s, Group I at about 160 s and Group
M/I at about 140 s. It appears that some of the treatments may be multiphasic as a population,
similar to what happened in some 6 month treated groups.

NMDA treatment at 12 months chronic treatments once again showed no significant
changes in overall fura-2 fluorescence in the population of CGCs (Figure 4.5B). As a whole, the
12 month NMDA treatments led to higher fluorescence at the end of imaging than 6 month
treatments, including up to 136% baseline in group O, 133% in group M, 161% in group I and
154% in group M/I, but due to variation, the two groups were not different, nor were the
treatment groups different from one another.

KCl depolarizations appeared to have a more noticeable effect in 12 month chronic
treatment slices, with population-wide transient depolarizations in all 4 chronic treatment groups
(Figure 4.5C). Increases occurred between 80-320 s for group O, between 120-340 s for group
M, between 120-400 s for group I, and between 140-500 s in group M/I. Due to high variability,

149

these chronic treatments were not significantly different compared to one another, nor were they
different from 6 month treatments.

In the last 5-7 batches of the 12 month treated mice, an additional acute treatment series
2+

was added. The first treatment of this series was continuous Mg -free ACSF, the starting buffer
for all other treatments, to try to establish a baseline for activity. Even without a stimulus
treatment, individual, transient fluorescence increases were observed in some cells in slices. In
addition to these spontaneous transients, some slices experienced population-wide depolarization
despite the lack of a stimulus. This occurred frequently enough that ACSF treatments as a whole
showed trends toward increases in the fluorescence ratio. This increase was significant in MeHg
treatments, where group M increased at 220 s onward to up to 247% baseline and group M/I
increased from 440 s onward to up to 274% baseline (Figure 4.6A). These results indicate that, at
least in 12 month chronic MeHg treatment, a change in Ca

2+

dynamics and homeostasis might be

occurring even without stimulus.

Following the ACSF control treatments, a treatment containing muscimol was added to
the slice. Muscimol activates GABAARs, and is of interest here. GABAAR activation is coupled
2+

to VGCCs, so it could contribute to changes in [Ca ]i even in adult CGCs, and may even
depolarize some neurons under stress (Eilers, et al., 2001, Chavas, et al., 2004, Galeffi, et al.,
2004). MeHg exposure appears to potentiate GABAARs at least temporarily (Arakawa, et al.,
1991,Yuan and Atchison, 2003). GABAARs internalize during stress, so differences in response
to muscimol could indicate changes to receptor expression and function (Lyons, et al., 2001,
Maguire, et al., 2009, Basu, et al., 2011). In fact, both groups M and M/I show significant
increases in fluorescence over baseline (Figure 4.6B). Group M increased significantly over

150

baseline at about 480s onwards and increased to a maximum of 400% of baseline. Group M/I
had a significant increase above baseline from 60-100 s, but grew too variable to continue to be
significant. As with all other treatments, population variability was too great to compare chronic
treatment groups directly.

Finally, bicuculline was used as an antagonist of GABAARs to complement treatment
with muscimol. Similarly to NMDA treatment, while there was a tendency for fluorescence to
increase in several groups, in this case Group O, Group M/I, and particularly in Group M, there
was sufficient variability to keep the increases from significance (Figure 4.6C).

Several factors of slice quality were checked in 12 month mouse slices. Up to 99% of
CGCs picked up by automatic tracking were unusable because they started at an inverted ratio of
340/380 fluorescence. This was not dependent on chronic treatment. All groups averaged about
35% of cells in this inverted state. Inverted ratios were not significantly related to the time taken
from the end of slicing to the start of imaging by linear regression, nor were they associated with
the batch (1-16) of the mouse (data not shown). Similarly, “extinguished” cells as tagged by
automated algorithms, displayed that about 15% and up to 58% of the remaining cells lost
significant total fluorescence during treatment series and imaging. As with inverted fluorescence
counts, extinguished cell counts were also not dependent on chronic treatment, acute treatment
or time from slicing (data not shown).

151

Figure 4.6. Mean relative fura-2 ratio values from acute ACSF control, 100 μM muscimol and 10
μM bicuculline applications as measured from high-speed (0.5 Hz) time course image series in
cerebellar slices from 15 month old mice. Mice were treated chronically for 12 months with
MeHg or isradipine. Separate lines represent different chronic treatments, including no MeHg
and no isradipine (Group O, black line, n=5, representative +SEM shown), MeHg and no
isradipine (Group M, red line, n=6), isradipine and no MeHg (Group I, blue line, n=4), and both
MeHg and isradipine (Group M/I, magenta line, n=6). Treatments are in order, and consisted of
treatment for 2 min, with approximately 20 s to reach the bath, imaging for an additional 8 min
and a 2 min wash between imaging. A. ACSF control treatment (no change from pretreatment)
of slices. Both chronically MeHg-treated groups (M and M/I) showed significant responses
during this treatment (*, P<0.05 compared to baseline). B. Muscimol treatment of slices. Both
chronically MeHg-treated groups showed significant increases in fluorescence, including the
largest increase of any treatment in Group M. C. Bicuculline treatment of slices, showing no
significant changes from baseline.

152

Figure 4.6 (cont’d)

153

E) DISCUSSION
2+

Both MeHg and isradipine chronic treatments in adult mice show effects on [Ca ]idependent mechanisms. None of the chronic treatments substantially protects against acute
2+

MeHg disruption of [Ca ] , baseline increases in fluorescence all groups at both 9 and 15
months of age are about 2-fold. This response indicates that chronic exposures do not greatly
change the response to acute exposures. However, chronic isradipine in diet with or without
2+

MeHg subtly reduce acute MeHg disruption of [Ca ]i, indicating some potential alterations to
the underlying mechanisms behind the disruption. At a population-wide level, differences in
responses to AMPA, NMDA, KCl depolarization, muscimol and bicuculline are also subtle, but
some common differences occur. First, baseline effects at 9 months of age with no chronic
treatments appear to be about a 2-fold increase in fluo4 fluorescence in response to AMPA, no
population-wide response to NMDA and a transient response to KCl depolarization, the last of
which is not seen in other chronic treatments. At 15 months of age, there is a larger increase in
fluorescence induced by AMPA, still no population-wide response to NMDA, and a larger
response to KCl. Additionally, there is no population-wide response to ACSF, no response to
muscimol and no response to bicuculline in this treatment group. Isradipine treatment as a whole
appears to reduce both the effects of acute exposure to MeHg and some of the effects of other
acute stimuli. Chronic MeHg treatments as a whole appear to enhance both spontaneous
depolarization seen in the high-speed imaging of ACSF treatment and the reaction to muscimol.

As for individual treatments, there are subtle differences as well. Acute application of 20
µM MeHg application on its own does not appear to be much different from results in neonatal
mice described in previous chapters, at least as far as fold change in relative fluorescence. 200%

154

changes to baseline are around the high end for CGCs in the neonatal IGL, but not particularly
different, and adult and geriatric mice do not have the more susceptible developing CGCs. There
is not a large difference between age groups within this project either. AMPA and KCl
2+

treatments display somewhat enhanced [Ca ]i responses in the 12 month group, but not
remarkably so, and not enough to indicate statistical differences. NMDA application was able to
cause responses in some CGCs, but not enough to cause any population-wide increases like other
2+

treatments. This may point to the difficulty of requiring Mg -free buffer; even though the slices
2+

spend over an hour in incubations and baths without Mg , endogenous intracellular Mg2+ may
still be present and may still block many NMDA receptors in the slice. Age effect was not
compared in NMDA treatment because of the lack of significant acute population-wide effects.
Age effects also were not compared in ACSF, muscimol and bicuculline treatments because they
were not used in the 6 month experiments.

Along with mostly subtle effects, some unexpected results complicate the conclusions
that could be taken from these sets of experiments. Control treatments were suggested early for
slices on the confocal and high-speed imaging platforms, but slices and time were limited, so
they were not incorporated until the 12 month treatment period. These control treatments were
helpful in finding subtle differences in the confocal results, and allowed for some
2+

characterization of [Ca ]i -based GABAAR response in the slices, but the untreated controls
produced a surprising result. Many untreated slices depolarized with no stimulus. It is unclear
what might have caused such a depolarization, it is possible the change from incubation in a 5%
CO2 incubator to perfusion with 95/5% O2/CO2 ACSF buffer could disturb the neurons enough
to depolarize. Even with a 10 min wash before imaging, this depolarization happened frequently

155

enough during the ACSF control treatment that the overall fluorescence means increased in some
treatment groups. The significant changes in fluorescence shown during the ACSF treatment of
slices are particularly problematic to the other results, because it indicates that at least in the
older mice that individual cells and even the whole slice can depolarize at any time, meaning that
at least some of the depolarization noted in other treatments may not be due to treatment. It is
particularly confusing when compared to NMDA and bicuculline responses, which both had
small, mostly linear responses over time. As a population, those responses might now be
2+

considered to be lower than no treatment, as they don’t cause any mean increases in [Ca ]i. It
might be possible to further characterize this issue, but it would require more adult mice and
possibly more chronically-treated mice. An ACSF treatment like this could at least be used as a
baseline for comparisons to the AMPA treatments from the same animals, as they were both the
first in a series. This could be performed similar to paired comparisons in Chapter Three and to
some extent in the second set of confocal results in this chapter. Spontaneous depolarization may
also contribute to the apparent multiphase traces of some AMPA depolarizations. Some of the
issues in comparing populations of CGCs in these experiments may also be solved by comparing
individual traces instead of population means.
2+

Few chronic treatment studies have directly characterized Ca -dependent biochemical
mechanisms. This is in part because many characterizations, including the Ca

2+

imaging

presented here, require euthanasia of the animal. Most rely on behavioral monitoring followed
by pathology or other fixed tissue or homogenized end-points, such as total neurotransmitter or
ROS concentration, instead of live or ex vivo experiments on tissue (Heath et al., 2010,
Bourdineaud, et al., 2012). In one experiment that isolated and plated live CGCs, these

156

dissociated CGCs from aged mice treated acutely with MeHg showed no difference in basal
2+

2+

[Ca ]i levels (Bellum, et al., 2012). This is not surprising as this basal [Ca ]i at the time
collected could easily be based more on the age of the animal than an acute dose of toxicant. The
differences in ROS and mitochondrial depolarization detected in that same set of experiments
2+

indicates that Ca -dependent mechanisms are still involved in acute toxicity. Here we
demonstrate changes to Ca

2+

dynamics imaged live in slices as opposed to single measurements

in dissociated CGCs. Results here of course are part of a larger project, and the results from
approaches such as electrophysiology, ATP production, gene expression, histology, and behavior
have not yet come together. Behavioral approaches in this collaboration are also coming
together, indicating that another Ca

2+

blocker, nimodipine, is able to reduce or eliminate

behavioral deficits in memory acquisition and performance caused by MeHg (Bailey, et al.,
2013). These comparisons are needed for this project to piece together many conclusions;
changes determined in other approaches will complement results here and these results will
complement the other approaches. For instance, we show here several differences in responses to
receptor agonists. If gene expression or immunohistochemistry results indicate a difference in
GABAAR expression with chronic MeHg treatment, then the differences in fura-2 response in
the muscimol results presented here could be interpreted as due to expression. Gene expression
may in fact be the keystone to this project, as chronic exposures, particularly those that don’t
cause overt effects, may still alter gene expression, signaling pathways and susceptibility to other
treatments. This collaboration of biochemical, behavioral and pathological approaches to
studying the toxicity of MeHg may redefine the level of complexity that can be assessed and
integrated in toxicological studies.

157

CHAPTER FIVE

HIGH-SPEED Ca

2+

IMAGING REVEALS SPONTANEOUS Ca

2+

TRANSIENTS IN

CEREBELLAR SLICES FROM YOUNG AND OLD MICE

158

A) ABSTRACT

Spontaneous neuronal and glial activity is a hallmark of many normal functions of the
central nervous system. Loss of this activity can have profound effects, such as those associated
with neurodegenerative diseases. In experiments meant to optimize imaging of small, transient
calcium responses in cerebellar slice preparations to depolarization, agonists and the
neurotoxicant methylmercury, we regularly visualized unexpected spontaneous increases in
fluorescence in some cells. We imaged fluo4 fluorescence, an indicator of internal Ca

2+

concentration, in slices of mouse cerebellum with a high-speed CCD camera and epifluorescence
microscope. Images were acquired at 2 Hz and 60x magnification and focused on cells from the
internal granule cell layer. Groups included young (PND 8-14) (+/+) and GABAAR α6 (-/-) mice
and 22 month-old (+/+) mice. The knockout line lacks tonic inhibition in mature cerebellar
granule cells and displayed increased calcium responses to MeHg exposure in previous
experiments. Older animals may also display differences in calcium homeostasis, contributing to
susceptibility to toxicity and neurodegeneration. Previous experiments determined that stimuli
such as NMDA receptor activation induce Ca

2+

transients and activity in cells, but stimulus-free

activity has not been detailed previously in cerebellar slices. During pre-treatment imaging in all
cases, spontaneous increases in fluorescence occurred in 4 - 33% of visible cells. Events were
rapid; typically within 2.5 seconds they increased fluorescence 50-300% compared to the mean
of the whole trace. Fluorescence returned to a steady baseline near the pre-event intensity over 212 seconds in young mouse cerebella. There was no significant difference between young (+/+)
and (-/-) number or intensity of spontaneous activity, suggesting these events are not in general
related to α6-containing GABAARs in of granule cell neurons. However, young (-/-) mice had a

159

larger number of shorter duration transients in their distribution than their (+/+) counterparts,
suggesting there might be different forms of these transient in the distribution. In slices from
older mice, it appears that there is no rapid return to baseline; spontaneous events remained near
their peak for several minutes of imaging, and the slices contained more long events than short.
In all groups, events were smaller, shorter and occurred in fewer cells than fluorescence
increases associated with depolarization of the whole slice. Among spontaneous events, cells
showed a variety of responses, including synchrony and repetition, unrelated to imaging
conditions. The identity of cells involved in these events is not certain, but based on location,
shape and relative abundance they could be glial cells or processes such as basket cell terminals
rather than neurons. These unexpected results show clearly that it is possible to regularly image
spontaneous live activity in slices of cerebellum in a relatively simple setup.

160

B) INTRODUCTION

As introduced in previous chapters, MeHg is a neurotoxicant that targets the cerebellum,
particularly developing CGCs. Molecular targets include GABAARs and several other regulators
of cellular Ca

2+

homeostasis. We have utilized multiple techniques for recording MeHg effects

on molecular targets, including electrophysiology and cell viability assays, and this dissertation
has already shown several approaches using cerebellar slices and Ca

2+

imaging. Chapters Two

2+

through Four all indicate that MeHg and other agents alter [Ca ]i in young and old animals.
Both Chapters Two and Three indicate subtle effects of GABAAR modulators in attenuating the
2+

effects of MeHg on [Ca ]i. Additionally, Chapter Three indicates that knockout of the α6
2+

subunit of the GABAAR in mice makes immature CGCs more susceptible to [Ca ]i increases
induced by MeHg.

Despite the multiple results indicating GABAARs and Ca

2+

signaling as important

components of MeHg toxicity, confocal imaging approaches do not offer optimal temporal
resolution for these events. This is particularly noticeable in the response of cerebellar slices to
muscimol versus bicuculline. In order for an agonist to desensitize a receptor, as has been
suggested for the reason muscimol appears to antagonize GABAARs in confocal images of
Chapters Two and Three, it must first activate the receptor. In this case GABAAR activation by
2+

muscimol is expected to temporarily increase [Ca ]i before desensitizing the receptor,
preventing some of the potentiation associated with MeHg. This connection between GABAAR
2+

activation and an increase in [Ca ]i has been established in several models (Eilers, et al., 2001,
Brodinsky et al., 2003) and has even been shown in previous imaging experiments in our lab

161

(Limke, dissertation observation). Despite this, confocal results in Chapters Two and Three do
2+

not indicate the presence of a transient increase in [Ca ]i.
With the possibility that confocal imaging is missing subtle alterations in Ca

2+

dynamics,

2+

we sought to visualize [Ca ]i in a system that acquires images rapidly. Ideally this approach
could still sample from a large population of cells in slice and utilize protocols already
established, such as the use of the bath-applied fluorophore fluo4. In addition, we sought to
2+

measure [Ca ]i activity in slices from multiple groups for comparison, including those in
developing mice, where CGCs are still migrating, in adult and geriatric mice and in a mouse line
lacking the GABAAR α6 subunit. Because higher temporal resolution requires larger imaging
data sets, we sought to automate the selection of cell regions of interest (ROIs) and analysis of
2+

[Ca ]i. activity where possible. These are the first steps to optimizing a new approach where we
could compare the Ca

2+

dynamics involved in MeHg and test hypotheses generated by other

experiments, such as whether or not there is a transient Ca

2+

influx in slices treated with

muscimol. The results here were surprising, in that we uncovered Ca

2+

activity that could be

measured without providing a stimulus. These experiments did not achieve the intended
outcomes and optimizations, such as a collection of depolarizations and MeHg treatments in
slices, but are presented primarily as a baseline for future experiments.

