The Pennsylvania State University

The Graduate School

College of Medicine

L-TYPE REGULATION IN THE P/Q-TYPE CALCIUM

CHANNEL MUTANT MOUSE, TOTTERING, A MODEL FOR EPISODIC

NEUROLOGICAL DISORDERS

A Thesis in

Neuroscience

by

Brandy Ellen Fureman

 2001 Brandy Ellen Fureman

Submitted in Partial Fulfillment

of the Requirements

for the Degree of

Doctor of Philosophy

August 2001 We approve the thesis of Brandy Ellen Fureman.

Date of Signature

______Ellen J. Hess Associate Professor of Neuroscience and Anatomy Thesis Advisor Co-Chair of Committee

______Robert J. Milner Professor of Neuroscience and Anatomy Director, Graduate Program in Neuroscience Co-Chair of Committee

______Melvin L. Billingsley Professor of Pharmacology

______Robert Levenson Professor of Pharmacology

______Theresa L. Wood Associate Professor of Neuroscience and Anatomy iii

ABSTRACT

Mutations in calcium channels cause episodic neurological disorders in humans and paroxysmal phenotypes in mice. Tottering mice display stress-induced attacks of a movement disorder due to a P/Q-type calcium channel mutation and cerebellar L-type calcium channel upregulation.L-type calcium channels were assessed in studies of calcium uptake, radioligand binding and in situ hybridization in tottering mice. These studies confirmed the increase in tottering cerebellar L-type calcium channels. Mutant mice also exhibited unique responses to repeated exposure to an L-type calcium channel antagonist () and agonist (BAY

K8644). These studies provide further evidence of L-type calcium channel misregulation in the tottering mouse cerebellum.

In developing tottering mice, restraint stress did not produce motor attacks until twenty- two days of age; attack frequencies at p22 were similar to adults. A second phenotype, aberrant cerebellar tyrosine hydroxylase mRNA expression, was not detected until after the onset of stress-induced motor attacks. These studies more precisely define the temporal relationship between these phenotypes, suggesting that aberrant calcium-responsive gene expression is a consequence of the intense cerebellar activation associated with motor attacks.

In adult tottering mice, attacks were reliably precipitated by stressful environmental disturbances, caffeine and administration; these agents are widely known triggers in human episodic disorders. Thus, tottering mice provide an excellent model to study common pathophysiological mechanisms underlying trigger phenomena in ion channelopathies. As iv activation of hormonal systems is a common feature of tottering mouse triggers, the influence of stress hormones on tottering mouse attacks was assessed. Glucocorticoids were neither necessary nor sufficient for the expression of this behavior; pharmacological blockade of the corticotropin- releasing hormone receptor type 1 (CRF-1) did not prevent stress-induced motor attacks.

Potential therapeutics including , and failed to prevent motor attacks. However, L-type calcium channel antagonist nimodipine prevented caffeine- and alcohol-induced attacks, and the glutamatergic antagonist MK-801 was effective in preventing attacks induced by restraint stress. The results of these studies suggest that abnormal ionic signaling triggered through calcium-dependent mechanisms in cerebellar networks produces periods of hyperexcitability in the tottering mouse brain. The unique neurocircuitry of the cerebellum may amplify these abnormal signals once they are produced. v

TABLE OF CONTENTS

List of Figures...... vii List of Tables ...... ix List of Abbreviations...... x Acknowledgments...... xiii Chapter 1: Introduction and Literature Review...... 1 Chapter Summary...... 1 Voltage Dependent Calcium Channels...... 3 Human Episodic Disorders...... 9 The tottering Mouse...... 13 Conclusions and Experimental Questions ...... 24 Chapter 2: Tottering mouse development ...... 26 Chapter Summary...... 26 Rationale...... 28 Materials and Methods ...... 32 Results ...... 38 Discussion...... 44 Chapter 3: L-type calcium channels in tottering mouse phenotypes ...... 50 Chapter Summary...... 50 Rationale...... 52 Materials and Methods ...... 54 Results ...... 58 Discussion...... 67 Chapter 4: Chronic L-type calcium channel activation in tottering mice ...... 73 Chapter Summary...... 73 Rationale...... 75 Materials and Methods ...... 77 Results ...... 80 Discussion...... 84 Chapter 5: Triggers and potential therapeutics in tottering mouse attacks ...... 89 Chapter Summary...... 89 Rationale...... 91 Materials and Methods ...... 94 Results ...... 98 Discussion...... 105 vi

Chapter 6: Conclusions ...... 111 Development of the tottering mouse phenotypes...... 113 Perturbed L-type calcium channel regulation in tottering mice ...... 115 Cerebellar TH mRNA expression in tottering mice...... 118 Attack triggers and potential therapeutics ...... 120 General model of hyperexcitability in the tottering mouse cerebellum...... 122 References ...... 126 vii

LIST OF FIGURES

Figure Title Page α 1.1 Predicted protein topology of voltage dependent calcium channel 1 subunits 4

1.2 Model diagram of predicted voltage dependent calcium channel subunit 5 interactions

2.1 Developmental onset of restraint-induced attacks in tottering mice 38

2.2 TH mRNA expression in the developing mouse brain 40

2.3 Calcium uptake in cerebellar synaptosomes from developing mice 41

2.4 Calcium uptake in forebrain synaptosomes from developing mice 42

3.1 Calcium uptake in cerebellar synaptosomes 58

3.2 Calcium uptake in hippocampal synaptosomes 59

3.3 Calcium uptake in cortical synaptosomes 60

3.4 Calcium uptake in striatal synaptosomes 61

3.5 Density of [3H]PN-200-110 binding sites after chronic nimodipine treatment 62 in control and tottering mice

3.6 Abnormal cerebellar TH mRNA expression in the adult tottering mouse 66

3.7 Cerebellar TH mRNA expression in tottering mice after chronic nimodipine 66 blockade

3.8 Effect of chronic nimodipine treatment on TH mRNA expression in tottering 65 mouse Purkinje cells

4.1 The effect of repeated 8 mg/kg BAY K8644 challenge on motor impairment 80 in control and tottering mice viii

Figure Title Page 4.2 Calcium uptake in cerebellar synaptosomes from control and tottering mice 81 repeatedly exposed to BAY K8644