162

C) MATERIALS AND METHODS

a, Chemicals and solutions

Chemicals and solutions used for these experiments have been described previously in
Chapters Two and Four. These include ACSF, sucrose slicing solution, fluo4 and ACSF with 40
mM KCl. Where handling is any different, the methods are described here in full. All solutions
were continuously bubbled with 95%/5% O2/CO2 during all experiments.
b. Animal care and use

All animal use was in accordance with NIH guidelines for laboratory animals and all
procedures were approved by the MSU Institutional Animal Care and Use Committee. As this
was primarily an attempt to develop protocols for use of a high-speed epifluorescence
microscope platform for use in the experiments described in Chapter Four, these experiments
included tissue from multiple sources. Tissue included cerebella from post-natal day (PND) 7-14
GABAAR α6 (-/-) and (+/+) mice included in the experiments of Chapter Two. Additionally, one
3 month (+/+) mouse from the same colony, and two 22 month IRC mice (Harlan, Frederick,
MD) from a separate experiment not described in this dissertation were used. The 3 month
mouse was grouped with the 22 month mice for the purposes of having two groups, and did
overlap with the young mice in any measures taken.

Collection of cerebella was described previously in Chapters Three and Four. For young
mice, cerebella were collected as in Chapter Three, with rapid decapitation and slicing in icecold slicing solution. For adult mice, collection was performed as described in Chapter Four,

163

with cardiac perfusion of near-frozen blended slicing solution, decapitation and slicing in
blended slicing solution. Slice incubations were similar, acclimation in ACSF contained in a
heated water bath at 37°C for 30 min, followed by 30 min at room temperature (23-25°C) and 1
h in an opaque container for fluo4 incubation before use. Multiple slices from the same animal
were used if possible.

c. Image acquisition and treatment

Slices were moved from incubation to a chamber on an inverse epifluorescence
microscope (Nikon Eclipse TE2000-U) and anchored in a bath of ACSF perfusing at about 2
mL/min. Slices were illuminated using a fast switching polychromatic light source (Polychrome
V, Till Photonics, GmbH), set at 488 nm excitation and 512 emission, through a custom dichroic
filter (Chroma 71000A Fura2 C59742), a 40x oil objective and either a 1x or 1.5x mirror. A low
fluorescence oil (FF, Cargill, Cedar Grove, NJ) was used for the oil immersion beneath the
chamber. Illumination time was minimized by switching to live software imaging (Till Vision,
Till Photonics) when possible rather than constant illumination through the eyepiece. Images
were captured using an electron-multiplying CCD camera (iXon EM+, Andor, South Windsor,
CT)

Regions of the slice were chosen that contained at least several dozen visible and clear
CGCs and no visible tissue damage or scarring. Exposure time was adjusted so that the bulk of
visible cells was at about 25% of the camera’s available maximum, to allow for optimum
dynamic range. Once a frame of interest was ready, 512 x 512 images were taken in series at 2-4
Hz for 150-600 s, depending on treatment. Some slices received treatment after 150 seconds,

164

consisting of 40 or 80 µM KCl ACSF, 20 or 100µM MeHg. Since none of these treatments were
repeated sufficiently often for quantitative comparisons, only the initial 150 s before treatment
were compared from these slices, and treatments are only noted as qualitative comparisons.

Image series were exported from TillVision software as tagged image files (.tif) and
imported into ImageJ. Image series were trimmed if they contained treatments and backgrounds
were subtracted using a 50 pixel rolling ball radius method. These processed files were moved
into a folder for batch processing by the LC_Pro plugin, which is designed to find intensity
transients in large populations of cells based only on the size of the cell (Francis, et al., 2012).
The total count of cells determined by the algorithm was obtained while the algorithm ran; it is
not present in the output. These cell counts also do not account for regional differences, such as
the EGL present in young mice, so they are not compared to framed region areas, as previous
experiments. Outputs from the algorithm include number, repetition, significant traces, and
location of each of the significant events. Means of several parameters, including number of total
cells, percent active cells, intensity of events and frequency of events was calculated for each of
the three groups and compared using T-tests. Duration of transients was plotted as a cumulative
% histogram and maximum distance (maxD) was calculated and compared to the critical
difference (critD). To calculate critD, the standard deviation (σ = 1.36 for α=0.05) was
multiplied by the square root of the sum of counts to be compared divided by the product of the
counts to be compared (σ * % (sum of counts / product of counts)). For comparing young (+/+)
(n=47 events) and young (-/-) (n=416 events), critD was 16%. Comparing young (+/+) and old
(+/+) (n=26 events), critD was 30%

165

D) RESULTS

a. Visualization of CGCs

CGCs labeled with fluo4 were characterized as small round somata clustered in the IGL,
and in young slices also in the EGL. A frame of interest contained these CGCs in at least 25% of
the frame. Pixel size of CGCs was dependent on magnification (40x or 60x), but equaled 5-10
μm, with the larger CGCs situated in the IGL compared to the precursors in the EGL. Unlike
confocal recording, the entire slice thickness could not be scanned, so some of the CGCs
remained out of focus. Slightly out of focus cells were still trackable. Slices of cerebellum from
young mice had characteristic layers as described in Chapter Three, with CGCs in the EGL, IGL,
and sometimes migrating through the ML, though without scanning of the slice these ML CGCs
were not always present.

b. Spontaneous Ca

2+

transients

Spontaneous flashes of fluo4 from some cells were noted even while acquiring a frame of
interest for treatments. These were concentrated in cells, with the brightness of the flash often
extending throughout the frame due to reflectivity. Some transients occurred in cell soma in
focus, presumably CGCs because of the size, shape and positions, while other flashes were out
of focus or out of the frame. These events were visible, but were shorter and less intense than
those caused by stimuli such as depolarization (Figure 5.1).

166

Figure 5.1. Example of two spontaneous events and KCl depolarization from a young (+/+) mouse cerebellar slice loaded with fluo4AM. Images are registered, cropped and contrast enhanced for clarity. Images were captured at 60x magnification; the scale bar equals
100 μm. Two nearly simultaneous events in the IGL are depicted in the first two images. The first image is just before the events and
the second at the peak intensity. Arrows indicate the two cells showing the events. This slice was later treated with high [KCl] ACSF
for the purposes of depolarizing the slice. Peak depolarization is depicted in the third image. The two cells that displayed spontaneous
transients were depolarized by this treatment, as were most cells in the slice.

167

Figure 5.2. Representative fluo4 intensity traces from cells in three groups. Each trace is
normalized to the mean intensity value of the individual trace, so units are relative. The first
group is traces from young (PND 4-14) GABAAR α6 (+/+) mice while the second is traces from
young (-/-) mice. The third set includes traces from older mice (22 months) from a separate line,
so they are considered (+/+) for the purposes of these results. Some qualities of the traces can be
seen from these examples. Note that several spontaneous events are repeated in young animals
and the fluo4 events in aged animals were much more protracted..

168

An ImageJ plugin, LC_Pro, was used to assess these flashes. From an image series with
background subtracted and images registered to avoid movement, LC_Pro took image series and
determined the presence of cells then proceeded to tally and characterize any significant
increases in live cell fluorescence, in this case fluo4.

Subjects were originally intended to be divided into 3 age groups, young, adult and
geriatric, however not enough adult and geriatric mice could be gathered in the available time
frame and hence the two groups were combined. The adult (3 mo) mouse and the geriatric (22
mo) mice had enough repetitions each to determine that the groups were not significantly
different in number of cells counted, % of active cells, and peak intensities. Mice were
subsequently divided into age and genotype groups, including (+/+) young ((+/+) Y), (-/-) young
((-/-) Y) and (+/+) aged ((+/+) A).

LC_Pro returned 482 events from frames of interest across a total of 23 slices used
(Figure 5.2). These events occurred in 389 unique ROIs out of a total of 1615 ROIs. Once
stratified into age and genotype groups, the groups did not differ in number of cells counted in
frames of interest (Figure 5.3), or intensity of events (Figure 5.4), (+/+) A mice had significantly
fewer active cells producing transients as a percent of total cells, (4 ± 3%) compared to (+/+) Y
(19 ± 4%) and (-/-) Y mice (27 ± 6%) (Figure 5.5).

Additionally, the duration of these transients appears to differ by group. Because of the
wide distribution and variability of events, even though the mean duration of events in aged mice
was over 20-fold longer than in young mice, the comparison could not be made from mean
values, but had to be calculated from cumulative distribution plots. (+/+) A mice have a longer

169

Figure 5.3. Slice characteristics from total cell and active cell regions of interest (ROI). Parameters from the automated event counter
included cell counts and unique event ROIs, allowing a count of total cells as well as active cells. Because of the differences between
layers in young (Y) and aged (A) mice, these counts are not divided into layers or corrected to calculate cell density as has been
performed in previous experiments in this dissertation. No significant differences were found in comparisons of total cell count.
Young α6 (+/+) and (-/-) slices showed 13-33% active cells while aged slices only showed about 4% compared to total fluorescent
cells, which is significantly fewer (P<0.05).
170

Figure 5.4. Mean fluo4 intensity of events within frames of interest in slices from α6 (+/+) and (-/-) genotypes and young and old age
groups. Event parameters were calculated relative to the mean value of an ROI over all time points, so values are relative. Even
though it appeared that (+/+) A had lower overall intensity, there were no significant differences.
171

Figure 5.5. Cumulative distribution of event durations in slices from α6 (+/+) and (-/-) genotypes
and young and old age groups. Events were distributed into 1 sec bins and plotted cumulatively
up to 100% and out to the longest duration of 174 sec. Because of the relative differences in
number of total events, young mice have a smoother distribution. Statistics were run comparing
the largest % difference between distributions and comparing to the critical difference (D)
calculated from event counts and standard deviation (σ). Slices from young (-/-) mice had
significantly more short duration events than (+/+) mice, while slices from aged (+/+) mice had
significantly more long duration events.

172

distribution than (+/+) Y, meaning that more of their transients last longer. At the maximum
difference, 97% of transients in (+/+) Y are less than 33 s long, while only 42% of (+/+) are less
than 33 s. In contrast, (-/-) Y mice have a shorter distribution compared to (+/+) Y. 70% of (-/-)
Y transients are less than 5 s long, while only 15% of (+/+) are that short. While this could have
been due to frequent repetition in (-/-), it does not appear that transients repeat significantly more
in (-/-) Y compared to (+/+) Y, while (+/+) A mice have almost no repetition of transients
(Figure 5.6).
2+

Overall this indicates that older mice have fewer, but longer Ca

transients, many of

which appear to be due to long tails. Additionally, while the GABAAR α6 (-/-) do not have a
phenotype, and do not differ from (+/+) in percent of active cells or frequency of repeated firing,
they do appear to differ in spontaneous activity, with more short duration events. The frequency
of events in (-/-) young, which would reflect an increase in repetition of the transients, is several
fold greater by mean, 0.23 Hz, than (+/+) young at 0.04 Hz, but is not significantly different.

173

Figure 5.6. Mean frequency of events in slices from α6 (+/+) and (-/-) genotypes and young and old age groups. Frequency was
calculated from the total events over total time, with some total times reduced to just pretreatment time to focus on the spontaneous
events. All slices had at least one significant event. Events occur about every 20 s in young (+/+) slices, about every 4 s in young (-/-)
slices and about every 150 s in (+/+) aged slices.
174

E) DISCUSSION
2+

The presence of spontaneous transient Ca

events in slices came as a surprise, while

electrophysiological monitoring of neurons and interneurons in slices frequently show
spontaneous activity, Ca

2+

transients without stimulation have not been characterized visually

with epifluorescence microscopy. In confocal imaging used in Chapters Two through Four,
because of the long time between image captures, the fine planar focus and linear scanning of the
slice, these events were not generally visible, though presumably they were occurring. With the
same fluorophore loading protocols in the same type of cerebellar slice, using epifluorescence
microscopy we observed Ca

2+

transients even through the microscope eyepieces. Here we show

some simple grouping of initial results and characterize several significant differences. The most
obvious is the differences in the overall activity of neurons in older slices compared to younger
slices, but we also demonstrate a difference in transient duration between genotypes.

In older mice, the lower percent active cells and frequency of events are likely related, as
fewer active cells means fewer overall events and less chance of more than one event occurring,
resulting in a frequency of 0. This does not appear to be a result of a higher baseline of
fluorescent cells, which could contribute to fewer significant events, nor do the slices from aged
mice appear to contain fewer overall cell counts due to death or dysfunction. It is possible that
fewer events occur because cells require more time to recover. This may be, but some events
appear to have multiple peaks during recovery, such as the final trace in young (-/-) mice in
Figure 5.2, as opposed to the long linear recovery of several of peaks from aged mice.
Alternatively, it is possible that cells in slices from the young mice are more active, as opposed
to those from the older mice being less so. Indeed, we would expect any CGCs that are still

175

migrating to express Ca

2+

oscillations critical for migration similar to what has been

characterized in CGC cell cultures (Komuro and Rakic, 1992, Komuro and Rakic, 1993, Komuro
and Kamada, 2005). It is possible that these oscillations make up the majority of the transient
events demonstrated here in acute slices. Even CGCs embedded in the IGL might still exhibit
these frequent oscillations if they haven’t matured to the point at which this mechanism stops.
Migratory transients do not explain the events visualized here in the adult and aged mice,
though. The protocols for young mice and old mice are sufficiently different, because of the need
for more preservative steps in collection and use of tissue, that the protocol or buffers used could
reduce the activity of slices from aged mice. Further characterization is necessary here, both to
establish the adult mouse brain as a baseline compared to young mice and aged mice, and to
determine any differences that protocols and buffers contribute to activity. These would be valid
approaches and optimizations, necessary even before testing the effects of stimuli such as MeHg
and depolarization.

The source of the difference in transient durations between young (-/-) and (+/+) is less
clear than are the reasons for the differences in the young versus aged animals. Without tonic
inhibition provided by extrasynaptic GABAARs in mature CGCs, those neurons could
conceivably be more active and display more events. This agrees with both research into the
molecular mechanisms of GABAAR α6 knockout mice (Brickley et al., 2001) and in the
pharmacology of the whole animal, in which potentiation of remaining GABAARs induces
dysfunction (Korpi, et al., 1998). Alternatively, less direct mechanisms could contribute to the
observations made. It is possible that altered signaling mechanisms from mature CGCs lacking
the GABAAR α6 subunit change the state of immature CGCs in the EGL, contributing to more
events from dividing or migrating CGC precursors in the the EGL or ML. The locations of ROIs
176

provided by the algorithm do not appear to indicate that this is the case, as (-/-) do not appear to
have increased numbers of ROIs in the EGL and ML. Even so, it is unclear if some of the active
cells in the IGL are still migrating, versus maturing and putting out axons. It would be valuable
to characterize the individual cells that have been highlighted for events further, and to
characterize the different layers in young slices, but this will require many more repetitions of
these experiments to optimize.

One of the drawbacks of automating the tracing of these cells is that it could conceivably
pick up events that are not based in CGCs. While the locations of the events have been traced
back to the original images and many are clearly CGC neurons, it is not always clear the type of
cell that was visualized. In some cases cells overlap or are out of focus or do not appear to be
quite the normal size or shape of CGCs. It is possible that other cell types contribute to the
spontaneous transients. Golgi interneurons are frequently firing action potentials during
development while making inhibitory synapses (Farrant and Brickley, 2003), and though larger
and more branched than CGCs, might be captured by the algorithm. Basket cell termini, located
in the PC layer, could also be dense enough to be confused for a cell body, and have been shown
to have spontaneous activity in intact two-photon imaging of cerebellum (Conti, et al., 2004).
Because not all of the ROIs associated with events can conclusively be ascribed to CGCs, they
are here referred to as cells.

There are a few potential solutions to determine the cell populations that are contributing
to spontaneous events, and to further characterize these transients. First, CGC-specific labels
would be an ideal solution, particularly if those labels can be associated with the live transients.
This could be achieved in a few ways, most notably by other live imaging methods. The

177

genetically-encoded calcium indicator GCaMP could be used with specific promoters instead of
the bath fluorophore fluo4 to indicate Ca

2+

dynamics in cell populations at any level of

specificity, from neurons versus glia, to CGCs, and perhaps even specific developmental stages
of CGCs. Alternatively, live cell fluorescent glucose analogues and live measurements of NADH
in neurons can also distinguish between neurons and glia, and would have the added benefit of
producing measures of energy metabolism associated with transient activity (Pancani, et al.,
2011). Fixation and immunohistochemistry of cells post-recording could also be used to
determine surface expression of receptors and other relevant factors. If the location imaged can
then be traced back from IHC, the specific cells could be associated with transient activity, not
just generalized. This might be accomplished with rapid fixation while the slice is still on the
microscope, but would have to be optimized.

Coupling of electrophysiology and Ca

2+

imaging is another potential approach. Patching

of neurons in slice is possible, and with up to 33% of neurons displaying fluorescent transients in
only a few minutes of recording, many cells could be recorded by both techniques. Readings
from an electrode could immediately determine if these transients are associated with action
potentials, and would characterize the events well. It is clear that these types of transients are not
all the same, electrophysiology would further characterize them. Some of the transients might be
migrational Ca

2+

oscillations, in which case there would not be action potentials, while signals

from glia or basket cell termini might be associated with characteristic neurotransmitter release.
Some of the fluorescent transients might be other signals entirely. Electrophysiology setups also
allow for other valuable tools, such as puff application of stimulants directly onto a neuron of
interest or even injection of labels. This high-speed imaging approach on its own has much

178

potential, but coupling high-speed imaging with electrophysiology would provide an excellent
direction to further experiments.

179

CHAPTER SIX

SUMMARY AND CONCLUSIONS

180

SUMMARY AND CONCLUSIONS
2+

This dissertation presents several projects designed to relate MeHg toxicity, [Ca ]i
homeostasis and the GABAA receptor. These projects build on the correlation we have
2+

established between MeHg toxicity and increases in [Ca ]i by demonstrating that the same
signaling mechanisms which occur in cultured neurons occur in functionally intact slices under a
variety of conditions, from developing rats and mice to chronically-treated aged mice.
Treatments focus on a receptor that is susceptible to MeHg but that has not been fully
characterized in the mechanisms of MeHg toxicity.

The results presented in this dissertation support the hypothesis that MeHg disrupts
2+

[Ca ]i partially through the actions of GABAAR signaling and that forcing block of GABAAR
2+

signaling can interfere with the actions of MeHg on [Ca ]i (Figure 6.1). MeHg acts on CGCs to
2+

increase [Ca ]i and CGCs show developmental and aging susceptibility. In developing CGCs in
2+

both rats and mice, where GABAAR activation would be depolarizing, increases in [Ca ]i are
enhanced compared to mature CGCs. In mature CGCs, where GABAAR activation would be
hyperpolarizing. GABAAR modulators reduce, but do not eliminate the immediate effects of
2+

MeHg on [Ca ]i and they act at all developmental stages, indicating that GABAARs are coupled
2+

2+

to [Ca ]i even in mature, hyperpolarized CGCs. While there is an increase in [Ca ]i by
muscimol alone, this effect is only temporary. This dissertation also offers a few intriguing
results that indicate that interactions between the effects of MeHg and GABAA receptors may be
more complex, based on the subtype of GABAARs involved and the duration of MeHg
treatment. Knockout of the GABAAR α6 subunit does not alter the effects of MeHg in the IGL,
where α6 is prominent in (+/+) but absent in (-/-), and not compensated by change in expression

181

2+

Figure 6.1. Cartoon depiction of GABAAR and MeHg interactions with [Ca ]i as highlighted by results of this dissertation.
182

2+

Figure 6.1 (cont’d) MeHg acts through GABAARs in part to increase [Ca ]i in all CGCs and all
2+
developmental stages. Both muscimol and bicuculline reduce MeHg effects on [Ca ]i through
GABAARs, apparently by desensitization in the case of muscimol and direct inhibition by
bicuculline. Loss of α6 subunits is disruptive to developing granule cells, possibly through
increased neurotransmission associated with mature CGCs with local synapses or possibly
2+
through alterations in GABAAR expression. [Ca ]i-dependent cell death, through enzyme
activation, rundown of ATP and aberrant nuclear signaling, is present, but not significant
compared to regular handling. Chronic MeHg appears to alter responses to muscimol, indicating
a role for changes in GABAAR surface expression, such as through internalization, as well as
possible changes in gene expression of GABAAR subunits (dashed arrows).

183

of other subunits (Jones, et al., 1997, Nusser et al., 1999). Knockout enhances the effect in
developing CGCs. These CGC neuronal precursors in α6 (-/-) mice would not be expected to be
any different from (+/+) because α2 is expressed in development. This unexpected result
indicates secondary effects of GABAAR expression and/or MeHg toxicity. Developing (-/-)
2+

CGCs also demonstrate an increase in short spontaneous [Ca ]i transients as shown by highspeed imaging. Alterations in the normal function of this critical mechanism could indicate
another susceptible target of MeHg toxicity in (-/-) mice. Finally, chronic MeHg treatment
2+

appears to unmask a large increase in the effectiveness of muscimol to increase [Ca ]i. These
results leave open several areas of potential further inquiry, both to build the characterizations of
GABAAR and MeHg signaling, and to use the techniques and optimizations established here to
probe additional interesting interactions.

In the initial set of experiments using rat cerebellar slices as the model, we first
characterized the concentration-response to MeHg in slices, as well as the viability of CGCs in
these slices. These experiments are the foundation for all the others used in this dissertation. In
studying the cellular and molecular toxicity of MeHg, it is extremely important to find balanced
concentrations to use that do not induce immediate cell death, but that can induce other nonlethal effects over the time period being observed, otherwise experiments lose relevance to actual
exposures. In time course experiments in rat neonatal slices, we determined that 20 μM MeHg is
2+

a valid concentration to use because it induces large increases in [Ca ]i-dependent fluo4
fluorescence, does not induce neuronal cell death more than trauma from slicing and time alone
do, and clearly displays developmental specificity, generating a larger increase in fluorescence in
developing CGCs in the EGL than in any other population. In other studies, concentrations of 0.2
to 5 µM MeHg have been used in neuronal culture treatments, including on cultured CGCs, and
184

induce a high incidence of cell death within a few hours (Marty and Atchison, 1998, Limke, et
al., 2004, Edwards, et al., 2005). This relative difference in MeHg concentration response
indicates that acute slice treatment may be more relevant to acute toxicity because the associated
tissue present buffers some of the MeHg compared to dissociated neurons. This is also an
interesting concentration to find because it falls within estimated blood MeHg concentrations
seen in the Iraq poisoning associated with acute adult toxicity (Bakir, et al., 1973).

In addition to the 20 μM MeHg concentration used in many parts of this dissertation,
several MeHg concentrations were added to this characterization of rat slice culture response,
2+

and it is likely that even 5 or 1 μM MeHg could have interesting [Ca ]i responses in slices with
extension of the experiments. Along that line, it appeared that monitoring of slices could easily
continue much farther than the extensions performed over the course of this dissertation. Slice
viability and maintenance of function is continually improving in slice culture protocols, such as
2+

2+

with the use of low [Ca ], high [Mg ] sucrose solutions during slicing and bath recovery
methods used in this dissertation, but could be improved in further experiments through methods
such as supplementation with n-methyl-d-glucamine in adult slicing solutions (Tanaka, et al.,
2008, Peca, et al., 2011) With the development and optimization of ACSF solutions and slice
protocols that went into this dissertation, as well as the automation of imaging at the microscope,
slice imaging could potentially extend to several hours, bridging the gap between acute slice
protocols and organotypic slice culturing.

The initial use of muscimol and bicuculline in Chapter One introduced one of the most
intriguing results of this dissertation, that an agonist and an antagonist of the same receptor could
both reduce the effect of MeHg. This was later confirmed in the experiments described in

185

Chapter Two. Previous experiments in the laboratory had demonstrated the expected opposite
effects of muscimol and bicuculline on electrophysiological spikes in hippocampal neurons
2+

(Yuan and Atchison, 1997) and on [Ca ]i in some cerebellar slices during optimization of the
bath application used in this dissertation (Mancini, unpublished dissertation report). It led to an
interesting internal hypothesis, that muscimol was desensitizing the receptor after an initial
activation and thus appearing to act as an antagonist in displaying no change at imaging time
points. This hypothesis led to many attempts to work out the timing of muscimol activation and
2+

attempts to capture an increase in [Ca ]i induced by muscimol on slices imaged with confocal
microscopy. Eventually, it led to using the high-speed imaging system for any attempts to find
such potentially transient responses. Because none of these attempts were successful or
repeatable, none of these experiments have been included in this dissertation. This agonist
appearance as antagonist mechanism still is not certain, and probably cannot be confirmed by
Ca

2+

imaging alone. Future work on this mechanism could include coupling imaging with

electrophysiology, either directly as with patch clamp studies or by monitoring network activity
using multi-electrode arrays. In the case of the former, it is much clearer to add a pulse of
muscimol directly onto a neuron and monitor Ca

2+

signaling along with electrophysiology, but

with the latter it would also be interesting to monitor the response as muscimol diffuses across a
bath.

In Chapter Two we switched models from rats to mice in order to add genotype as a
factor, since knockout mice had been established well as models compared to rats. Switching
models always has the potential to complicate protocols, but it appeared that for the most part
there was no change in responses. Slices from mice appear to have a somewhat smaller response
2+

to acute MeHg, displaying up to about a 2-fold [Ca ]i increase in CGC populations whereas rats
186

displayed up to 3-fold increases, but mouse cerebellar slices still respond significantly compared
to control. It was because of this smaller response, and inspired by seeing electrophysiological
and other biochemical recordings from two complementary phrenic nerves from lung diaphragm,
that it was determined that experiments on cerebellar slices should be paired whenever possible.

The GABAAR α6 (-/-) mouse model requires more characterization to fully understand
the results presented in this dissertation. Going into this with the MeHg model of preferential
block of α6-containing GABAARs, and with the previous results showing block by muscimol
and bicuculline in the presence of MeHg as protective, the expected result would have been
protection in adult CGCs in IGL. It is in the IGL where α6 is prominently expressed (Takayama
and Inoue, 2004), but the biggest differences in the results here were within EGL, where α6
would not normally be expressed. Since those precursor cells in the EGL have not yet normally
developed expression of α6, the (-/-) model would not be hypothesized to show any difference
between those and the (+/+). We need to characterize what is happening in the (-/-) EGL beyond
2+

the changes in [Ca ]i before this project can really be complete. Several approaches could be
2+

used. In the discussion of Chapter Three, it was suggested that the increase in [Ca ]i in EGL
CGCs might in part be due to enhanced excitotoxic release of glutamate from (-/-) GCG axons. It
would be possible to probe this hypothesis using pharmacological or toxic agents that block
neurotransmission without interfering with other mechanisms of MeHg toxicity. Other released
factors might also be measured in the bath solution or blocked as well. If the susceptibility of
developing (-/-) neurons is not based on neurotransmission, then it may be based on gene
expression alterations. Even if α6 is not yet expressed, the expression profiles of developing
(+/+) and (-/-) CGCs may still be different.

187

2+

In Chapter Four, the hypotheses that changes to [Ca ]i homeostasis in aged animals
contributes to susceptibility was tested by bringing the principles established in previous
chapters to mice chronically-treated as adults. Some of these experiments used a different
2+

[Ca ]i-dependent fluorophore, fura-2, which is ratiometric, but the same principles apply to
Ca

2+

imaging as with fluo4. The results of fura-2 Ca

2+

imaging in cerebellum slices from these

2+

mice indicate that baseline [Ca ]i is very similar between 6 month and 12 month chronic
treatment groups, but that responses to stimuli such as AMPA and KCl are slightly enhanced.
Because many MeHg response pathways, including GABAARs, might have been altered with
chronic MeHg exposure, Ca

2+

channel blockers, or with aging when compared to other groups

such as the neonatal mice characterized in Chapter Three, this project also explored acute MeHg
2+

treatment in these slices using fluo4 and confocal imaging. MeHg-induced [Ca ]i response in
these aged slices was greater than responses in IGL in neonatal mice, about 2-fold increase in
aged while neonatal slices only display about 1.5-fold increases. A direct comparison is more
complex because these are not the same mouse lines nor did they receive the same treatments
2+

throughout life. Chronic treatments only slightly altered MeHg-induced [Ca ]i responses, but
were enough to push responses above or below significance. For example, all chronic treatment
group slices displayed a lack of concerted depolarization, as would be indicated by a transient
2+

[Ca ]i increase, when treated with bath application of high [KCl] as compared to the control
aged mice. In 12 month treatment groups, treatment responses were enhanced slightly, but
enough that all groups remained above baseline with AMPA treatment and all groups displayed
significant increases during KCl depolarization. These readings may be complicated because
additional experiments performed indicated that a large enough portion of CGCs in at least two
treatment groups (M and M/I) depolarize spontaneously during imaging that this response may

188

boost the readings from other treatments. This issue was discussed in Chapter Four, and it may
require further analysis on the data to determine if and how results need to be corrected.

Chapter Four also includes results suggesting a role for GABAARs in mechanisms of
susceptibility to MeHg in aged mice, but only in the 12 month treatment groups. Because the
2+

temporal resolution has been improved, these results show an effect of muscimol on [Ca ]i,
both at the individual cellular level and at the population level as indicated by the means used in
2+

Figure 4.6 B. Muscimol increases [Ca ]i in these slices, particularly in chronic MeHg treatment
groups (M and M/I). This muscimol application in MeHg-treated mice, while highly variable,
was the largest mean increase in fura-2 fluorescence ratio of any treatment, up to about a 4-fold
increase. Bicuculline, applied shortly after muscimol, did not significantly change fluorescence.
Of all the muscimol results in this dissertation, this was the largest effect by the agent, and
clearly points to a change in chronically MeHg-treated mice. These mice have some alteration
that effectively potentiates GABAARs, making activation much more effective. These results
suggest widespread changes in chronic MeHg treatment and will be useful in strengthening
arguments in the literature against conclusions that adult exposure to MeHg displays no
functional change (Bellum, et al., 2012). It will take additional characterizations to determine the
source of this difference rather than speculate, and these characterizations may come from
additional approaches used within the framework of the group project.

The group project of which Chapter Four was a part is a massive undertaking and results
from many researchers involved in the four major parts, biochemical, behavioral,
2+

electrophysiological, and pathological, are still being processed. Ca

imaging was part of the

biochemical approach to this project, and results from other approaches within this laboratory,

189

including gene expression, immunohistochemical, electrophysiological, and metabolic assays
and analyses are of particular interest to the conclusions presented here, and will contribute to an
overall integration of results. As of now, the results presented in this dissertation must stand
independently, but some general conclusions can be applied, and several other 6 month results
from the study are available from within this lab. For instance, chronic treatments did not display
detectable differences in rota-rod motor performance or general health and weight, but MeHg
treatment did increase partial weakness in the hind limbs as indicated by the hind-limb cross test,
so the collaboration with a behavioral monitoring lab may have interesting behavioral
phenotypes to add soon. ATP-level assays indicated that the combination of chronic MeHg and
isradipine increase maximum ATP production in striatal neurons (unpublished observation). If
this metabolic change is also present in CGCs, then it would suggest the extra ATP capacity is
2+

needed to counteract toxicity brought about by large increases in [Ca ]i observed in this
dissertation. Many more connections between the approaches will be made as results become
available.

Chapter Five addresses some issues and hypotheses that came up during confocal
imaging studies. While the intent of the high-speed imaging approach was to determine if
2+

muscimol increased [Ca ]i in a way that was missed in the low temporal resolution of confocal
imaging, the opportunity to test muscimol in this imaging approach did not occur until the
experiments described in Chapter Four. Even so, the collection of results showed an interesting
2+

reaction in slices, spontaneous [Ca ]i transients, which have been shown in advanced
microscopy approaches such as two-photon microscopy, but have not been characterized in
epifluorescent microscopy of slices (Conti, et al., 2004). By dividing into age groups, these
2+

results were enough to determine some interesting interactions between GABAARs and [Ca ]i
190

transients in slices. Neonatal GABAAR α6 (-/-) mice, as used in Chapter Three, while otherwise
similar to matched (+/+) mice when measuring transients, had a distribution of events that had
many more short events. The significance of these shorter events is unclear, but as with the
results of Chapter Three, if genotype alters the activity of mature CGCs and thus
neurotransmission, then a response in developing CGCs to that activity could have altered the
signaling that regulates spontaneous events.