4.3 Calcium uptake in striatal synaptosomes from control and tottering mice 82 repeatedly exposed to BAY K8644

5.1 Environmental disturbances trigger attacks in tottering mice 98

5.2 Effect of anticonvulsant administration on the frequency of attacks 100

5.3 Effect of MK-801 administration on the frequency of tottering mouse attacks 102 induced by restraint

6.1 Model diagram describing neuroantomical connections that may contribute 124 to hyperexcitability in the tottering mouse cerebellar cortex ix

LIST OF TABLES

Table Title Page 1.1 Molecular and pharmacological classification of high voltage-activated 5 calcium channel subtypes in the central nervous system

1.2 Selected human disorders with episodic expression 10

1.3 Voltage dependent calcium channel mutations in mice 13

1.4 Behavioral and cellular phenotypes in calcium channel mutant mice 14

α 2.1 1A mRNA expression in developing tottering and control mice 42

α 2.2 1C mRNA expression in developing tottering and control mice 43

α 3.1 1C mRNA expression following chronic nimodipine treatment in tottering 63 and control mice

4.1 Rating system for motor impairment 78

α 4.2 Cerebellar L-type calcium channel 1C subunit mRNA expression following 83 chronic BAY K8644 administration in tottering and control mice

5.1 Effect of caffeine on tottering attacks 99

5.2 Effect of on tottering attacks 100

5.3 Effect of nimodipine pre-treatment on caffeine-induced tottering attacks 102

5.4 Effect of nimodipine pre-treatment on ethanol-induced tottering attacks 102

5.5 Efficacy of triggers following surgical adrenalectomy in tottering mice 104

5.6 Effect of exogenous glucocorticoid on the frequency of tottering attacks 104

5.7 Effect of CRF-1 receptor blockade on the frequency of tottering attacks 105 x

LIST OF ABBREVIATIONS

A1 Adenosine receptor type 1

A2A Adenosine receptor type 2A ACTH Adrenocorticotropin hormone ADX Adrenalectomized AMPA α-amino-3-hydroxy-5-methylisoxazole-4-propionic acid ATP Adenosine triphosphate BCA Bicinchoninic acid bp Basepair C Cytosine CB Cerebellum CORT Corticosterone cDNA Complementary deoxyribonucleic acid Ci Curie CNS Central nervous system CRF Corticotropin-releasing factor CRF-1 CRF receptor type 1 CRF-2 CRF receptor type 2 cRNA Complementary ribonucleic acid CTP Cytosine triphosphate DCN Deep cerebellar nuclei DHP Dihydropyridine DNA Deoxyribonucleic acid DTT Dithiothreitol EA2 Episodic ataxia type 2 EDTA Ethylenediaminetetraacetic acid EGFP Enhanced green fluorescent protein EEG Electroencephalogram EPSP Excitatory postsynaptic potential EtOH Ethanol ETn Early transposon insertion FHM Familial hemiplegic migraine g Gram(s) g Standard acceleration of gravity (9807mm/sec2) GABA Gamma-aminobutyric acid

GABAA GABA receptor type A xi

GC Glucocorticoid GTP Guanocine triphosphate HCL Hydrochloric acid HEPES N-[2-Hydroxyethyl]piperazine-N’-[2-ethanesulfonic acid] HPA Hypothalamic-pituitary-adrenal hr Hour(s) kb Kilobase pairs kg Kilogram IEG Immediate early gene IO Inferior olive i.p. Intraperitoneal IPSP Inhibitory postsynaptic potential LC Locus coeruleus M Molar MCID Microcomputer imaging device mg Milligram µg Microgram min Minute(s) ml Milliliter µl Microliter mM Millimolar µM Micromolar mMol Millimole(s) µMol Micromole(s) Mol Mole(s) mRNA Messenger ribonucleic acid n Number NE Norepinephrine ng Nanogram(s) NIH National Institutes of Health NMDA N-methyl-d-aspartatic acid nMol Nanomole(s) OB Olfactory bulb Os Oligosyndactylism p Probability PBS Phosphate buffered saline PC Purkinje cell xii pcd Purkinje cell degeneration PCR Polymerase chain reaction PTZ Pentalenetetrazole pX Postnatal day (X) RN Red nucleus ROD Raw optical density RNA Ribonucleic acid RPM Rotations per minute S Segment SC Spinal cord s.c. Subcutaneous sec Second(s) SEM Standard error of the mean SN Substantia nigra SSC Standard saline citrate T Thymidine tg Tottering TH Tyrosine hydroxylase tRNA Transfer ribonucleic acid U Unit(s) UTP Uridine-triphosphate VDCC Voltage dependent calcium channel VL Ventrolateral VPL Ventroposteriolateral v:v Volume:volume w:v Weight:volume 0C Degrees Celcius [3H] Tritium-labeled + Control % Per cent xiii

ACKNOWLEDGMENTS

Over the past five years, I have been extremely fortunate to learn from a great number of individuals: challenging teachers, patient mentors, expert assistants and exceptional friends. I have benefited professionally and personally from their interest in my development, and I offer my deep gratitude for their efforts.

Dr. Ellen J. Hess, my thesis advisor, has unstintingly shared her experience, resources and vision while supporting me in my exploration of the scientific endeavor. I thank her for her willingness to foster creativity and encourage independence during my pre-doctoral training in her laboratory.

The members of my thesis committee, Drs. Billingsley, Levenson, Milner and Wood, have consistently provided insightful comments and advice. Each has generously provided me with opportunities for grant support, technical training, or teaching experience in addition to the use of equipment and reagents.