Together, the results from Chapter Five and the spontaneous response in slices from aged
animals in Chapter Four to extended untreated imaging indicate a need to characterize
spontaneous events in these slices further. The discussion in Chapter Five mentioned several
approaches that could be used to improve characterization of each experiment performed on the
high-speed imaging system, including coupling imaging to electrophysiology, adding other live
fluorescent probes, or fixation of slices post experiment to determine expression profiles and
other characteristics. It would be useful to consider these approaches if the types of experiment
performed in Chapters Four and Five are continued, because spontaneous activity could continue
to affect results. As for directions this type of project could go in the future, the groups used in
Chapter Five could be expanded to include neonatal, young, adult and geriatric mice to
determine the baseline characteristics of spontaneous events in far more detail than has already
been shown in these results. As GABAAR α6 (-/-) mice have already been shown here to be
significantly different at a neonatal age, it would be particularly useful to continue using this
genotype model in further age groups. Further research could extend this into other mouse lines
with aberrant Ca

2+

dynamics in the cerebellum, including the natural Ca

2+

channel mutation

lines lethargic, tottering, or stargazer, or other GABAAR subunit knockouts. From Chapter
Four, an interesting future direction would be to continue characterizing MeHg treatment and
191

2+

other chronic and acute treatments. Where MeHg acute treatment increases [Ca ]i on its own,
perhaps lower concentrations and longer treatment times would affect the spontaneous transients
clearly. Slice culture and high-speed imaging are powerful tools, and with continued
optimization and exploration, the possibilities to characterize the mechanism of spontaneous
events and how they relate to MeHg toxicity and GABAAR are unlimited.
2+

The mechanisms of action summarized in Figure 6.1 (and 1.1) indicate that [Ca ]i is a
2+

central feature of MeHg toxicity, but non-Ca -dependent effects may also be critical. For
2+

example, MeHg treatment also increases [Zn ]i in synaptosomes, (Denny, et al., 1993, Denny
and Atchison, 1994), neuronal cell lines (Hare, et al., 1993, Marty and Atchison, 1997) and
2+

cultured CGCs (Edwards, et al., 2005). In these cultured CGCs, the use of the Zn -specific
chelator TPEN diminished the first phases of fura-2 fluorescence increases during MeHg
treatment, but did not alter other effects of MeHg. Within this dissertation, experiments
presented in Chapters Two and Three did not use TPEN in order to examine all possible effects
2+

of MeHg, while 4 and 5 used TPEN in order to examine Ca -dependent effects. ROS are also
produced excessively in the cerebellum by MeHg treatment, likely through the actions of excess
2+

[Ca ]i on mitochondria, and may contribute to toxicity (Yee and Choi, 1994, Yee and Choi,
1996, Bellum, et al., 2007). This ROS production and MeHg toxicity can be reduced by
antioxidants such as probucol and vitamin E in some neuronal models (Gassó, et al., 2001,
Usuki., et al., 2001), but not others (Sarafian and Verity, 1991, Ou, et al., 1999, Edwards, et al.,
2005), suggesting that ROS production is not a critical factor in all neuron types. Interestingly, a
recent study of a fungal neurotoxin Penitrem A indicates toxicity in CGCs through GABAARs
2+

and ROS production, possibly through a [Ca ]i-dependent mechanism (Berntsen, et al., 2013).
2+

While the interaction between GABAARs and [Ca ]i is still not clear, this supports the
192

conclusions of this dissertation in demonstrating that GABAARs are coupled to Ca

2+

signaling at

least in CGCs.
2+

There are certainly many more components of the relationships between MeHg, [Ca ]i
and GABAARs that may be characterized. In several discussion sections, expression is
mentioned as an important component of this characterization, and this dissertation has only
touched on expression, both at the genetic level and at the surface of neurons, as it relates to the
ongoing project associated with Chapter Four. An interesting potential project here would be
characterizing surface expression of GABAARs as they relate to MeHg-induced stress. As
mentioned in the introduction, GABAARs can be internalized during stress, and these changes
can be seen through impermeable labeling of GABAARs as compared to permeable labels
(Maguire, et al., 2009, Poulter, et al., 2009). Alterations in surface expression like this might be
behind the change in muscimol response in chronically MeHg-treated mice. Muscimol and
2+

bicuculline are certainly useful tools in probing GABAAR effects on [Ca ]i and MeHg toxicity,
but in order to fully characterize these mechanisms and determine the specific subunits involved,
more specific GABAAR modulators should be used, and many are now just available for
research. GABAARs are receiving more attention for potential involvement in many neurological
disorders as finely tuned pharmacological targets. As the GABAAR structure, dynamics and
interactions are explored, the potential for more specific drugs and interactions with MeHg will
become available. Unlike muscimol, which is psychoactive and sedative, and bicuculline, which
can induce epilepsy, these new drugs may be used in future therapies for many disorders. As
described in Chapter One, GABAAR signaling disorders include epilepsy and anxiety, but may
be important in disorders such as schizophrenia, Alzheimer’s disease and Parkinson’s disease.
Due to the many additional mechanisms of MeHg toxicity, such as through M3 receptors,
193

VGCCs and mitochondria (Figure 1.1), at best the therapeutic potential of GABAAR modulators
for MeHg poisoning would be partial, much like Ca

2+

channel blockers such as nifedipine

(Sakamoto, et al., 1996, Marty and Atchison, 1998).
2+

In conclusion, this dissertation shows that MeHg rapidly increases [Ca ]i in CGCs in a
way that is dependent in part on GABAAR signaling. This disruption is dependent on the
2+

development of CGCs, and is enhanced in migrating CGCs, where migration is critically [Ca ]idependent. GABAAR block by bicuculline and desensitization by muscimol both delay the
immediate actions of MeHg, much as Ca

2+

channel blockers are demonstrated to do. Genetic

knockout of the prominent α6 subunit of GABAARs in mature CGCs enhances the effect of
MeHg in immature CGCs, indicating an uncharacterized interaction between CGC axons and
immature CGCs. Finally, chronic MeHg treatment is shown to greatly enhance the effectiveness
2+

of muscimol in producing [Ca ]i transients on aged cerebellar slices. Along with MeHg actions
on VGCCs, muscarinic receptors and mitochondria, the actions on and through the GABAAR
2+

contribute to [Ca ]i -dependent toxicity of MeHg.

194

REFERENCES

195

REFERENCES

Abdalla, F.H., Bellé, L.P., Bitencourt, P.E.R., da Silva, J.E., Roman, S., da Rosa, C., Schetinger,
M.R., and Moretto, M.B. (2012). Methylmercury-induced changes in target organs of
suckling rat pups. Exp. Toxicol. Path. 64, 605-609.
Abramian, A.M., Comenencia-Ortiz, E., Vithlani, M., Verena Tretter, E., Sieghart, W., Davies
P.A., and Moss, S.J. (2010). Protein Kinase C phosphorylation regulates membrane
insertion of GABAA receptor subtypes that mediate tonic inhibition. J. Biolog. Chem.
285, 41795-41805.
Akagi, H., Grandjean, P., Takizawa, Y., and Weihe, P. (1998). Methylmercury Dose Estimation
from Umbilical Cord Concentrations in Patients with Minamata Disease. Environ. Res.,
Sect. A 77, 98-103.
Amico-Ruvio, S.A., Murthy, S.E., Smith, T.P., Popescu, G.K. (2011). Zinc effects on NMDA
receptor gating kinetics. Biophys. J. 100, 1910-1918.
Amin-Zaki, L., Elhassani, S., Majeed, M.A., Clarkson, T.W., Doherty, R.A., and Greenwood, M.
(1974). Intra-uterine methylmercury poisoning in Iraq. Pediatrics 54, 587-595.
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, 58-63.
Araki, T., Kanai, Y., Kato, H., Kogure, K., Shuto, K., and Ishida, Y. (1993). Effect of vinconate
against regional age-related changes in the gerbil brain. Pharmacol. Biochem. Behav. 44,
17-25.
Aremu, D.A., Madejczyk, M.S., and Ballatori, N. (2008) N-acetylcysteine as a potential antidote
and biomonitoring agent of methylmercury exposure. Environ Health Perspect. 116,
26-31.
Aschner, M., and Aschner, J.L. (1990) Mercury neurotoxicity: mechanisms of blood-brain
barrier transport. Neurosci Biobehav Rev.14,169-76.
Aschner, M., Conklin, D.R., Yao, C.P., Allen, J.W., Tan, K.H. (1998). Induction of astrocyte
metallothioneins (MTs) by zinc confers resistance against the acute cytotoxic effects of
methylmercury on cell swelling, Na+ uptake, and K+ release. Brain Res. 813, 254-261.

196

Aschner, M., Eberle, N.B., Miller, K., and Kimelberg, H.K. (1990). Interactions of
methylmercury with rat primary astrocyte cultures: inhibition of rubidium and glutamate
uptake and induction of swelling. Brain Res. 530, 245-250.
Aschner, M., Mullaney, K.J., Wagoner, D.E. Jr., Lash, L.H., and Kimelberg, H.K. (1995).
Adenosine modulates methylmercuric chloride (MeHgCl)-induced D-aspartate release
from neonatal rat primary astrocyte cultures. Brain Res. 689, 1-8.
Atchison, W.D. (1986). Extracellular calcium-dependent and -independent effects of
methylmercury on spontaneous and potassium-evoked release of acetylcholine at the
neuromuscular junction. JPET 237, 672-680.
Atchison, W.D. (1987). Effects of activation of sodium and calcium entry on spontaneous release
of acetylcholine induced by methylmercury. J Pharmacol Exp Ther. 241, 131-9.
Atchison, W.D. (2005) Is chemical neurotransmission altered specifically during
methylmercury-induced cerebellar dysfunction. TRENDS in Pharmacol. Sci. 26, 549-557.
Atchison, W.D., and Hare, M.F. (1994). Mechanisms of methylmercury-induced neurotoxicity.
FASEB J. 8, 622-9.
Atlas, D., Wiser, O., and Trus, M. (2001). The voltage-gated Ca2+ channel is the Ca2+ sensor of
fast neurotransmitter release. Cell. Mol. Neurobiol. 21, 717-731.
Augustine, G.J., Charlton, M.P., and Smith, S.J. (1985). Calcium entry and transmitter release at
voltage-clamped nerve terminals of squid. J. Physiol. 367, 163-181.
Babot, Z., Teresa Vilaró, M., and Suñol, C. (2007). Long-term exposure to dieldrin reduces γaminobutyric acid type A and N-methyl-D-aspartate receptor function in primary cultures
of mouse cerebellar granule cells. J. Neurosci. Res. 85, 3687-3695.
Bailey, J.M., Hutsell, B.A., and Newland, M.C. (2013). Dietary nimodipine delays the onset of
methylmercury neurotoxicity in mice. Neurotoxicology 37, 108-117.
Bakir, F., Damluji, S.F., Amin-Zaki, L., Murtadha, M., Khalidi, A., Al-Rawi, N.Y., Tikriti, S.,
Dhahir, H.I., Clarkson, T.W., Smith, J.C., and Doherty, R.A. (1973). Methylmercury
poisoning in Iraq. Science 181, 230-241.
Basu, N., Scheuhammer, A.M., Rouvinen-Watt, K., Grochowina, N., Evans, R.D., O'Brien, M.,
and Chan, H.M. (2007) Decreased N-methyl-D-aspartic acid (NMDA) receptor levels are
associated with mercury exposure in wild and captive mink.NeuroToxicology, 28, 587-93
Basu, N., Scheuhammer, A.M., Rouvinen-Watt, K., Evans, R.D., Trudeau, V.L., and Chan,
L.H.M. (2010). In vitro and whole animal evidence that methylmercury disrupts

197

GABAergic systems in discrete brain regions in captive mink. Comp. Biochem. and
Physiol., part C. 151, 379-385.
Baumgartner, H.K., Gerasimenko, J.V., Thorne, C., Ferdek, P., Pozzan, T., Tepikin, A.V.,
Petersen, O.H., Sutton, R., Watson, A.J., Gerasimenko, O.V. (2009). Calcium elevation
in mitochondria is the main Ca2+ requirement for mitochondrial permeability transition
pore (mPTP) opening. J. Biol. Chem. 284, 20796-20803.
Baur, R., Simmen, U., Senn, M., Séquin, U., and Sigel, E. (2005). Novel plant substances acting
as beta subunit isoform-selective positive allosteric modulators of GABAA receptors.
Mol. Pharmacol. 68, 787-792.
Bearzatto, B., Servais, L., Roussel, C., Gall, D., Baba-Aïssa, F., Schurmans, S., de Kerchove
d'Exaerde, A., Cheron, G., and Schiffmann, S.N. (2006). Targeted calretinin expression
in granule cells of calretinin-null mice restores normal cerebellar functions. FASEB J. 20,
380-382.
Belelli, D., Harrison, N.L., Maguire, J., Macdonald, R.J., Walker, M.C., and Cope, D.W. (2009).
Extrasynaptic GABAA Receptors: Form, Pharmacology, and Function. J. Neurosci. 29,
12757-12763.
Bellanger, M., Pichery, C., Aerts, D., Berglund, M., Castaño, A., Cejchanová, M., Crettaz, P.,
Davidson, F., Esteban, M., Fischer, M.E., Gurzau, A.E., Halzlova, K., Katsonouri, A.,
Knudsen, L.E., Kolossa-Gehring, M., Koppen, G., Ligocka, D., Miklavčič, A., Reis,
M.F., Rudnai, P., Tratnik, J.S., Weihe, P., Budtz-Jørgensen, E., Grandjean, P.;
DEMO/COPHES. (2013). Economic benefits of methylmercury exposure control in
Europe: monetary value of neurotoxicity prevention. Environ. Health 12, 3.
Bellum, S., Bawa, B., Thuett, K.A., Stoica, G., Abbott, L.C. (2007). Changes in biochemical
processes in cerebellar granule cells of mice exposed to methylmercury. Int. J. Toxicol.
26, 261-269.
Bellum, S., Thuett, K.A., Bawa, B., and Abbott, L.C. (2012). The effect of methylmercury
exposure on behavior and cerebellar granule cell physiology in aged mice. J. Appl.
Toxicol. Advanced access published August 10, 2012. doi:10.1002/jat.2786.
Ben-Ari, Y., Woodin, M.A., Sernagor, E., Cancedda, L., Vinay, L., Rivera, C., Legendre, P.,
Luhmann, H.K., Bordey, A., Wenner, P., Fukuda, A., van den Pol, A.N., Gaiarsa, J., and
Cherubini, E. (2012). Refuting the challenges of the developmental shift of polarity of
GABA actions: GABA more exciting than ever! Frontiers in Cell. Neurosci. 6.35, 1-18.
Berends, M., Maex, R., De Schutter, E. (2004). A detailed three-dimensional model of the
cerebellar granule cell layer. Neurocomputing. 58-60, 587-592.

198

Berntsen, H.F., Wigestrand, M.B., Bogen, I.L., Fonnum, F., Walaas, S.I., and Moldes-Anaya, A.
(2013). Mechanisms of penitrem-induced cerebellar granule neuron death in vitro:
possible involvement of GABAA receptors and oxidative processes. Neurotoxicology 35,
129-136.
Bianchi, M.T., Haas, K.F., and Macdonald, R.L. (2002). α1 and α6 subunits specify distinct
desensitization, deactivation and neurosteroid modulation of GABAA receptors
containing the δ subunit. Neuropharmacology 43, 492-502.
Bornschein, G., Arendt, O., Hallermann, S., Brachtendorf, S., Eilers, J., and Schmidt, H. (2013).
Paired-pulse facilitation at recurrent Purkinje neuron synapses is independent of
calbindin and parvalbumin during high-frequency activation. J. Physiol. 591, 3355-3370.
Bortone, D., and Polleux, F. (2009). KCC2 expression promotes the termination of cortical
interneuron migration in a voltage-sensitive calcium-dependent manner. Neuron 62, 5371.
Bracamontes, J., McCollum, M., Esch, C., Li, P., Ann J., Steinbach, J.H., and Akk, G. (2011).
Occupation of either site for the neurosteroid allopregnanolone potentiates the opening of
the GABAA receptor induced from either transmitter binding site. Mol Pharmacol. 80,
79-86.
Bregetovski, P., and Bernard, C. (2012). Excitatory GABA: how a correct observation may turn
out to be an experimental artifact. Front. Pharmacol. 3, 65.
Brewer, L.D., Thibault, O., Staton, J., Thibault, V., Rogers, J.T., Garcia-Ramos, G., Kraner, S.,
Landfield, P.W., Porter, N.M. (2007). Increased vulnerability of hippocampal neurons
with age in culture: temporal association with increases in NMDA receptor current,
NR2A subunit expression and recruitment of L-type calcium channels. Brain Res. 1151,
20-31.
Brickley, S.G., Cull-Candy, S.G., and Farrant, M. (1996). Development of a tonic form of
synaptic inhibition in rat cerebellar granule cells resulting from activation of GABAA
receptors. J. Physiol. 497.3, 753-759.
Brickley, S.G., Farrant, M., Swanson, G.T., and Cull-Candy, S.G. (2001). CNQX increases
GABA-mediated synaptic transmission in the cerebellum by an AMPA/kainite receptor
independent mechanism. Neuropharmacology. 41, 730-736.
Brickley, S.G., Revilla, V., Cull-Candy, S.G., Wisden, W., and Farrant, M. (2001). Adaptive
regulation of neuronal excitability by a voltage-independent potassium conductance.
Nature 409, 88-92.

199

Brickley, S.G., and Mody, I. (2011). Extrasynaptic GABAA receptors: their function in the CNS
and implications for disease. Neuron. 73, 23-34.
Bright, D.P. and Brickley, S.G. (2008). Acting locally but sensing globally: impact of
GABAergic synaptic plasticity on phasic and tonic inhibition in the thalamus. J. Physiol.
586.21, 5091-5099.
Bright, D.P., Renzi, M., Bartram, J., McGee, T.P., MacKenzie, G., Hosie, A.M., Farrant, M., and
Brickley, S.G. (2011). Profound desensitization by ambient GABA limits activation of δcontaining GABAA receptors during spillover. J. Neurosci. 31, 753-763.
Briz, V., Parkash, J., Sánchez-Redondo, S., Prevot, V., and Suñol, C. (2012). Allopregnanolone
prevents Dieldrin-induced NMDA receptor internalization and neurotoxicity by
preserving GABAA receptor function. Endocrinology. 153, 847-860.
Brodinsky, L.N., O’Leary, D., Neale, J.H., Vicini, S., Coso, O.A., and Fiszman, M.L. (2003).
GABA-induced neurite outgrowth of cerebellar granule cells is mediated by GABAA
receptor activation, calcium influx and CaMKII and erk1/2 pathways. J. Neurochem. 84,
1411-1420.
Cace, I.B., Milardovic, A., Prpic, I., Krajina, R., Petrovic, O., Vukelic, P., Spiric, Z., Horvat, M.,
Mazej, D., and Snoj, J. (2011). Relationship between the prenatal exposure to low-level
of mercury and the size of a newborn's cerebellum. Med. Hypotheses. 76, 514-516.
Campbell, C.A., Mackay, K.B., Patel, S., King, P.D., Stretton, J.L., Hadingham, S.J., and
Hamilton, T.C. (1997). Effects of isradipine, an L-type calcium channel blocker on
permanent and transient focal cerebral ischemia in spontaneously hypertensive rats. Exp.
Neurol. 148, 45-50.
Cannell, M.B., and Cody, S.H. (2006). Fluorescent ion measurement. In Handbook of Biological
Confocal Microscopy (J.B. Pawley, ed.), 3rd ed., pp.736-745. Springer Science+Business
Media, New York, NY.
Castejón, O.J., Fuller, L., and Dailey, M.E. (2004). Localization of synapsin-I and PSD-95 in
developing postnatal rat cerebellar cortex. Brain Res Dev Brain Res. 151, 25-32.
Castoldi, A.F., Barni, S., Turin, I., Gandini, C., and Manzo, L. (2000) Early acute necrosis,
delayed apoptosis and cytoskeletal breakdown in cultured cerebellar granule neurons
exposed to methylmercury. J. Neurosci. Res. 59, 775-787.
Castoldi, A.F., Johansson, C., Onishchenko, N., Coccini, T., Roda, E., Vahter, M., Ceccatelli, S.,
Manzo, L. (2008). Human developmental neurotoxicity of methylmercury: impact of
variables and risk modifiers. Regul Toxicol Pharmacol 51, 201-14.