I would like to thank both Drs. Joan Lakoski and Patricia Sue Grigson for serving as personal mentors; their support during stressful periods was much appreciated. I am grateful to the Graduate Program in Neuroscience, especially Director Dr. Robert Milner and assistant

Linda Flickenger for their special efforts to provide assistance at every step of the process. My fellow classmates, Corey Hart, Michelle Jones and Robert Wheeler have been loyal comrades-in- arms; their support and friendship has enriched my graduate school experience in countless ways.

Technical assistance came from many sources; I appreciate the help provided by Crystal

Anglen, Dr. Robert Bonneau, Dr. Daniel Campbell, Dr. Victor Canfield, Roger Carroll, Corey

Hart, Dr. Hyder Jinnah, Dr. J. Kyle Krady, Dr. Kang Li, Ridwan Lin, Dr. Lynn Maines, Dr.

Steven O’Donnell, Jennifer Thompson, Robert Twining and Robert Wheeler. xiv

Within the Hess laboratory, Marianne Klinger, Jesse North, Matthew Deardorff and

Bryan Lee have been wonderful “teammates” and friends. I am indebted to my treasured companions, Kristy Bruno, Carolyn Pizoli and Michelle Jones for providing inspiration, strength and laughter during our shared journey. I am also grateful for the endless encouragement from my friends Sarah Gabriele, Marci Kauffman and Jennifer Wells-Marani. Thank you, all.

Finally, I thank my family, Rogie, John, Jared and Sasha. They know why.

Research supported by PHS NS33592, NS34845, NS40470 and 5 T32 ES07312. 1

CHAPTER 1: INTRODUCTION AND LITERATURE REVIEW

Chapter Summary

Genetic mutations that cause human neurological disorders are being identified at a rapid rate. Unfortunately, the identification of a putative disease-causing mutation often does not reveal the underlying pathophysiological mechanism in many neurological disorders. Conditions arising from mutations in genes, or “channelopathies”, exemplify the difficulty in predicting a priori how a mutation in a single protein ultimately leads to deficits in the neural systems regulating behavior.

Ion channels control the movement of specific ions (such as sodium, potassium, chloride, and calcium) across the cell membrane in excitable tissues; these processes are fundamental to membrane excitation, activity-dependent gene transcription, intracellular signaling and neurotransmitter release. Clearly, ion channels play an essential role in the complex coordination of the nervous system. Surprisingly however, many patients with an ion channel mutation display only occasional neurological deficits, precipitated by common triggers such as stress, caffeine, alcohol, hormonal changes, and fatigue. Between attacks, patients are often asymptomatic despite substantial abnormalities detected in the activity of the mutated channel.

Thus, the mechanism by which a persistent abnormality in an ion channel leads to episodic neurological dysfunction remains unclear. Still less is understood about how trigger factors, often 2 agents with apparently divergent effects on the nervous system (such as alcohol and caffeine), can act on perturbed ion channel systems to generate attacks in a channelopathy disorder.

An animal model of ion channelopathy is an excellent experimental system in which to ask how an ion channel mutation might affect the development and regulation of other ion channels, and how particular trigger factors interact with calcium channels to elicit attacks.

Ideally, an animal model of ion channelopathy should have both genetic and phenotypic similarity to human channelopathies, specifically with regard to the transient nature of the neurological impairments. The tottering (tg/tg) mouse displays attacks of a movement disorder characterized by involuntary twisting movements and sustained abnormal postures due to an ion channel mutation. Mutations in the same ion channel gene in humans cause migraine and paroxysmal movement disorders. The tottering mouse appears to be an excellent candidate system to study the consequences of an ion channel mutation on the mammalian central nervous system as well as the factors that precipitate the transient neurological impairments characteristic of channelopathy disorders. 3

Voltage Dependent Calcium Channels

Strict control over intracellular calcium concentration in excitable cells generates a system by which changes in the external environment are rapidly communicated to the cell via membrane depolarization and subsequent changes in calcium influx (reviewed by Janis et al.,

1987). The intracellular calcium concentration influences membrane excitability, neurotransmitter release, hormonal secretion, activation of intracellular second messengers, and gene transcription (reviewed by Jones, 1998; Catterall, 2000). Six different voltage-dependent calcium channel subtypes mediate these diverse processes. These subtypes (L-, N-, P-, Q-, R-, and T-type) have distinct expression patterns, physiologic responses and developmental profiles, and appear to have unique functions in the central nervous system.

Although there is a wealth of data demonstrating ion channel function in vitro, a great deal of uncertainty regarding the basic biology of ion channels in the intact mammalian nervous system remains. This is particularly true of voltage-gated calcium channels, largely due to a lack of specific drugs and assays suited to in vivo study. As a result of these limitations, it is difficult to determine the in vivo effects of a mutation on calcium channel function, or how those abnormalities might affect the regulation of other calcium channel subtypes. Since several related disorders in humans are linked to mutations in the genes encoding voltage-dependent calcium channel subunits (Ophoff et al., 1996; Jodice et al., 1997), a better understanding of the calcium channel system in vivo may help in understanding and treating the symptoms of calcium channelopathies. 4

Structure of voltage dependent calcium channels

Neuronal voltage-dependent calcium channels are comprised of at least four, and in some

α β α δ γ cases five subunits, including the pore-forming 1 subunit and modulatory , 2, , and

α subunits. The 1 subunit

consists of four internal-repeat

domains (I-IV) composed of six

transmembrane segment repeats

(S1-S6) (Fig. 1.1) and is

structurally very similar to

sodium and

Fig 1.1 Predicted protein topology of voltage-dependent calcium pore-forming subunits α channel 1 subunits. + indicates the position of the voltage sensor, * indicates the approximate position of the tottering α (reviewed by Catterall, 2000) . mutation in the murine 1A subunit (Adapted from Hess, 1996).

The a1 subunit is responsible for

channel pore-formation, ion-selectivity, voltage-sensitivity and ligand binding, while the other

subunits modulate channel function.

α Although 1 subunits are sufficient to form the channel pore, they typically do not

conduct appreciable amounts of calcium in the absence of modulatory subunits. Modulatory

subunits increase the peak current amplitude of the channel, accelerate the rates of channel

α activation and inactivation, and enhance ligand binding to the 1 subunit (reviewed by Walker 5

and De Waard, 1998). A fully functional

voltage dependent calcium channel complex

in the mammalian central nervous system is

α believed to (minimally) consist of single 1,

β α δ , and 2 subunits (Figure 1.2).

Figure 1.2 Model diagram of predicted voltage- dependent calcium channel subunit interactions. α Classification of voltage dependent The 1 subunit forms the channel pore and controls ion selectivity and voltage sensitivity, calcium channel subtypes while β and α2δ subunits modulate these properties. Drawing by K. Bruno. Functional diversity among the subtypes of voltage dependent calcium channels is due to the existence of multiple genes for

α individual subunits and alternative splicing of subunit genes. To date, a total of nine 1 subunit

Table 1.1 Molecular and pharmacological classification of high voltage-activated calcium channel subtypes in the central nervous system. Subunit Gene Channel Subtype Pharmacologic Sensitivity α ω ω 1A-a P -agatoxin IVA, -agatoxin TK α ω 1A-b Q - MVIIC α ω ω 1B N -conotoxin MVIIA, -conotoxin GIVA α 1C L dihydropyridines, benzothiazepines α 1D and phenylalkylamines α 2+ 2+ 1E RNi, Cd

genes (A-I) have been cloned. Of these, five genes expressed in the central nervous system encode voltage-dependent calcium channels activated by large voltage changes; the others 6

α α encode calcium channels that are low voltage-activated ( 1G-I) or expressed peripherally ( 1S).

The high voltage-activated calcium channel subtypes expressed in the central nervous system can

α be categorized into five different classes based on pharmacological sensitivity of the 1 subunit

(Table 1.1).

Table 1.1 Molecular and pharmacological classification of high voltage-activated calcium channel subtypes in the central nervous system. Subunit Gene Channel Subtype Pharmacologic Sensitivity α ω ω 1A-a P -agatoxin IVA, -agatoxin TK α ω 1A-b Q -conotoxin MVIIC α ω ω 1B N -conotoxin MVIIA, -conotoxin GIVA α 1C L dihydropyridines, benzothiazepines α 1D and phenylalkylamines α 2+ 2+ 1E RNi, Cd

In addition, multiple genes have been identified for modulatory subunit types as well. In

vitro, each member of a modulatory subunit group can interact with and modify the conductance

α properties of each 1 subunit, suggesting an enormous potential for physiological diversity

among the calcium channel subtypes based on subunit interactions. However, although mRNA

and protein expression for the various subunits has been demonstrated in vivo (Hillman et al.,

1991; Chin et al., 1992; Castellano et al., 1993; Hell et al., 1993; Stea et al., 1994; Williams et

al., 1994; Ludwig et al., 1997; Klugbauer et al., 2000), much of the information regarding effects

of subunit interactions on channel properties has, of necessity, been studied using heterologous

expression systems in vitro. As a result, it is not known whether all of these interactions are

functionally important in the mammalian central nervous system. 7

Voltage dependent calcium channel subtype localization and predicted function

The discrete regional and subcellular localization of voltage-dependent calcium channel subtypes suggests that different subtypes mediate diverse functions within neuronal cells. The non-dihydropyridine sensitive subtypes P/Q-, N-, and R-type calcium channels are functionally coupled to the synaptic vesicle release machinery (reviewed by Sheng et al., 1998) and provide the calcium trigger for synchronous neurotransmitter release (Luebke et al., 1993; Takahashi and

α Momiyama, 1993; Turner et al., 1993) Within the central nervous system, 1A subunit mRNA encoding P/Q-type calcium channels is most abundantly expressed in the cerebellum,

α hippocampus, dentate gyrus and inferior colliculus, while N-type 1B mRNA distribution is

α diffuse with moderate levels of expression in many brain regions. The 1E subunit, predicted to encode R-type calcium channels, is abundantly expressed throughout the brain, particularly in the olfactory bulbs, hippocampus, dentate gyrus, interpeduncular nucleus, and cerebellum

(Ludwig et al., 1997).

L-type calcium channels are most densely expressed at the soma and proximal dendrites,

α α in a clustered ( 1C) or continuous ( 1D) pattern (Hell et al., 1993). This subcellular position may better enable L-type calcium channels to provide a link between neuronal activity and both short and long-term changes in neuronal gene expression, as calcium influx through L-type calcium channels has been shown to influence transcription in several brain areas (Morgan and Curran,

1986; Vidal et al., 1989; Brosenitsch et al., 1998; Cigola et al., 1998). L-type calcium channel 8 subunit mRNA is expressed in the olfactory bulbs, hippocampus, dentate gyrus, superior colliculus and cerebellum, as well as in neuroendocrine regions such as hypothalamic nuclei and the pituitary and pineal glands (Chin et al., 1992; Ludwig et al., 1997). While L-type calcium channels are not required for synchronous neurotransmitter release, they have been implicated in neuroendocrine secretion (Laird et al., 1991; Borycz et al., 1993), influence membrane excitability and may have a unique role during development.

Voltage dependent calcium channel regulation during development

α The expression patterns of all calcium channel 1 subunits tend to be low during early postnatal development of the rodent nervous system (Tanaka et al., 1995). Studies using autoradiography and radioligand binding of tritiated L-type calcium channel ligands, calcium uptake and calcium channel subunit mRNA expression in Xenopus oocytes have all documented a peak in L-type calcium channel expression, ranging from postnatal days 9-21 (Erdman et al.,

1983; Kazazoglou et al., 1983; Mourre et al., 1987; Litzinger et al., 1993; Tanaka et al., 1995).