200

Catterall, W.A. and Few, A.P. (2008). Calcium channel regulation and presynaptic plasticity.
Neuron 59, 882-901.
Catterall, W.A., Perez-Reyes, E., Snutch, T.P., and Striessnig, J. (2005). International Union of
Pharmacology. XLVII. Nomenclature and structure-function relationships of Voltagegated Calcium channels. Pharmacol. Rev. 57, 411-425.
Cerella, C., Diederich, M., and Ghibelli, L. (2010). The dual role of Calcium as messenger and
stressor in cell damage, death, and survival. Int. J. Cell Biol. Article 546163.
Chad, J.E., and Eckert, R. (1984). Calcium domains associated with individual channels can
account for anomalous voltage relations of CA-dependent responses. Biophys. J. 45, 993999.
Chan, S.L., and Mattson, M.P. (1999) Caspase and calpain substrates: roles in synaptic plasticity
and cell death. J. Neurosci. Res. 58, 168-190.
Chandra, D., Halonen, L.M., Linden, A.M., Procaccini, C., Hellsten, K., Homanics, G.E., and
Korpi, E.R. (2010). Prototypic GABA(A) receptor agonist muscimol acts preferentially
through forebrain high-affinity binding sites. Neuropsychopharmacology 35, 999-1007.
Chapman, E.R. (2008). How does synaptotagmin trigger neurotransmitter release? Annu. Rev.
Biochem. 77, 615–41.
Chavas, J., Forero, M.E., Collin, T., Llano, I., and Marty, A. (2004). Osmotic tension as a
possible link between GABAA receptor activation and intracellular calcium elevation.
Neuron, 44, 701-713.
Cheng, A., Gonçalves, J.T., Golshani, P., Arisaka, K., and Portera-Cailliau, C. (2011).
Simultaneous two-photon calcium imaging at different depths with spatiotemporal
multiplexing. Nature Methods. 8, 139-142.
Cherubini, E. (2012). Refuting the challenges of the developmental shift of polarity of GABA
actions: GABA more exciting than ever! Frontiers in Cell. Neurosci. 6.35, 1-18.
Chiu, C., Brickley, S., Jensen, K., Southwell, A., Mckinney, S., Cull-Candy, S., Mody, I., and
Lester, H.A. (2005). GABA transporter deficiency causes tremor, ataxia, nervousness,
and increased GABA-induced tonic conductance in cerebellum. J. Neurosci. 25, 32343245.
Choe, M., Ortiz-Mantilla, S., Makris, N., Gregas, M., Bacic, J., Haehn, D., Kennedy, D., Pienaar,
R., Caviness, Jr., V.S., Benasich, A.A., amd Grant, P.E. (2012). Regional infant brain
development: an MRI-based morphometric analysis in 3 to 13 month olds. Cerebral
Cortex. Advanced access published September 6, 2012. doi:10.1093/cercor/bhs197.

201

Choi, B.H., Yee, S., and Robles, M. (1996). The effects of glutathione glycoside in methyl
mercury poisoning. Toxicol Appl Pharmacol. 141,357-64.
Chung, Y.H., Shin, C.M., Kim, M.J., Shin, D.H., Yoo, Y.B., and Cha, C.I. (2001). Differential
alterations in the distribution of voltage-gated calcium channels in aged rat cerebellum.
Brain Res. 903, 247-252.
Chung, Y.H., Shin, C., Park, K.H., and Cha, C.I. (2000). Immunohistochemical study on the
distribution of neuronal voltage-gated calcium channels in the rat cerebellum. Brain Res.
865, 278-282.
Clancy, B., Darlington, R.B., and Finlay, B.L. Translating developmental time across
mammalian species. Neuroscience 105, 7-17.
Coccini, T., Roda, E., Castoldi, A.F., Poli, D., Goldoni, M., Vettori, M.V., Mutti, A., and Manzo,
L. (2011). Developmental exposure to methylmercury and 2,2’,4,4’,5,5’hexachlorobiphenyl (PCB153) affects cerebral dopamine D1-like and D2-like receptors
of weanling and pubertal rats. Arch. Toxicol. 85, 1281-1294.
Conti, R., Tan, Y.P., and Llano, I. (2004). Action potential-evoked and ryanodine-sensitive
spontaneous Ca2+ transients at the presynaptic terminal of a developing CNS inhibitory
synapse. J. Neurosci. 24, 6946-6957.
Coon, A.L., Wallace, D.R., Mactutus, C.F., and Booze, R.M. (1999). L-type calcium channels in
the hippocampus and cerebellum of Alzheimer's disease brain tissue. Neurobiol. Aging
20, 597-603.
Corda, M.G., 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, 264-269.
Csordás, G., Thomas, A.P., Hajnóczky, G. (1999). Quasi-synaptic calcium signal transmission
between endoplasmic reticulum and mitochondria. EMBO J. 18, 96-108.
Davidson, P.W., Sloane-Reeves, J., Myers, G.J., Hansen, O.N., Huang, L., Georger, L.A., Cox,
C., Thurston, S.W., Shamlaye, C.F., and Clarkson, T.W. (1998). Association between
prenatal exposure to methylmercury and visuospatial ability at 10.7 years in the
seychelles child development study. NeuroToxicology 29, 453-459.
Dahlem, Y.A., Wolf, G., Siemen, D., and Horn, T.F. (2004). Combined modulation of the
mitochondrial ATP-dependent potassium channel and the permeability transition pore
causes prolongation of the biphasic calcium dynamics. Cell Calcium 39, 387-400.

202

De Schutter, E. (2002). Cerebellar cortex: computation by extrasynaptic inhibition? Curr. Biol.
12, R363-365.
De Zeeuw, C.I., Hoebeek, F.E., Bosman, L.W.J., Schonewille, M., Witter, L., and Koekkoek,
S.K. (2011). Spatiotemporal firing patterns in the cerebellum. Nat. Rev.Neurosci. 12,
327-344.
Debes, F., Budtz-Jørgensen, E., Weihe, P., White, R.F., Grandjean, P. (2006). Impact of prenatal
methylmercury exposure on neurobehavioral function at age 14 years. Neurotox. and
Teratol. 28, 363-375.
Demarchi, F., and Schneider, C. (2007). The calpain system as a modulator of stress/damage
response. Cell Cycle 6, 136-138.
Deniaud, A., Sharaf el dein, O., Maillier, E., Poncet, D., Kroemer, G., Lemaire, C., and Brenner,
C. (2008). Endoplasmic reticulum stress induces calcium-dependent permeability
transition, mitochondrial outer membrane permeabilization and apoptosis. Oncogene 27,
285-299.
Denny, M.F., and Atchison, W.D. Methylmercury-induced elevations in intrasynaptosomal zinc
concentrations: an 19F-NMR study. J. Neurochem. 63, 383-386.
Denny, M.F., Hare, M.F., and Atchison, W.D. (1993). Methylmercury alters intrasynaptosomal
concentrations of endogenous polyvalent cations. Toxicol. Appl. Pharmacol. 122, 222232.
D’hulst, C., Atack, J.R., and Kooy, R.F. (2009). The complexity of the GABAA receptor shapes
unique pharmacological profiles. Drug Discovery Today 14, 866-875.
Ding, J., Zhang, L., Qu, F., Ren, X., Zhao, X., and Liu, Q. (2011). Cell activity analysis by
capillary zone electrophoresis combined with specific cell staining. Electrophoresis. 32,
455-463.
Doherty, D. (2009) Joubert syndrome: insights into brain development, cilium biology, and
complex disease. Semin. Pediatr. Neurol. 16, 143–154.
Doroshenko, P.A., Woppmann, A., Miljanich, G., and Augustine, G.J. (1997). Pharmacologically
distinct presynaptic calcium channels in cerebellar excitatory and inhibitory synapses.
Neuropharmacology 36, 865-872.
Drexler, B., Jurd, R., Rudolph, U., Antkowiak, B. (2009). Distinct actions of etomidate and
propofol at beta3-containing gamma-aminobutyric acid type A receptors.
Neuropharmacology 57, 446-455.

203

Earl, D.E. and Tietz, E.I. (2011). Inhibition of recombinant L-type voltage-gated calcium
channels by positive allosteric modulators of GABAA receptors. JPET. 337, 301-311.
Eccles, J.C., Ito, M., Szenta´gothai, J. (1967). The cerebellum as a neuronal machine.
Springer-Verlag, Berlin.
Edwards, J.R., Marty, M.S., and Atchison, W.D. (2005). Comparative sensitivity of rat cerebellar
neurons to dysregulation of divalent cation homeostasis and cytotoxicity caused by
methylmercury. Toxicol Appl Pharmacol 208, 222-232.
Egawa, K., Kitagawa, K., Inoue, K., Takayama, M., Takayama, C., Saitoh, S., Kishino, T.,
Kitagawa, M., and Fukuda, A. (2012). Decreased tonic inhibition in cerebellar granule
cells causes motor dysfunction in a mouse model of Angelman Syndrome. Science Trans.
Med. 4, 163ra157.
Eilers, J., Plant, T.D., Marandi, N., and Konnerth, A. (2001). GABA-mediated Ca2+ signalling
in developing rat cerebellar Purkinje neurones. J. Physiol. 536.2, 429-437.
Eimerl, S., and Schramm, M. (1991). Acute glutamate toxicity in cultured cerebellar granule
cells: agonist potency, effects of pH, Zn2+ and the potentiation by serum albumin. Brain
Res. 560, 282-290.
Engel, G.L., Delwig, A., and Rand, M.D. (2012). The effects of methylmercury on Notch
signaling during embryonic neural development in Drosophila melanogaster. Toxicol. in
Vitro. 26, 485-492.
Engelman, H.S., and MacDermott, A.B. (2004) Presynaptic ionotropic receptors and control of
transmitter release. Nat. Rev. Neurosci. 5, 135-145.
Erceg, S., Ronaghi, M., Zipancic, I., Lainez, S., Roselló, M.G., Xiong, C., Moreno-Manzano, V.,
Rodríguez-Jiménez, F.J., Planells, R., Alvarez-Dolado, M., Bhattacharya, S.S., and
Stojkovic, M. (2010). Efficient differentiation of human embryonic stem cells into
functional cerebellar-like cells. Stem Cells and Dev. 19, 1745-1756.
Erreger, K., Traynelis, S.F. (2008). Zinc inhibition of rat NR1/NR2A N-methyl-D-aspartate
receptors. J. Physiol. 586, 763-778.
Exton, J.H. (1990). Signaling through phosphatidylcholine breakdown. J. Biol. Chem. 265, 1-4.
Fahrion, J.K., Komuro, Y., Li, Y., Ohno, N., Littner, Y., Raoult, E., Galas, L., Vaudry, D., and
Komuro, H. (2012). Rescue of neuronal migration deficits in a mouse model of fetal
Minamata disease by increasing neuronal Ca2+ spike frequency. PNAS. 109, 5057-5062.

204

Fang, M., Li, J., Tiu, S.C., Zhang, L., Wang, M., Yew, D.T. (2005). N-methyl-D-aspartate
receptor and apoptosis in Alzheimer's disease and multiinfarct dementia. J. Neurosci.
Res. 81, 269-274.
Faro, L.R., do Nascimento, J.L., Alfonso, M., and Durán, R. (2002). Protection of
methylmercury effects on the in vivo dopamine release by NMDA receptor antagonists
and nitric oxide synthase inhibitors. Neuropharmacology 42, 612-618.
Farrant, M., and Brickley, S.G. (2003). Properties of GABAA receptor-mediated transmission at
newly formed Golgi-granule cell synapses in the cerebellum. Neuropharmacology. 44,
181-189.
Feng, H., Kao, C., Gallagher, M.J., Jansen, E.D., Mahadevan-Jansen, A., Konrad, P.E., and
Macdonald, R.L. (2010). Alteration of GABAergic neurotransmission by pulsed infrared
laser stimulation. J Neurosci. Methods. 30, 110-114.
Findeisen, F., and Minor, Jr., D.L. (2009). Disruption of the IS6-AID linker affects voltage-gated
Calcium channel inactivation and facilitation. J. Gen. Physiol. 133, 327-343.
Fogelson, A.L., and Zucker, R.S. (1985). Presynaptic calcium diffusion from various arrays of
single channels. Implications for transmitter release and synaptic facilitation. Biophys. J.
48, 1003-1017.
Fonfría, E., Rodríguez-Farré, E., and Suñol, C. (2001). Mercury interaction with the GABAA
receptor modulates the benzodiazepine binding site in primary cultures of mouse
cerebellar granule cells. Neuropharmacology. 41, 819-833.
Fonfria, E., Dare, E., Benelli, M. Suñol, C., and Ceccatelli, S. (2002). Translocation of
apoptosis-inducing factor in cerebellar granule cells exposed to neurotoxic agents
inducing oxidative stress. Eur J Neurosci 16, 2013-2016.
Fox, D.A., Grandjean, P., de Groot, D., and Paule, M.G. (2012). Developmental origins of adult
diseases and neurotoxicity: epidemiological and experimental studies. Neurotoxicology
33, 810-816.
Gall, D., Roussel, C., Susa, I., D'Angelo, E., Rossi, P., Bearzatto, B., Galas, M.C., Blum, D.,
Schurmans, S., and Schiffmann, S.N. (2003). Altered neuronal excitability in cerebellar
granule cells of mice lacking calretinin. J. Neurosci. 23, 9320-9327.
Gall, D., Roussel, C., Nieus, T., Cheron, G., Servais, L., D'Angelo, E., and Schiffmann, S.N.
(2005). Role of calcium binding proteins in the control of cerebellar granule cell neuronal
excitability: experimental and modeling studies. Prog Brain Res. 148, 321-8.