Following this sharp increase, L-type calcium channel expression and functional calcium uptake begins to decline to reach moderate levels in the adult. This developmental profile mirrors the timing of synaptic maturation in the central nervous system, as sodium (Sashihara et al., 1995;

Felts et al., 1997), potassium channels (Downen et al., 1999) and glutamate receptors (Takayama et al., 1996) show similar patterns. 9

Human Episodic Disorders

Episodic disorders result from ion channel mutations

Episodic disorders have a significant impact on the health of otherwise normal individuals. Although symptoms of individual disorders are quite diverse, including migraine headache, epilepsy, cardiac arrythmia, paralysis or hemi-paralysis, ataxia, and dyskinesia, there are astonishing similarities between both the genetic etiology and the factors capable of triggering symptomatic attacks in many episodic disorders. There is a growing consensus that most episodic disorders are caused by mutations in ion channels (Ptacek, 1999); several episodic neurological disorders caused by ion channelopathy are listed in Table 1.2. The distinguishing feature of these channelopathies is the transient nature of the neurological dysfunction, characterized by paroxysmal episodes of abnormal motor control and/or muscle tone.

While many of the monogenic channelopathies listed in Table 1.2 are rare, it has been suggested that understanding the pathogenesis of attacks in these genetically simple disorders may provide the basis for unraveling genetically complex episodic disorders such as migraine and epilepsy (Ptacek, 1998; Terwindt et al., 1998; Jen, 1999). 10

Table 1.2 Selected human disorders with episodic expression Affected Ion Channel Disorder Calcium channel Familial hemiplegic migraine (Ophoff et al., 1996) Episodic ataxia (type 2) (Ophoff et al., 1996; Jodice et al., 1997; Jen et al., 1999) Hypokalemic periodic paralysis (Ptacek et al., 1994b)

Sodium channel Hyperkalemic periodic paralysis (Ptacek et al., 1991) Paramyotonia congenita (Ptacek et al., 1992) Potassium-aggravated myotonia (Ptacek et al., 1994a)

Potassium channel Episodic ataxia (type 1) (Browne et al., 1994) Benign familial neonatal convulsions Ð 1 (Singh et al., 1999a) Benign familial neonatal convulsions Ð 2 (Charlier et al., 1998)

Chloride channel Myotonia congenita (Zhang et al., 1996)

Acetylcholine receptor Myasthenic syndrome (Ohno et al., 1995) Autosomal dominant nocturnal frontal lobe epilepsy (Steinlein et al., 1995)

Although the channelopathies have historically been regarded as independent clinical entities, more recently it has been realized that co-occurrence of various features, such as migraine, epilepsy and movement disorders, can exist in subsets of patients diagnosed with a single channelopathy (Neville et al., 1998; Singh et al., 1999b). Irregular ionic signaling resulting from a channel mutation may have multiple effects on different brain regions and even in different types of excitable tissues, thus generating two or more distinct syndromes in a single channelopathy patient.

Episodic disorders share common triggers

Though the features of different channelopathies are quite diverse, similar stimuli are reported as triggers in human episodic disorders. Psychological or emotional stress, fatigue, 11 exercise, caffeine, alcohol, and hormonal fluctuations are among the most commonly noted triggers (Mount and Reback, 1940; Richards and Barnett, 1968; Lance, 1977; Boel and Casaer,

1988; Bressman et al., 1988; Ptacek, 1998; Battistini et al., 1999; Cooper and Jan, 1999; Jen,

1999; Ptacek, 1999; Jarman et al., 2000; Pittock et al., 2000). The existence of shared triggers suggests that they activate a common pathway that ultimately leads to expression of neurological dysfunction. Understanding the mechanisms by which various triggers contribute to the onset of attacks in otherwise normal individuals is a key step in understanding the pathogenesis of attacks and in developing effective prevention strategies.

Treatment strategies for episodic disorders

Most episodic disorders are treated with anticonvulsants, with variable success. In the particular case of paroxysmal dyskinesia, the efficacy of drug therapy varies widely among the three major subclasses (Demirkiran and Jankovic, 1995; Bhatia, 1999). Patients with paroxysmal kinesiogenic dyskinesia exhibit short-lived attacks induced by sudden movements and respond well to anticonvulsant therapies, including carbamazepine and phenytoin. Patients with attacks of longer duration include those with paroxysmal exercise-induced dyskinesia and paroxysmal non- kinesiogenic dyskinesia, disorders which typically are not controlled by anticonvulsants.

Currently, treatment strategies for paroxysmal non-kinesiogenic dyskinesia consist of diet and lifestyle modification, and although an occasional patient will respond to carbamazepine or the benzodiazepine clonazepam, the response to drug therapy can be extremely variable between individuals with the same diagnosis. A lack of reliable animal models with which to study 12 episodic movement disorders such as paroxysmal dyskinesia contributes to the paucity of effective treatment options for these patients. 13

The tottering Mouse

An animal model of episodic neuronal dysfunction could provide a powerful tool to examine the basic pathophysiology in paroxysmal neurological disorders, and may prove helpful in testing candidate drug therapies. Mutations in genes encoding calcium channel subunits have been identified in several neurological mouse mutants, including tottering, leaner, rolling mouse

α Nagoya, rocker, 1A null mutant, lethargic and stargazer mice (Table 1.3).

Table 1.3 Voltage dependent calcium channel mutations in mice Mouse mutant Affected Mutation Predicted physiological effect(s) References gene/protein Tottering Cacna1a C→T n1802 ↓ calcium current density (Fletcher et al., 1996; α 1A P601L Doyle et al., 1997; Wakamori et al., 1998) Leaner Cacna1a G→A n5762 ↓ calcium current density (Fletcher et al., 1996; α 1A (splice site error) Dove et al., 1998; 1922, 1968 Lorenzon et al., 1998; Wakamori et al., 1998)] Rolling Mouse Cacna1a C→G n3784 ↓ calcium current density, (Mori et al., 2000) α Nagoya 1A R1262G ↓ voltage sensitivity

Rocker Cacna1a C→A n3929 ? (Zwingman et al., α 1A T1310K 2001)

α 1A Subunit Cacna1a Neo cassette gene Elimination of P/Q-type calcium (Jun et al., 1999) α ↑ Null Mutant 1A disruption currents, L- and N-type current in Purkinje cells Lethargic Cacnb4 4 bp insertion ? (Burgess et al., 1997) β 4 (splice site error)

Stargazer Cacnγ2 ETn insertion ? (Letts et al., 1998) γ 2 mutation ETn, early transposon insertion; TH, tyrosine hydroxylase

These mouse mutants offer the chance to study phenotypic heterogeneity arising from different mutations within voltage dependent calcium channels, a significant clinical issue emerging as more mutations are identified in human syndromes. The various mutant mice also 14 display some intriguing similarities (Table 1.4) that may reflect common responses to calcium channel mutations.

Table 1.4 Behavioral and cellular phenotypes in voltage dependent calcium channel mutant mice Mouse Behavioral Phenotypes Cellular Phenotypes mutant ~ age of onset Ataxia Absence seizures Dyskinesia ↑ NE Adult cerebellar TH Cerebellar innervation expression degeneration Tottering + + ++ +++ - ~p21-28 ~p16-18 (episodic, (See Chapter 1) see Chapter 1) Leaner* +++ + +++ + +++ +++ ~p9-10 (chronic) (anterior granule and Purkinje cells) Rolling ++ +++++++ Mouse ~p9-10 (episodic) (anterior granule Nagoya cells only) Rocker + +-?- - ~p21-28 (but see Chapter 6) α 1A Subunit +++ ? +++ ? (present at p21, ? null mutant* ~p10 (show behavioral (chronic, adults not examined) arrest) >p10) Lethargic ++ +++?++? ~p15 (episodic, >p15) Stargazer + +++ -- - ? ~p16-18

Symbols indicate phenotype is: mild (+), moderate (++), severe (+++), absent (-), or not examined (?). *mice display increased mortality beginning at ~3 weeks of age References: tottering (Green and Sidman, 1962; Kaplan et al., 1979; Noebels and Sidman, 1979; Levitt and Noebels, 1981; Hess and Wilson, 1991; Austin et al., 1992; Fletcher et al., 1996), leaner (Herrup and Wilczynski, 1982; Hess and Wilson, 1991; Austin et al., 1992), rolling mouse Nagoya (Muramoto et al., 1981; Muramoto et al., 1982; Tamaki et al., 1986; Sawada et al., 1999), rocker α (Zwingman et al., 2001), 1A null mutant (Jun et al., 1999), lethargic (Burgess et al., 1997; Yoshor et al., 1997), stargazer (Noebels, 1995; (Yoshor et al., 1997))

α The tottering mouse (tg/tg) is the most widely studied of the 1A subunit mouse mutants.

A dramatic behavioral phenotype in these mice is episodic, generalized dyskinesia similar to paroxysmal movement disorders in humans (Green and Sidman, 1962; Campbell and Hess, 15

1999; Campbell et al., 1999). EEG analyses of unanesthetized, freely-moving tottering mice appear normal during the motor attacks (Kaplan et al., 1979; Noebels and Sidman, 1979), suggesting that this phenotype is better characterized as a movement disorder than as a motor seizure. Attacks of dyskinesia impair motor function in tottering mice approximately once or twice per day, but they have long been regarded as spontaneous events (Green and Sidman,

1962); few triggers have been described. As the tottering mouse may provide an excellent pre- clinical model of episodic neurological dysfunction, further investigation into the tottering mouse phenotypes is warranted.

The tottering mutation affects P/Q-type calcium channels

Tottering mice were first identified in 1962 after three DBA/2J littermates with an

“abnormal, wobbly gait” and attacks of a motor disorder were observed at the Jackson

Laboratory (Green and Sidman, 1962). The tottering mutation was subsequently identified as a

α single base pair mutation in the 1Asubunit gene (Fletcher et al., 1996; Doyle et al., 1997). The C to T mutation causes substitution of a leucine for a proline residue in the S5-6 region of domain

II, a region close to the pore-forming domain of the P/Q-type calcium channel. P/Q-type calcium channels are abundantly expressed in both the cerebellum and hippocampus (Hillman et al.,

1991; Starr et al., 1991); neurotransmitter release is thought to be highly dependent on P/Q-type calcium channels in these regions (Gaur et al., 1994; Momiyama and Takahashi, 1994;

Doroshenko et al., 1997). Specifically, a 40% reduction in the whole-cell calcium current has been described in acutely dissociated cerebellar Purkinje cells from developing tottering mice 16

(Wakamori et al., 1998). At hippocampal Schaffer collateral synapses, P/Q-type calcium channel transients are profoundly reduced (Qian and Noebels, 2000). In the cortex and thalamus, regions

α of moderate 1A expression, glutamatergic transmission is diminished (Caddick et al., 1999;

Ayata et al., 2000). Because the mutation does not appear to alter the mRNA or protein

α expression of 1A (Fletcher et al., 1996; Campbell and Hess, 1999) it is thought that the reduction in calcium current is due to a change in the gating properties of the mutant P/Q-type calcium

α channel. However, the connections between the 1A mutation, P/Q-type calcium channel defects and the abnormal cellular and behavioral phenotypes in tottering mice are not yet well defined.