205

Galeffi, F., Sah, R., Pond, B.B., George, A., and Schwartz-Bloom, R.D. (2004). Changes in
intracellular chloride after oxygen-glucose deprivation of the adult hippocampal slice:
effect of diazepam. J. Neurosci. 24, 4478-4488.
Galofré, M., Babot, Z., García, D.A., Iraola, S., Rodríguez-Farré, E., Forsby, A., and Suñol, C.
(2010). GABAA receptor and cell membrane potential as functional endpoints in cultured
neurons to evaluate chemicals for human acute toxicology. Neurotoxicol. Teratol. 32, 5261.
Gault, L.M., and Siegel, R.E. (1997). Expression of the GABAA receptor δ subunit is selectively
modulated by depolarization in cultured rat cerebellar granule neurons. J. Neurosci. 17,
2391-2399.
Gassó, S., Cristòfol, R.M., Selema, G., Rosa, R., Rodríguez-Farré, E., and Sanfeliu, C. (2001).
Antioxidant compounds and Ca2+ pathway blockers differentially protect against
methylmercury and mercuric chloride neurotoxicity. J. Neurosci. Res. 66, 135-145.
Geppert, M., Goda, Y., Hammer, R.E., Li, C., Rosahl, T.W., Stevens, C.F., and Sudhof T.C.
(1994). Synaptotagmin I: a major Ca2+ sensor for transmitter release at a central synapse.
Cell 79, 717-727.
Gibon, J., Tu, P., Frazzini, V., Sensi, S.L., and Bouron, A. (2010). The thiol-modifying agent
N-ethylmaleimide elevates the cytosolic concentration of free Zn(2+) but not of Ca(2+)
in murine cortical neurons. Cell Calcium 48, 37-43.
Glazner, G.W., and Mattson, M.P. (2000). Differential effects of BDNF, ADNF9, and TNFalpha
on levels of NMDA receptor subunits, calcium homeostasis, and neuronal vulnerability
to excitotoxicity. Exp. Neurol. 161, 442-452.
Gleichmann, M., and Mattson, M.P. (2011). Neuronal calcium homeostasis and dysregulation.
Antioxid. Redox Signal. 14, 1261-73.
Grabrucker, A.M., Knight, M.J., Proepper, C., Bockmann, J., Joubert, M., Rowan, M., Nienhaus,
G.U., Garner, C.C., Bowie, J.U., Kreutz, M.R., Gundelfinger, E.D., and Boeckers, T.M.
(2011). Concerted action of zinc and ProSAP/Shank in synaptogenesis and synapse
maturation. EMBO J. 30, 569-581.
Grishin, A., Li, H., Levitan, E.S., and Zaks-Makhina, (2006). Identification of γ-aminobutyric
acid receptor-interacting factor 1 (TRAK2) as a trafficking factor for the K+ channel
Kir2.1. J. Biol. Chem. 281, 30104-30111.
Hajela, R.K., Peng, S.Q., and Atchison, W.D. (2003). Comparative effects of methylmercury and
Hg2+ on human neuronal N- and R- type high-voltage activated calcium channels

206

transiently expressed in Human Embryonic Kidney 293 cells. J Pharmacol Exp Ther 306,
1129-1136.
Haller, V.L., Bernstein, M.A., and Welch, S.P. (2005). Chronic morphine treatment decreases
the CaV1.3 subunit of the L-type calcium channel. Eur. J. Pharmacol. 578, 101-107.
Harada, M. (1978). Congenital Minamata disease: intrauterine methylmercury poisoning.
Teratology. 18, 285-288.
Hare, M.F., McGinnis, K.M., and Atchison, W.D. (1993). Methylmercury increases intracellular
concentrations of Ca++ and heavy metals in NG108-15 cells. J. Pharmacol. Exp. Ther.
266, 1626-1635.
Hashimoto, M., and Hiri, M. (2012). Development and evolution of cerebellar neural circuits.
Develop. Growth Differ. 54, 373–389.
Heath, J.C., Banna, K.M., Reed, M.N., Pesek, E.F., Cole, N., Li, J., and Newland, M.C. (2010).
Dietary selenium protects against selected signs of aging and methylmercury exposure.
Neurotoxicology 31, 169-179.
Helmcke, K.J., Aschner, M. (2010). Hormetic effect of methylmercury on Caenorhabditis
elegans. Toxicol. Appl.Pharmacol. 248, 156-164.
Herden, C.J., Pardo, N.E., Hajela, R.K., Yuan, Y., and Atchison, W.D. (2008). Differential
effects of methylmercury on γ-aminobutyric acid type A receptor currents in rat
cerebellar granule and cerebral cortical neurons in culture. J Pharmacol Exp Ther 324,
517-528.
Homanics, G.E., Ferguson, C., Quinlan, J.J., Daggett, J., Snyder, K., Lagenaur, C., Mi, Z.P.,
Wang, X.H., Grayson, D.R., Firestone, L.L. (1997) Gene knockout of the alpha6 subunit
of the gamma-aminobutyric acid type A receptor: lack of effect on responses to ethanol,
pentobarbital, and general anesthetics. Mol. Pharmacol. 51, 588-596.
Hoogland, T.M., Kuhn, B., and Wang, S.S.-H. (2011). Preferential loading of Bergmann glia
with synthetic acetoxymethyl calcium dyes. Cold Spring Harb. Protoc. 2011, 1228-1231.
Houston, C.M., McGee, T.P., MacKenzie, G., Troyano-Cuturi, K., Mateos Rodriguez, P.,
Kutsarova, E., Diamanti, E., Hosie, A.M., Franks, N.P., and Brickley, S.G. (2012). Are
extrasynaptic GABAA receptors important targets for sedative/hypnotic drugs? J.
Neurosci. 32, 3887-3897.
Huang, C., Liu, S., Hsu, C., Lin-and Shiau, S. (2011). Neurotoxicological effects of low-dose
methylmercury and mercuric chloride in developing offspring mice. Toxicol. Lett. 201,
196-204.

207

Huang, W.C., Young, J.S., and Glitsch, M.D. (2007). Changes in TRPC channel expression
during postnatal development of cerebellar neurons. Cell Calcium 42, 1-10.
Hunter, D., and Russell, D.S. (1954). Focal cerebral and cerebellar atrophy in a human subject
due to organic mercury compounds. J Neurol Neurosurg Psychiat 17, 235-241.
Inada, H., Watanabe, M., Uchida, T., Ishibashi, H., Wake, H., Nemoto, T., Yanagawa, Y.,
Fukuda, A., and Nabekura, J. (2011). GABA regulates the multidirectional tangential
migration of GABAergic interneurons in living neonatal mice. PLoS One. 6, e27048.
Inoue, T., and Bryant, B.P. (2005). Multiple types of sensory neurons respond to irritating
volatile organic compounds (VOCs): Calcium fluorimetry of trigeminal ganglion
neurons.
Isaev, N.K., Lozier, E.R., Novikova, S.V., Silachev, D.N., Zorov, D.B., and Stelmashook, E.V.
(2012). Glucose starvation stimulates Zn2+ toxicity in cultures of cerebellar granule
neurons. Brain Res. Bull. 87, 80-84.
Ito, M. (2006). Cerebellar circuitry as a neuronal machine. Prog. In Neurobiol. 78, 272-303.
Ito, M. (2008). Control of mental activities by internal models in the cerebellum. Nat. Rev.
Neurosci. 9, 304-313.
Jacob, T.C., Moss, S.J., and Jurd, R. (2008). GABA(A) receptor trafficking and its role in the
dynamic modulation of neuronal inhibition. Nat. Rev. Neurosci. 9, 331-43.
Jiang, D., Sullivan, P.G., Sensi, S.L., Steward, O., Weiss, J.H. (2001). Zn(2+) induces
permeability transition pore opening and release of pro-apoptotic peptides from neuronal
mitochondria. J. Biol. Chem. 276, 47524-47529.
Johnson, F.O., Yuan, Y., Hajela, R.K., Chitrakar, A., Parsell, D.M., and Atchison, W.D. (2011).
Exposure to an environmental neurotoxicant hastens the onset of amyotrophic lateral
sclerosis-like phenotype in human Cu2+/Zn2+ superoxide dismutase 1 G93A mice:
glutamate-mediated excitotoxicity. J. Pharmacol. Exp. Ther. 338, 518-527.
Joksimović, S., Varagic, Z., Kovačević, J., Van Linn, M., Milić, M., Rallapalli, S., Timić, T.,
Sieghart, W., Cook, J.M., and Savić, M.M. (2013). Insights into functional pharmacology
of α1 GABAA receptors: how much does partial activation at the benzodiazepine site
matter? Psychopharmacology. Advanced Access published May 18, 2013.
doi:10.1007/s00213-013-3143-4.
Jones, R.S., and Heinemann, U.H. (1987). Differential effects of calcium entry blockers on preand postsynaptic influx of calcium in the rat hippocampus in vitro. Brain Res. 416, 257266.

208

Jones, A., Korpi, E.R., McKernan, R.M., Pelz, R., Nusser, Z., Mäkelä, R., Mellor, J.R., Pollard,
S., Bahn, S., Stephenson, F.A., Randall, A.D., Sieghart, W., Somogyi, P., Smith, A.J.H.,
and Wisden, W. (1997). Ligand-gated ion channel subunit partnerships: GABAA
receptor α6 subunit gene inactivation inhibits δ subunit expression. J. Neurosci. 17, 13501362.
Joyce, C.J. (2007). In silico comparative genomic analysis of GABAA receptor transcriptional
dynamics. BMC Genomics 8, 203.
Juárez, B.I., Portillo-Salazar, H., González-Amaro, R., Mandeville, P., Aguirre, J.R., and
Jiménez, M.E. (2005). Participation of N-methyl-D-aspartate receptors on
methylmercury-induced DNA damage in rat frontal cortex. Toxicology 207, 223-9.
Kass, G.E., and Orrenius, S. (1999). Calcium signaling and cytotoxicity. Environ. Health
Perspect. 107, 25-35.
Kasischke, K., Büchner, M., Ludolph, A.C., and Riepe, M.W. (2001). Nuclear shrinkage in live
mouse hippocampal slices. Acta Neuropathol. 101, 483-490.
Kerper, L.E., Ballatori, N., and Clarkson, T.W..(1992) Methylmercury transport across the
blood-brain barrier by an amino acid carrier. Am. J. Physiol. 262, R761-765.
Khom, S., Strommer, B., Ramharter, J., Schwarz, T., Schwarzer, C., Erker, T., Ecker, G.F.,
Mulzer, J., and Hering, S. (2010). Valerenic acid derivatives as novel subunit-selective
GABAA receptor ligands - in vitro and in vivo characterization. Br. J. Pharmacol. 161,
65-78.
Knight, M.M., Roberts, S.R., Lee, D.A., and Bader, D.L. (2003). Live cell imaging using
confocal microscopy induces intracellular calcium transients and cell death. Am J Physiol
Cell Physiol 284, C1083–C1089.
Komuro, H., and Kumada, T. (2005). Ca2+ transients control CNS neuronal migration. Cell
Calcium 37, 387-393.
Komuro, H., and Rakic, P. (1992). Selective role of N-type calcium channels in neuronal
migration. Science 257, 806-809.
Komuro, H., and Rakic, P. (1993). Modulation of neuronal migration by NMDA receptors.
Science 260, 95-97.
Komuro, H., and Rakic, P. (1998). Distinct modes of neuronal migration in different domains of
developing cerebellar cortex. J Neurosci 18,1478-1490.

209

Korpi, E.R., Koikkalainen, P., Vekovischeva, O.Y., Mäkelä, R., Kleinz, R., Uusi-Oukari, M.,
and Wisden, W. (1999). Cerebellar granule-cell-specific GABAA receptors attenuate
benzodiazepine-induced ataxia: evidence from alpha 6-subunit-deficient mice. Eur. J.
Neurosci. 11, 233-240.
Kuhn, B., Hoogland, T.M., and Wang, S.S.-H. (2011). Cerebellar craniotomy for in vivo calcium
imaging of astrocytes. Cold Spring Harb. Protoc. 2011, 1224-1227.
Kukulski, W., Schorb, M., Welsch, S., Picco, A., Kaksonen, M., and Briggs, J.A.G. (2011).
Correlated fluorescence and 3D electron microscopy with high sensitivity and special
precision. J. Cell Biol. 192, 111-119.
Kullman, D.M., Ruiz, A., Rusakov, D.M., Scott, R., Semyanov, A., and Walker, M.C. (2005).
Presynaptic, extrasynaptic and axonal GABAA receptors in the CNS: where and why?
Prog. Biophys. Mol. Bio. 87, 33-46.
Kumada, T., and Komuro, H. (2004). Completion of neuronal migration regulated by loss of
calcium transients. PNAS 101, 8479-8484.
Kumada, T., Komuro, Y., Li, Y., Hu, T., Wang, Z., Littner, Y., and Komuro, H. (2010).
Inhibition of cerebellar granule cell turning by alcohol. Neuroscience 170, 1328-1344.
Laha, K.T., and Wagner, D.A. (2011). A state-dependent salt-bridge interaction exists across the
β/α inter-subunit interface of the GABAA receptor. Mol. Pharmacol. 79, 662-671.
Landfield, P.W., Pitler, T.A. (1984). Prolonged Ca2+-dependent afterhyperpolarizations in
hippocampal neurons of aged rats. Science 226, 1089-1092.
Laure-Kamionowska, M., and Maślińska, D. (2011). Cerebellar cortical neurons misplaced in the
white matter due to disturbed migration during development of human brain. Folia
Neuropathol. 49, 282-294.
Leclerc, C., Néant, I., and Moreau, M. (2011). Early neural development in vertebrates is also a
matter of calcium. Biochimie 93, 2102-2111.
Lee, H.H.C., Walker, J.A., Williams, J.R., Goodier, R.J., Payne, J.A., and Moss, S.J. (2007).
Direct protein kinase C-dependent phosphorylation regulates the cell surface stability and
activity of the potassium chloride cotransporter KCC2. J. Biol. Chem. 282, 29777-29784.
Lehoullier, P.F., and Ticku, M.K. (1989). The pharmacological properties of GABA receptorcoupled chloride channels using 36Cl-influx in cultured spinal cord neurons. Brain Res.
487, 205-214.

210

Letts, V.A., Felix, R., Biddlecome, G.H., Arikkath, J., Mahaffey, C.L., Valenzuela, A., Bartlett,
F.S. 2nd., Mori, Y., Campbell, K.P., and Frankel, W.N. (1998). The mouse stargazer gene
encodes a neuronal Ca2+-channel gamma subunit. Nat. Genet. 19, 340-347.
Levesque, P.C., and Atchison, W.D. (1987). Interactions of mitochondrial inhibitors with
methylmercury on spontaneous quantal release of acetylcholine. Toxicol. Appl.
Pharmacol. 87, 315-324.
Limke, T.L., Bearss, J.J., and Atchison, W.D. (2004). Acute exposure to methylmercury causes
Ca2+ dysregulation and neuronal cell death in rat cerebellar granule cells through an M3
muscarinic receptor-linked pathway. Toxicol Sci 80, 60-68.
Limke, T.L., and Atchison, W.D. (2002). Acute exposure to methylmercury opens the
mitochondrial permeability transition pore in rat cerebellar granule cells. Toxicol Appl
Pharmacol 178, 52-61.
Limke, T.L., Otero-Montañez, J.K., and Atchison, W.D. (2003). Evidence for interactions
between intracellular calcium stores during methylmercury-induced intracellular calcium
dysregulation in rat cerebellar granule cells. J Pharmacol Exp Ther 304, 949-958.
Lo, W., Lagrange, A.H., Hernandez, C.C., Harrison, R., Dell, A., Haslam, S.M., Sheehan, J.H.,
and Macdonald, R.L. (2010). Glycosylation of β2 subunits regulates GABAA receptor
biogenesis and channel gating. J. Biol. Chem. 285, 31348-31361.
Long, P., Mercer, A., Begum, R., Stephens, G.J., Sihra, T.S., and Jovanovic, J.N. (2009). Nerve
terminal GABAA receptors activate Ca2+/Calmodulin-dependent signaling to inhibit
voltage-gated Ca2+ influx and glutamate release. J. Biol. Chem. 284, 8726-8737.
Lyons, H.R., Land, M.B., Gibbs, T.T., and Farb, D.H. (2001). Distinct signal transduction
pathways for GABA-induced GABAA receptor down-regulation and uncoupling in
neuronal culture: a role for voltage-gated calcium channels. J. Neurochem. 78, 11141125.
MacDermott, A.B., Role, L.W., and Siegelbaum, S.A. (1999). Presynaptic ionotropic receptors
and the control of transmitter release. Annu. Rev. Neurosci. 22, 443-485.
Macdonald, R.L., Kang, J., and Gallagher, M.J. (2010). Mutations in GABAA receptor subunits
associated with genetic epilepsies. J Physiol. 588.11, 1861-1869.
Magos, L., and Clarkson, T.W. (2006) Overview of the clinical toxicity of mercury. Ann Clin
Biochem. 43, 257-68.
Maguire, J., Ferando, I., Simonsen, C., and Mody, I. (2009) Excitability changes related to
GABAA receptor plasticity during pregnancy. J. Neurosci. 29, 9592-601.

211

Mahapatra, C.T., Bond, J., Rand, D.M., and Rand, M.D. (2010). Identification of methylmercury
tolerance gene candidates in Drosophila. Toxicol. Sci. 116, 225-238.
Mäkelä, R., Uusi-Oukari, M., Homanics, G.E., Quinlan, J.J., Firestone, L.L., Wisden, W., and
Korpi, E.R. (1997). Cerebellar gamma-aminobutyric acid type A receptors:
pharmacological subtypes revealed by mutant mouse lines. Mol. Pharmacol. 52, 380-388.
Mancini, J.D., Autio, D.M., and Atchison, W.D. (2009). Continuous exposure to low
concentrations of methylmercury impairs cerebellar granule cell migration in organotypic
slice culture. NeuroToxicology 30, 203-208.
Manev, H., Kharlamov, E., Uz, T., Mason, R.P., Cagnoli, C.M. (1997). Characterization of
zinc-induced neuronal death in primary cultures of rat cerebellar granule cells. Exp.
Neurol. 146, 171-178.
Mao, X., Ma, P., Cao, D., Sun, C., Ji, Z., Min, D., Sun, H., Xie, N., Cai, J., and Cao, Y. (2011).
Altered expression of GABAA receptors (α4, γ2 subunit), potassium chloride
cotransporter 2 and astrogliosis in tremor rat hippocampus. Brain Res. Bull. 25, 373-379.
Marchionni, I., Kasap, Z., Mozrzymas, J.W., Sieghart, W., Cherubini, E., and Zacchi, P. (2009).
New insights on the role of gephyrin in regulating both phasic and tonic GABAergic
inhibition in rat hippocampal neurons in culture. Neuroscience. 164, 552-562.
Marczynski, T.J. (1998). GABAergic deafferentation hypothesis of brain aging and Alzheimer’s
disease revisited. Brain Res. Bull 45, 341-379.
Marty, M.S., and Atchison, W.D. (1997). Pathways mediating Ca2+ entry in rat cerebellar
granule cells following in vitro exposure to methyl mercury. Toxicol Appl Pharmacol.
147, 319-30.
Marty, M.S., and Atchison, W.D. (1998). Elevations of Ca2+ as a probable contributor to
decreased viability in cerebellar granule cells following acute exposure to
methylmercury. Toxicol Appl Pharmacol 150, 98-105.
Mehlin, C., Crittenden, C., and Andreyka, J. (2003). No-wash dyes for calcium flux
measurement. BioTechniques 34, 164-166.
Mennerick, S., He, Y., Jiang, X., Manion, B.D., Wang, M., Shute, A., Benz, A., Evers, A.S.,
Covey, D.F., Zorumski, C.F. (2005). Selective antagonism of 5alpha-reduced
neurosteroid effects at GABA(A) receptors. Mol. Pharmacol. 65, 1191-1197.
Métin, C., Vallee, R.B., Rakic, P., and Bhide, P.G. (2008). Modes and mishaps of neuronal
migration in the mammalian brain. J. Neurosci. 28, 11746-11752.