Cerebellar L-type calcium channel upregulation in tottering mice

Following the identification of the tottering mutation in the α1A gene of P/Q-type calcium channels (Fletcher et al., 1996; Doyle et al., 1997), investigations using radioligand binding revealed an apparent upregulation of binding sites for the L-type calcium channel-specific ligand, , in the tottering mouse cerebellum (Campbell and Hess, 1999). Further,

α α increased expression of an L-type calcium channel 1 subunit, 1C, was observed in tottering

α mouse cerebellar Purkinje cells with no change in 1D subunit mRNA. A compensatory increase in another calcium channel subtype may help to explain the relatively mild effects of the tottering mutation on the behavior and viability of the mouse. Both L-type calcium channels and the cerebellum have been implicated in the attacks of dyskinesia exhibited by tottering mice, discussed below. 17

Paroxysmal dyskinesia in tottering mice

Green and Sidman were the first to carefully describe the episodic motor abnormalities in developing tottering mice (Green and Sidman, 1962). In addition to ataxia, they observed

“flattening of the trunk in the sacral area, spastic abduction and extension of one hind limb for a few seconds, recurring in the same pattern every few minutes” in developing tg/tg C57BL6 mice at approximately three weeks after birth. As the mice age, the authors note that the attacks begin to involve more of the body but that the inter-attack interval grows longer. In adult tottering mice, the typical attack begins with contraction of the hindlimbs, causing a waddling gait but little change in mobility. As the episode progresses, the mouse loses mobility and begins to extend and contract the hindlimbs. Gradually the upper limbs, head and neck are severely flexed, in addition to paddling movements of both hindlimbs. As the lower body recovers, the mouse assumes a tripod position on the hindlimbs and tail while the upper body (forelimbs, neck and jaw) is involved in extension/contraction movements, with frequent falls. The entire episode usually lasts from 20-40 minutes, with no apparent refractory period (Campbell and Hess, 1998) and no loss of consciousness (Noebels and Sidman, 1979).

Attack triggers

Largely regarded as spontaneous events, few reports have carefully described the stimuli that trigger attacks in tottering mice. Kaplan and colleagues noted that attacks are often elicited by moving the mice into an unfamiliar environment (Kaplan et al., 1979). Syapin observed that tottering mice repeatedly watched over 60 minutes by a human observer had “spontaneous” 18 attacks 50% of the time, while the occurrence of attacks in mice observed using a video camera for 60 minutes was less than 5%. The author attributes these data to anxiety, stress or fear triggered by the presence of the observer (Syapin, 1983a). Consistent with this interpretation,

Campbell and Hess demonstrated that short-term restraint, a procedure known to cause stress- induced activation of the hypothalamic-pituitary-adrenal (HPA) axis (Barlow et al., 1975;

Armario and Castellanos, 1984; Hauger et al., 1988), reliably induces attacks in tottering mice

(Campbell and Hess, 1998, 1999; Campbell et al., 1999). Taken together, these results suggest that changes in the stress and/or arousal systems may play a role in precipitating tottering mouse attacks.

Convulsant drugs also trigger attacks in tottering mice. Administration of pentylenetetrazol (PTZ), a GABAergic antagonist that causes generalized clonic convulsions in mice, elicits attacks in tottering mice at doses below the threshold for clonic convulsions in wild type and heterozygous mice (Noebels, 1979; Syapin, 1983a). A low dose of the calcium channel agonist, BAY K8644 induces a characteristic tottering mouse attack without features of seizures exhibited by wild-type mice at higher doses of agonist (squeaking, jumping and catatonia)

(Campbell and Hess, 1999). The ability to induce attacks with a calcium channel agonist provides an important link between the calcium channel mutation in tottering mice and the abnormal behavioral phenotype.

Although collectively these studies span twenty years of investigation into the tottering mouse phenotypes, thus far only three factors have been implicated in the initiation of tottering 19 mouse attacks; stress, antagonism of GABA-mediated neuronal inhibition and calcium channel activation. A better understanding of the factors that trigger tottering mouse attacks will likely be applicable to episodic neurological disorders caused by ion channelopathy in humans.

Inhibition of attacks

Using the PTZ-induced attacks as an experimental design, Syapin tested several drugs, including ethosuximide, diazepam, phenytoin, sodium phenobarbital, naloxone, valproic acid, sodium , and aminooxyacetic acid for their ability to prevent PTZ-induced attacks in tottering mice. Of these, the anticonvulsant phenytoin caused “spontaneous” attacks. Only diazepam was effective in blocking attacks induced by PTZ (Syapin, 1983a). As PTZ is a

noncompetitive antagonist at the GABAA receptor while diazepam is a benzodiazepine receptor

agonist, it is possible that diazepam pre-treatment merely reduced PTZ effects at GABAA receptors to below the threshold required for attack induction. Anxiolytic or sedative effects of diazepam may also modify the expression of the attacks. The author notes unpublished observations that diazepam blocks “spontaneous” attacks, which suggests that neuronal

inhibition via GABAA receptors can prevent attacks, but this result has not been formally documented.

To date, the only drugs shown to block tottering mouse attacks other than diazepam are inhibitors of the L-type calcium channel. Peripheral administration of dihydropyridines nimodipine, , or nitrendipine, or the benzothiazepine , inhibits the expression of restraint stress-induced attacks in tottering mice in a dose-dependent manner (Campbell and 20

Hess, 1999). The dihydropyridines readily cross the blood brain barrier to exert central effects

(Janicki et al., 1988; Larkin et al., 1992). The phenylalkylamine, , penetrates the blood brain barrier poorly, and when given peripherally has no effect on tottering mouse attacks.

However, intracerebroventricular administration of verapamil blocks attacks, demonstrating that peripheral effects of L-type calcium channel blockade do not influence attack behavior

(Campbell and Hess, 1999). These results and those described above for the L-type calcium channel activator BAY K8644 suggest that calcium influx is required for the initiation of attacks induced by restraint in tottering mice. However, the tottering mutation in P/Q-type calcium channels reduces calcium influx in dissociated Purkinje cells (Wakamori et al., 1998), initially making this pharmacological data difficult to reconcile with the predicted effects of the mutation.

Evidence for cerebellar involvement in attacks

Overall, the gross morphology of the tottering brain is normal (Green and Sidman, 1962), and initially few anatomical changes were found that could reasonably account for the occasional attacks of dyskinesia. The only behavioral clue to a potential cellular abnormality is that tottering mice consistently display mild ataxia, with a splayed stance and slight tremor associated with movement. Behaviors that require more complex motor skills, such as swimming and maintaining position on a rotorod, are severely impaired in tottering mice (Syapin, 1982). This deficit in coordinated movement is associated with a small but significant decrease in the size of the cerebellum (Isaacs and Abbott, 1992) due to decreased volume of the molecular layer (Isaacs and Abbott, 1995). Tottering mice are often classified as cerebellar mutants based on these 21 findings, however, as the cerebellum is not traditionally associated with the generation of paroxysmal events, a link between cerebellar abnormalities and attacks of dyskinesia was not immediately apparent.