212

Micheva, K.D., Busse, B., Weiler, N.C., O’Rourke, N., and Smith, S.J. (2010). Single-synapse
analysis of a diverse synapse population: proteomic imaging methods and markers.
Neuron. 68, 639-653.
Minnema, D.J., Cooper, G.P., and Greenland, R.D. (1989). Effects of methylmercury on
neurotransmitter release from rat brain synaptosomes. Toxicol. Appl. Pharmacol. 99, 510521.
Morishima, T., Uematsu, M., Furukawa, T., Yanagawa, Y., Fukuda, A., and Yoshida, S. (2010).
GABA imaging in brain slices using immobilized enzyme-linked photoanalysis.
Neurosci. Res. 67, 347-353.
Myers, G.J., Thurston, S.W., Pearson, A.T., Davidson, P.W., Cox, C., Shamlaye, C.F.,
Cernichiari, E., and Clarkson, T.W. (2009). Postnatal exposure to methyl mercury from
fish consumption: a review and new data from the Seychelles Child Development Study.
Neurotoxicology 30, 338-349.
Nusser, Z., Ahmad, Z., Tretter, V., Fuchs, K., Wisden, W., Sieghart, W., and Somogyi, P.
(1999). Alterations in the expression of GABAA receptor subunits in cerebellar granule
cells after disruption of the α6 subunit gene. Eur. J. Neurosci. 11, 1685-1697.
Nusser, Z., Roberts, J.D., Baude, A., Richards, J.G., and Somogyi, P. (1995). Relative densities
of synaptic and extrasynaptic GABAA receptors on cerebellar granule cells as
determined by a quantitative immunogold method. J. Neurosci. 15, 2948-2960.
Nusser, Z., Sieghart, W., and Somogyi, P. (1998). Segregation of different GABAA receptors to
synaptic and extrasynaptic membranes of cerebellar granule cells. J Neurosci 18, 16931703.
Ohtsuki, G., and Hirano, T. (2008). Bidirectional plasticity at developing climbing fiber-Purkinje
neuron synapses. Eur. J. Neurosci. 28, 2393-2400.
Olsen, R.W., Bureau, M.H., Endo, S., and Smith, G. (1991). The GABAA receptor family in the
mammalian brain. Neurochem. Res. 16, 317-325.
Olsen, R.W., and Sieghart, W. (2009). GABA A receptors: subtypes provide diversity of
function and pharmacology. Neuropharmacology. 56,141-8.
Olsen, R.W., and Tobin, A.J. (1990). Molecular biology of GABAA receptors. FASEB J. 4,
1469-1480.

213

Onishchenko, N., Karpova, N., Sabri, F., Castrén, E., and Ceccatelli, S. (2008). Long-lasting
depression-like behavior and epigenetic changes of BDNF gene expression induced by
perinatal exposure to methylmercury. J. Neurochem. 106, 1378-1387.
Ophoff, R.A., Terwindt, G.M., Vergouwe, M.N., van Eijk, R., Oefner, P.J., Hoffman, S.M.,
Lamerdin, J.E., Mohrenweiser, H.W., Bulman, D.E., Ferrari, M., Haan, J., Lindhout, D.,
van Ommen, G.J., Hofker, M.H., Ferrari, M.D., Frants, R.R. (1996). Familial hemiplegic
migraine and episodic ataxia type-2 are caused by mutations in the Ca2+ channel gene
CACNL1A4. Cell 87, 543-552.
Ortinski, P.I., Lu, C., Takagaki, K., Fu, Z., and Vicini, S. (2004). Expression of distinct alpha
subunits of GABAA receptor regulates inhibitory synaptic strength. J. Neurophysiol. 92,
1718-1727.
Ou, Y.C., White, C.C., Krejsa, C.M., Ponce, R.A., Kavanagh, T.J., and Faustman, E.M. (1999).
The role of intracellular glutathione in methylmercury-induced toxicity in embryonic
neuronal cells. Neurotoxicology 20, 793-804.
Owens, D.F., Boyce, L.H., Davis, M.B.E., and Kriegstein, A.R. (1996). Excitatory GABA
responses in embryonic and neonatal cortical slices demonstrated by gramicidin
perforated-patch recordings and calcium imaging. J. Neurosci. 16, 6414-6423.
Paletz, E.M., Day, J.J., Craig-Schmidt, M.C., and Newland, M.C. (2007). Spatial and visual
discrimination reversals in adult and geriatric rats exposed during gestation to
methylmercury and n-3 polyunsaturated fatty acids. Neurotoxicology. 28, 707-19.
Pancani, T., Anderson, K.L., Porter, N.M., Thibault, O. (2011). Imaging of a glucose analogue,
calcium and NADH in neurons and astrocytes: dynamic responses to depolarization and
sensitivity to pioglitazone. Cell Calcium. 50, 548-558.
Pardo, N.E., Hajela, R.K., and Atchison, W.D. (2006). Acetylcholine release at neuromuscular
junctions of adult tottering mice is controlled by N- (Cav2.2) and R-type (Cav2.3) but not
L-type (Cav1.2) Ca2+ channels. JPET. 319, 1009-1020.
Payne, H.L., Connelly, W.M., Ives, J.H., Lehner, R., Furtmuller, B., Sieghart, W., Tiwari, P.,
Lucocq, J.M., Lees, G., and Thompson, C.L. (2007). GABAA α6-containing receptors
are selectively compromised in cerebellar granule cells of the ataxic mouse, Stargazer. J.
Biol. Chem. 282, 29130-29143.

214

Payne, H.L., Ives, J.H., Sieghart, W., and Thompson, C.L. (2008). AMPA and kainate receptors
mediate mutually exclusive effects on GABAA receptor expression in cultured mouse
cerebellar granule neurones. J. Neurochem. 104, 173-186.
Peça, J., Feliciano, C., Ting, J.T., Wang, W., Wells, M.F., Venkatraman, T.N., Lascola, C.D., Fu,
Z., Feng, G. (2011). Shank3 mutant mice display autistic-like behaviours and striatal
dysfunction. Nature 472, 437-442.
Pelkey, K.A., Topolnik, L., Lacaille, J.C., and McBain, C.J. (2006). Compartmentalized Ca(2+)
channel regulation at divergent mossy-fiber release sites underlies target cell-dependent
plasticity. Neuron 52, 497-510.
Peng, S.Q., Hajela, R.K., and Atchison, W.D. (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, 424-432.
Petersen, M.S., Halling, J., Bech, S., Wermuth, L., Weihe, P., Nielsen, F., Jørgensen, P.J.,
Budtz-Jørgensen, E., and Grandjean, P. (2008). Impact of dietary exposure to food
contaminants on the risk of Parkinson's disease. Neurotoxicology 29, 584-590.
Petroni, D., Tsai, J., Mondal, D., and George, W. (2011). Attenuation of low dose
methylmercury and glutamate induced-cytotoxicity and tau phosphorylation by an
N-methyl-D-aspartate antagonist in human neuroblastoma (SHSY5Y) cells. Environ.
Toxicol. Early Access: doi: 10.1002/tox.20765.
Poulter, M.O., Du, L., Zhurov, V., Merali, Z., and Anisman, H. (2009) Plasticity of the
GABA(A) receptor subunit cassette in response to stressors in reactive versus resilient
mice. Neuroscience 165, 1039-51.
Prodanov, D., Heeroma, J., and Marani E. (2006). Automatic morphometry of synaptic boutons
of cultured cells using granulometric analysis of digital images. J. Neurosci. Methods
151, 168-177.
Pugh, J.R., and Jahr, C.E. (2011). Axonal GABAA receptors increase cerebellar granule cell
excitability and synaptic activity. J. Neurosci. 31, 565-574.
Qian, Z., Yakhnitsa, V., and Barmack, N.H. (2011). Climbing fiber-evoked Purkinje cell
discharge reduces expression of GABA(A) receptor-associated protein (GABARAP) and
decreases its interaction with GABA(A) receptors. J. Neurochem. 117, 197-208.

215

Raiteri, L., Schmid, G., Prestipino, S., Ratieri, M., and Bonanno, G. (2001). Activation of α6
GABAA receptors on depolarized cerebellar parallel fibers elicits glutamate release
through anion channels. Neuropharmacology. 41, 943-951.
Rakotomamonjy, J., Levenes, C., Bailieu, E.E., Schumacher, M., and Ghoumari, A.M. (2011)
Novel protective effect of mifepristone on detrimental GABAA receptor activity in
immature Purkinje cells. FASEB J. 25, 3999-4010.
Rand, M.D. (2010). Drosophotoxicology: the growing ppotential for Drosophila in
neurotoxicology. Neurotoxicol. Teratol. 32, 74.
Rasband, W.S. “ImageJ”, US National Institutes of Health, Bethesda, Maryland, USA,19972011. rsb.info.nih.gov/ij/index.html Accessed 17 April 2012.
Rasband, W., Dougherty, B., Collins, T., Karperian, A. and Stervbo, U. “Multi-Measure”
Optinav, Inc. 2006. www.optinav.com/imagej.html accessed 17 April 2012.
Rice, D., and Barone, S. Jr. (2000). Critical periods of vulnerability for the developing nervous
system: evidence from human and animal models. Environ Health Perspect 108,
Suppl:511-533.
Rizzuto, R., Bastianutto, C., Brini, M., Murgia, M., and Pozzan, T. (1994). Mitochondrial Ca2+
homeostasis in intact cells. J. Cell Biol. 126, 1183-1194.
Roda, E., Coccini, T., Acerbi, D., Castoldi, A., Bernocchi, G., and Manzo, L. (2008). Cerebellum
cholinergic muscarinic receptor (subtype-2 and -3) and cytoarchitecture after
developmental exposure to methylmercury: an immunohistochemical study in rat. J
Chem Neuroanat 35, 285-294.
Roda, E., Manzo, L., and Coccini, T. (2012) Application of chemical markers for assessing
health effects after developmental methylmercury and PCB coexposure. J. Toxicol. 2012,
216032.
Rüegg, P.C., and Nelson, D.J. (1989). Safety and efficacy of isradipine, alone and in
combination, in the treatment of angina pectoris. Am. J. Med. 86, 70-74.
Ruiz, A., Matute, C., and Alberdi, E. (2009) Endoplasmic reticulum Ca(2+) release through
ryanodine and IP(3) receptors contributes to neuronal excitotoxicity. Cell Calcium 46,
273-281.
Sabatini, B.L., and Regehr, W.G. (1995). Detecting changes in calcium influx which contribute
to synaptic modulation in mammalian brain slice. Neuropharmacology 34, 1453-1467.
216

Sainlos, M., Tigaret, C., Poujol, C., Olivier, N.B., Bard, L., Breillat, C., Thiolon, K., Choquet,
D., and Imperialli, B. (2010). Biomimetic divalent ligands for the acute disruption of
synaptic AMPAR stabilization. Nature Chem. Biol. 7, 81-91.
Sakamoto, M., Ikegami, N., and Nakano A. (1996). Protective effects of Ca2+ channel blockers
against methyl mercury toxicity. Pharmacol. Toxicol. 78, 193-199.
Sakamoto, M., Wakabayashi, K., Kakita, A., Takahashi, H., Adachi, T., and Nakano, A. (1998).
Widespread neuronal degeneration in rats following oral administration of
methylmercury during the postnatal development stage: a model of fetal-type Minamata
disease. Brain Res 784, 351-354.
Sakamoto, M., Kakita, A., de Oliviera, R.B., Pan, H.S., and Takahashi, H. (2004). Dosedependent effects of methylmercury administered during neonatal brain spurts in rats.
Dev Brain Res 152, 171-176.
Sanna, E., Mascia, M.P., Klein, R.L., Whiting, P.J., Biggio, G., and Harris, R.A. (1995). Actions
of the general anesthetic propofol on recombinant human GABAA receptors: influence of
receptor subunits. JPET 274, 353-360.
Sanna, E., Murgia, A., Casula, A., Usala, M., Maciocco, E., Tuligi, G., and Biggio, G. (1996).
Direct activation of GABAA receptors by loreclezole, an anticonvulsant drug with
selectivity for the beta-subunit. Neuropharmacology 35, 1753-1760.
Sarafian, T., and Verity, M.A. (1991). Oxidative mechanisms underlying methyl mercury
neurotoxicity. Int. J. Dev. Neurosci. 9, 147-153.
Sauter, A., and Rudin, M. (1990). Calcium antagonists for reduction of brain damage in stroke. J.
Cardiovasc. Pharmacol. 15 Suppl. 1, S43-S47.
Sauter, A., and Rudin, M. (1991). Prevention of stroke and brain damage with calcium
antagonists in animals. Am. J. Hypertens. 4, 121S-127S.
Saxena, N.C., and Macdonald, R.L. (1996). Properties of putative cerebellar gammaaminobutyric acid A receptor isoforms. Mol Pharmacol 49, 567-579.
Schmid, G., Chittolini, R., Raiteri, L., and Bonanno, G. (1999). Differential effects of zinc on
native GABA(A) receptor function in rat hippocampus and cerebellum. Neurochem. Int.
34, 399-405.
Schneggenberger, R., Han, Y., and Kochubey, O. (2012). Ca2+ channels and transmitter release
at the active zone. Cell Calcium 52, 199-207.
217

Schonewille, M., Belmeguenai, A., Koekkoek, S.K., Houtman, S.H., Boele, H.J., van Beugen,
B.J., Gao, Z., Badura, A., Ohtsuki, G., Amerika, W.E., Hosy, E., Hoebeek, F.E.,
Elgersma, Y., Hansel, C., and De Zeeuw, C.I. (2010). Purkinje cell-specific knockout of
the protein phosphatase PP2B impairs potentiation and cerebellar motor learning.
Neuron. 67, 618-628.
Scott, J.A., Hamzelou, K.S., Rajagopalon, V., Habas, P.A., Kim, K., Barkovich, A.J., Glenn,
O.A., and Studholme, C. (2012). 3D morphometric analysis of human fetal cerebellar
development. Cerebellum, 11, 761-770.
Sensi, S.L., Yin, H.Z., and Weiss, J.H. (2000). AMPA/kainate receptor-triggered Zn2+ entry into
cortical neurons induces mitochondrial Zn2+ uptake and persistent mitochondrial
dysfunction. Eur. J. Neurosci. 12, 3813-3818.
Shafer, T.J., and Atchison, W.D. (1991). Methylmercury blocks N- and L-type Ca++ channels in
nerve growth factor-differentiated pheochromocytoma (PC12) cells. JPET 258, 149-157.
Shafer, T.J., Bushnell, P.J., Benignus, V.A., and Woodward, J.J. (2005). Perturbation of voltagesensitive Ca2+ channel function by volatile organic solvents. JPET. 315, 1109-1118.
Shevtsova, O., and Leitch, B. (2013). Selective loss of AMPA receptor subunits at inhibitory
neuron synapses in the cerebellum of the ataxic stargazer mouse. Brain Res. 1427, 54-64.
Shoshan-Barmatz, V., Zalk, R., Gincel, D., and Vardi, N. (2004). Subcellular localization of
VDAC in mitochondria and ER in the cerebellum. Biochim. Biophys. Acta. 1657, 105114.
Shrivastava, A.N., Triller, A., and Sieghart, W. (2011). GABAA receptors: post-synaptic
colocalization and cross-talk with other receptors. Fronteirs in Cell. Neurosci. 5.7, 1-12.
Sigel, E., and Baur, R. (2000). Electrophysiological evidence for the coexistence of alpha1 and
alpha6 subunits in a single functional GABA(A) receptor. J. Neurochem. 74, 2590-2596.
Simeone, T.A., Wilcox, K.S., and White, H.S. (2006). Subunit selectivity of topiramate
modulation of heteromeric GABA(A) receptors. Neuropharmacology 50, 845-857.
Sirois, J.E., and Atchison, W.D. (1996) Effects of mercurials on ligand- and voltage-gated ion
channels: a review. NeuroToxicology 17, 63-84.
Sirois, J.E., and Atchison, W.D. (2000). Methylmercury affects multiple subtypes of calcium
channels in rat cerebellar granule cells. Toxicol Appl Pharmacol 167, 1-11.

218

Smith, A.J., Oxley, B., Malpas, S., Pillai, G.V., and Simpson, P.B. (2004). Compounds
exhibiting selective efficacy for different beta subunits of human recombinant
gamma-aminobutyric acid A receptors. JPET 311, 601-609.
Spät, A., Szanda, G., Csordás, G., and Hajnóczky, G. (2008). High- and low-calcium-dependent
mechanisms of mitochondrial calcium signalling. Cell Calcium 44, 51-63.
Spurny, R., Ramerstorfer, J., Price, K., Brams, M., Ernst, M., Nury, H., Verheij, M., Legrand, P.,
Bertrand, D., Bertrand, S., Dougherty, D.A., de Esch, I.J.P., Corringer, P., Sieghart, W.,
Lummis, S.C.R., and Ulens, C. (2012). Pentameric ligand-gated ion channel ELIC is
activated by GABA and modulated by benzodiazepines. PNAS. 109, E3028-E3034.
Stanley, E.F. (1993). Single calcium channels and acetylcholine release at a presynaptic nerve
terminal. Neuron 11, 1007-1011.
Steel, R.G.D., and Torrie, J.H. (1960). Principles and procedures of statistics. New York:
McGraw.
Stern, S., Cox, C., Cernichiari, E., Balys, M., and Weiss, B. (2001). Perinatal and lifetime
exposure to methylmercury in the mouse: blood and brain concentrations of mercury to
26 months of age. Neurotoxicology 22, 467-477.
Ström, S., Helmfrid, I., Glynn, A., and Berglund, M. (2011). Nutritional and toxicological
aspects of seafood consumption – An integrated exposure and risk assessment of
methylmercury and polyunsaturated fatty acids. Environ. Res. 111, 274-280.
Stutzmann, G.E. (2007). The pathogenesis of Alzheimers disease is it a lifelong
"calciumopathy"? Neuroscientist 13, 546-559.
Suñol, C., Babot, Z., Fonfría, E., Galofré, M., García, D., Herrera, N., Iraola, S., and Vendrell, I.
(2008). Studies with neuronal cells: from basic studies of mechanisms of neurotoxicity to
the prediction of chemical toxicity. Toxicol. In vitro. 22, 1350-1355.
Tai, H.C., and Lim, C. (2006). Computational studies of the coordination stereochemistry,
bonding, and metal selectivity of mercury. J Phys Chem A, 110, 452-62.
Takayama, C., and Inoue, Y. (2004). Transient expression of GABAA receptor α2 and α3
subunits in differentiating cerebellar neurons. Dev Brain Res 148,169-177.
Takayama, C., and Inoue, Y. (2006). Developmental localization of potassium chloride cotransporter 2 in granule cells of the early postnatal mouse cerebellum with special
reference to the synapse formation. Neuroscience 143, 757-767.
219

Takeuchi T. (1968) Pathology of Minamata disease. In Minamata Disease, Study Group of
Minamata disease, pp. 141–228. Kumamoto University, Shuhan Co., Japan.
Takeuchi, T., Morikawa, N., Matsumoto, H., and Shiraishi, Y. (1962). A pathological study of
Minamata Disease in Japan. Acta Neuropathologica 2, 40-57.
Tanaka, Y., Tanaka, Y., Furuta, T., Yanagawa, Y., and Kaneko, T. (2008). The effects of cutting
solutions on the viability of GABAergic interneurons in cerebral cortical slices of adult
mice. J. Neurosci. Methods 171, 118-125.
Tang, S.C., Arumugam, T.V., Cutler, R.G., Jo, D.G., Magnus, T., Chan, S.L., Mughal, M.R.,
Telljohann, R.S., Nassar, M., Ouyang, X., Calderan, A., Ruzza, P., Guiotto, A., and
Mattson, M.P. (2007). Neuroprotective actions of a histidine analogue in models of
ischemic stroke. J. Neurochem.101, 729-36.
Tashiro, A., Aaron, G., Aronov, D., Cossart, R., Dumitriu, D., Fenstermaker, V., Goldberg, J.,
Hamzei-Sichani, F. Ikegaya, Y., Konur, S., MacLean, J., Nemet, B., Nikolenko, V.,
Portera-Cailliau, C., and Yuste R. (2006). Imaging brain slices. In Handbook of
Biological Confocal Microscopy (J.B. Pawley, ed.), 3rd ed., pp.722-735. Springer
Science+Business Media, New York, NY.
Thévenaz, P., Ruttimann, U.E., and Unser, M. (1998). A pyramid approach to subpixel
registration based on intensity. IEEE Trans on Image Proc 7, 27-41.
Toescu, E.C., and Vreugdenhil, M. (2010). Calcium and normal brain aging. Cell Calcium 47,
158-164.
Tokuomi, H., Uchino, M., Imamura, S., Yamanaga, H., Nakanishi, R., and Ideta, T. (1982)
Minamata Disease (organic mercury poisoning): neurologic and electrophysiologic
studies. Neurology 32, 1369-1375.
Traxinger, D., and Atchison, W.D. (1987). Comparative effects of divalent cations on the
methylmercury-induced alterations of acetylcholine release. JPET 240, 451-459.
Tretter, V., Mukherjee, J., Maric, H., Schindelin, H., Sieghart, W., and Moss, S.J. (2012).
Gephyrin, the enigmatic organizer at GABAergic synapses. Frontiers in Cell. Neurosci.
6.23, 1-16.
Tsang, S.Y., Ng, S.K., Xu, Z., and Xue, H. (2007). The evolution of GABAA receptor-like
genes. Mol. Biol. Evol. 24, 599-610.

220

Tsetlin, V., Kuzmin, D., and Kasheverov, I. (2010) Assembly of nicotinic and other Cys-loop
receptors. J Neurochem. 116, 734-741.
Tymianski, M., Charlton, M.P., Carlen, P.L., and Tator, C.H. (1993). Secondary Ca2+ overload
indicates early neuronal injury which precedes staining with viability markers. Brain Res.
607, 319-323.
Uchino, M., Tanaka, Y., Ando, Y., Yonehara, T., Hara, A., Mishima, I., Okajima, T., and Ando,
M. (1995). Neurologic features of chronic minamata disease (organic mercury poisoning)
and incidence of complications with aging. J. Environ. Sci. Health B 30, 699-715.
US EPA, (1997). Methylmercury study report to congress. EPA-452/R-97-006.
Usuki, F., Yasutake, A., Umehara, F., Tokunaga, H., Matsumoto, M., Eto, K., Ishiura, S.,
Higuchi, I. (2001). In vivo protection of a water-soluble derivative of vitamin E, Trolox,
against methylmercury-intoxication in the rat. Neurosci. Lett. 304, 199-203.
Vale, C., Fonfría, E., Bujons, J., Messeguer, A., Rodríguez-Farré, E., and Suñol, C. (2003). The
organochlorine pesticides γ-hexachlorocyclohexane (lindane), α-endosulfan and dieldrin
differentially interact with GABAA and glycine-gated chloride channels in primary
cultures of cerebellar granule cells. Neuroscience. 117, 397-403.
Varagic, Z., Wimmer, L., Schnürch, M., Mihovilovic, M.D., Huang, S., RallaPalli, S., Cook,
J.M., Mirheydari, P., Ecker, G.F., Sieghart, W., and Ernst, M. (2013). Identification of
novel positive allosteric modulators and null modulators at the GABAA receptor α+βinterface. Brit. J. Pharmacol. 169, 371-383.
Vendrell, I., Carrascal, M., Vilaró, M., Abián, J., Rodríguez-Farré, E., and Suñol C. (2007). Cell
viability and proteomic analysis in cultured neurons exposed to methylmercury. Human
& Exp. Toxicol. 26, 263-272.
Vendrell, I., Carrascal, M., Campos, F., Abián, J., and Suñol C. (2010). Methylmercury disrupts
the balance between phosphorylated and non-phosphorylated cofilin in primary cultures
of mice cerebellar granule cells A proteomic study. Toxicol and Applied Pharmacol. 242,
109-118.
Veng, L.M., Mesches, M.H., and Browning, M.D. (2003). Age-related working memory
impairment is correlated with increases in the L-type calcium channel protein alpha1D
(Cav1.3) in area CA1 of the hippocampus and both are ameliorated by chronic
nimodipine treatment. Brain Res. Mol. Brain Res. 110, 193-202.

221

Vitalone, A., Catalani, A., Cinque, C., Fattori, V., Matteucci, P., Zuena, A.R., and Costa, L.G.
(2010) Long-term effects of developmental exposure to low doses of PCB 126 and
methylmercury. Toxicol Lett. 197, 38-45.
Volpe, J.J. (2009). Cerebellum of the premature infant: rapidly developing, vulnerable, clinically
important. J Child Neurol. 24,1085-104.
Wafford, K.A., Thompson, S.A., Thomas, D., Sikela, J., Wilcox, A.S., and Whiting, P.J. (1996).
Functional characterization of human gamma-aminobutyric acidA receptors containing
the alpha 4 subunit. Mol. Pharmacol. 50, 670-678.
Wall, M.J. (2002). Furosemide reveals heterogeneous GABA(A) receptor expression at adult rat
Golgi cell to granule cell synapses. Neuropharmacology 43, 737-749.
Wall, M.J. (2005). A role for zinc in cerebellar synaptic transmission? Cerebellum 4, 224-229.
Wang, Z., Danscher, G., Kim, Y.K., Dahlstrom, A., Mook Jo, S. (2002). Inhibitory zinc-enriched
terminals in the mouse cerebellum: double-immunohistochemistry for zinc transporter 3
and glutamate decarboxylase. Neurosci. Lett. 321, 37-40.
Wankerl, K., Weise, D., Gentner, R., Rumpf, J.J., and Classen, J. (2010). L-type voltage-gated
Ca2+ channels: a single molecular switch for long-term potentiation/long-term
depression-like plasticity and activity-dependent metaplasticity in humans. J. Neurosci.
30, 6197-6204.
Weiss, B. (1990). Risk assessment: the insidious nature of neurotoxicity and the aging brain.
Neurotoxicology 11, 305-313.
Weiss, B. (2010). Lead, manganese, and methylmercury as risk factors for neurobehavioral
impairment in advanced age. Int. J. Alzheimers Dis. 2011, 607543.
Weiss, B., Clarkson, T.W., and Simon, W. (2002). Silent latency periods in methylmercury
poisoning and in neurodegenerative disease. Environ. Health Perspect. 110 Suppl. 5,
851-854.
Weiss, B., Stern, S., Cox, C., and Balys, M. (2005). Perinatal and lifetime exposure to
methylmercury in the mouse: behavioral effects. Neurotoxicology 26, 675-690.
Westenbroek, R.E., Sakurai, T., Elliott, E.M., Hell, J.W., Starr, T.V.B., Snutch, T.P., and
Catterall W.A. (1995). Immunochemical identification and subcellular distribution of the
α1A subunits of brain calcium channels. J. Neurosci. 15, 6403-6418.

222

Wisden, W., Cope, D., Klausberger, T., Hauer, B., Sinkkonen, S.T., Tretter, V., Lujen, R., Jones,
A., Korpi, E.R., Mody, I., Sieghart, W., and Somogyi, P. (2002). Ectopic expression of
the GABAA receptor α6 subunit in hippocampal pyramidal neurons produces
extrasynaptic receptors and an increased tonic inhibition. NEuropharmacology. 43, 530549.
Wong, L.C., Lu, B., Tan, K.W., and Fivaz, M. (2010) Fully-automated image processing
software to analyze calcium traces in populations of single cells. Cell Calcium. 48, 270274.
Workman, A.D., Charvet, C.J., Clancy, B., Darlington, R.B., Finlay, B.L. (2013). Modeling
transformations of neurodevelopmental sequences across mammalian species. J
Neurosci. 33, 7368-7383.
Wu, X., Wu, Z., Ning, G., Guo, Y., Ali, R., Macdonald, R.L., De Blas, A.L., Luscher, B., and
Chen, G. (2012). γ-aminobutyric acid type A (GABAA) receptor α subunits play a direct
role in synaptic versus extrasynaptic targeting. J. Biol. Chem. 287, 27417-27430.
Xu, J., Mashimo, T., and Sudhof T.C. (2007). Synaptotagmin-1, -2, and -9: Ca(2+) sensors for
fast release that specify distinct presynaptic properties in subsets of neurons. Neuron 64,
667-681.
Xu, M., Yan, C., Tian, Y., Yuan, X., and Shen, X. (2010). Effects of low level of methylmercury
on proliferation of cortical progenitor cells. Brain Res. 1359, 272-290.
Xu, X., Weber, D., Carvan, M.J.3rd, Coppens, R., Lamb, C., Goetz, S., and Schaefer, L.A.
(2012). Comparison of neurobehavioral effects of methylmercury exposure in older and
younger adult zebrafish (Danio rerio). NeuroToxicology, 33, 1212-1218.
Yaginuma-Sakurai, K., Murata, K., Iwai-Shimada, M., Nakai, K., Kurokawa, N., Tatsuta, N.,
Satoh, H. (2012). Hair-to-blood ratio and biological half-life of mercury: experimental
study of methylmercury exposure through fish consumption in humans. J. Toxicol. Sci.
37, 123-130.
Yamada, J., Okabe, A., Toyoda, H., Kilb, W., Luhmann, H.J., and Fukuda, A. (2004). Cl- uptake
promoting depolarizing GABA actions in immature rat neocortical neurones is mediated
by NKCC1. J Physiol 557, 829-841.
Yasutake, A., Nakano, A., Miyamoto, K., and Eto, K. (1997). Chronic effects of methylmercury
in rats. I. Biochemical aspects. Tohoku J Exp. Med. 182, 197-205.

223

Yee, S., Choi, B.H. (1994). Methylmercury poisoning induces oxidative stress in the mouse
brain. Exp. Mol. Pathol. 60, 188-196.
Yee, S., and Choi, B.H. (1996). Oxidative stress in neurotoxic effects of methylmercury
poisoning. Neurotoxicology 17, 17-26.
Yin, Z., Milatovic, D., Aschner, J.L., Syversen, T., Rocha, J.B., Souza, D.O., Sidoryk, M.,
Albrecht, J., and Aschner, M. (2007). Methylmercury induces oxidative injury,
alterations in permeability and glutamine transport in cultured astrocytes. Brain Res.
1131, 1-10.
Yoshida, M., Shimizu, N., Suzuki, M., Watanabe, C., Satoh, M., Mori, K., and Yasutake, A.
(2008). Emergence of delayed methylmercury toxicity after perinatal exposure in
metallothionein-null and wild-type C57BL mice. Environ. Health Perspect. 116, 746751.
Yorifuji, T., Tsuda, T., Inoue, S., Takao, S., and Harada, M. (2011). Long-term exposure to
methylmercury and psychiatric symptoms in residents of Minamata, Japan. Environ. Int.
37, 907-913.
Yuan, Y., and Atchison, W.D. (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 282, 6473.
Yuan, Y., and Atchison, W.D. (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, W.D. (2003). Methylmercury differentially affects GABAA receptormediated spontaneous IPSCs in Purkinje and granule cells of rat cerebellar slices. J
Physiol 550.1, 191-204.
Yuan, Y., and Atchison, W.D. (2007). Methylmercury-induced increase of intracellular Ca2+
increases spontaneous synaptic current frequency in rat cerebellar slices. Mol Pharmacol
71, 1109-1121.
Yuan, Y., Otero-Montañez, J.K., Yao, A., Herden, C.J., Sirois, J.E., and Atchison, W.D. (2005).
nwardly rectifying and voltage-gated outward potassium channels exhibit low sensitivity
to methylmercury. Neurotoxicology 26, 439-454.

224

Yuste, R., MacLean, J., Vogelstein, J. and Paninski, L. (2011). Imaging action potentials with
calcium indicators. Cold Spring Harb. Protoc. 2011, 985-989.
Zariwala, H.A., Borghuis, B.G., Hoogland, T.M., Madisen, L., Tian, L., De Zeeuw, C.I., Zeng,
H., Looger, L.L., Svoboda, K., and Chen, T.-W. (2012). A Cre-dependent CGaMP3
reporter mouse for neuronal imaging in vivo. J. Neurosci. 32, 3131-3141.
Zheng, J.Q., and Poo, M. (2007). Calcium signaling in neuronal motility. Annu. Rev. Cell Dev.
Biol. 23, 23375–23404.
Zheng, T., Santi, M.R., Bovolin, P., Marlier, L.N.J., and Grayson, D.R. (2003). Developmental
expression of the α6 GABAA receptor subunit mRNA occurs only after cerebellar granule
cell migration. Dev Brain Res 75, 91-103.
Zhou, X., Wei, J., Song, M., Francis, K., and Yu, S.P. (2010). Novel role of KCNQ2/3 channels
in regulating neuronal cell viability. Cell Death Differ. 18, 493-505.
Zhuchenko, O., Bailey, J., Bonnen, P., Ashizawa, T., Stockton, D.W., Amos, C., Dobyns, W.B.,
Subramony, S.H., Zoghbi, H.Y., and Lee, C.C. (1997). Autosomal dominant cerebellar
ataxia (SCA6) associated with small polyglutamine expansions in the alpha
1A-voltage-dependent calcium channel. Nat. Genet. 15, 62-69.
Zimmer, B., Kuegler, P.B., Baudis, B., Genewsky, A., Tanavde, V., Koh, W., Tan, B.,
Waldmann, T., Kadereit, S., and Leist, M. (2010). Coordinated waves of gene expression
during neuronal differentiation of embryonic stem cells as basis for novel approaches to
developmental neurotoxicity testing. Cell Death Differ. 18, 383-395.

225