A study using 2-deoxyglucose uptake during attacks in tottering mice revealed increased activation of brainstem and bilateral thalamic nuclei (Noebels, 1979). These results suggested a subcortical origin of the abnormal neuronal signals mediating the attacks. A more recently developed technique using c-fos immediate-early gene activation to measure changes in polysynaptic neural activity during tottering attacks has produced results consistent with those of the metabolic study. Thirty minutes after the onset of a restraint-induced attack in a tottering mouse, c-fos expression is dramatically increased in the cerebellar cortex, deep cerebellar nuclei, inferior olivary nucleus, pontine nuclei, red nucleus, vestibular nuclei and cerebral cortex

(Campbell and Hess, 1998). A timed comparison between activation of the cerebral cortex and cerebellar cortex reveals that the cerebellum is activated prior to activation of the cerebral cortex.

This result suggests that the cerebellum, rather than the motor cortex, is heavily involved in generating the abnormal signals that lead to attacks in tottering mice (Campbell and Hess, 1998).

This interpretation is consistent with the results of the 2-deoxyglucose study, as brainstem nuclei and the thalamus receive cerebellar afferent innervation.

Output from the cerebellar cortex is required for the expression of tottering mouse attacks. This conclusion is supported by evidence from Purkinje cell ablation experiments; tottering mice that also carry the pcd (Purkinje cell degeneration) mutation lose cerebellar 22

Purkinje cells after the maturation of the cerebellum. These mice did not exhibit spontaneous or restraint-induced attacks (Campbell et al., 1999). In addition, chemical and surgical lesions of the cerebellum have revealed that the frequency and duration of tottering mouse attacks are reduced after the anterior, but not posterior, cerebellum is damaged (Abbott et al., 2000). This result is especially intriguing in light of the fact that the anterior cerebellum is the site of ectopic TH expression in several calcium channel mutant mice (see Table 1.4 and discussion below), and that cell loss in developing leaner mice is most severe in the anterior cerebellum (Herrup and

Wilczynski, 1982). Collectively, the lesion studies support the idea that the cerebellum is a key site in the pathogenesis of attacks in tottering mice.

Cerebellar tyrosine hydroxylase expression in tottering mice

In 1991, a dramatic abnormality in tyrosine hydroxylase (TH) expression was identified in a non-catecholaminergic region of the tottering mouse brain, the cerebellum (Hess and

Wilson, 1991). The TH enzyme is the rate-limiting step in catecholamine biosynthesis, and is normally abundant in the locus coeruleus, striatum and olfactory bulb. Although control mice transiently express TH in posterior Purkinje cells during early postnatal development, this expression is greatly reduced in the adult animal, as mature Purkinje cells rely on GABAergic neurotransmission. In adult tottering mutant mice, posterior parasagittal bands of cerebellar

Purkinje cells were labeled by probes for TH mRNA and protein (Hess and Wilson, 1991). It appears that tottering mice fail to repress TH expression and maintain high levels of both TH mRNA and protein in subsets of posterior Purkinje cells throughout adulthood. These initial 23 findings have been replicated and extended to include Purkinje cells in the anterior cerebellum of tottering mice as well (Austin et al., 1992; Abbott et al., 1996). TH expression in the anterior cerebellum is truly ectopic, since controls never express TH in Purkinje cells from this region.

Aberrant TH expression is observed in leaner (Austin et al., 1992), rolling (Sawada et al., 1999), and lethargic (Yoshor et al., 1997) mouse cerebellar Purkinje cells as well (Table 1.4).

The cause and significance of abnormal TH expression in the tottering mouse cerebellum is unknown, as other enzymes required for catecholamine synthesis are absent from tottering

Purkinje cells (Hess and Wilson, 1991). However, TH mRNA expression is highly dependent on calcium influx, as calcium channel activation promotes and blockade diminishes TH mRNA expression in vitro (Vidal et al., 1989; Brosenitsch et al., 1998; Cigola et al., 1998). Presumably alterations in the calcium-signaling system due to calcium channel mutations play a role in abnormal TH mRNA expression in tottering, leaner, lethargic and rolling mice. In this respect,

TH-containing neurons in the adult mouse cerebellum could be considered a molecular marker for the effects of the mutation on the calcium signaling system. A better understanding of the molecular events that give rise to abnormal cerebellar TH expression may provide clues into pathophysiology common to the calcium channel mutant mice. 24

Conclusions and Experimental Questions

The tottering mouse appears to be a good candidate system in which to investigate the cellular mechanisms underlying human episodic disorders caused by ion channelopathy. The studies presented herein address the following questions:

1 How does the tottering mutation in the P/Q-type calcium channel affect voltage dependent

calcium channel regulation and function in various brain regions in adult and developing

mice?

2 When is the phenotypic onset of motor attacks and aberrant cerebellar TH mRNA expression

in tottering mice, and does calcium influx through L-type calcium channels influence

aberrant gene expression?

3 Which of the trigger factors and treatment strategies involved in human episodic disorders

also influence the expression of paroxysmal dyskinesia in tottering mice; do these agents

share a common mechanism?

A goal of these studies was to provide further insight into the largely unexplored question of calcium channel regulation in the context of development and disease in the mammalian central nervous system. In addition, investigations into possible links between trigger factors and the precipitation of attacks were conducted to examine this intriguing aspect of many human disorders arising from ion channelopathies. The results of these studies suggest that a genetic mouse model of ion channelopathy retains features remarkably similar to human episodic 25 disorders, and that a complex interaction between several neural systems may be involved in the periodic expression of symptoms. Further study using this and other models of ion channelopathy will bring to light new aspects of this unique group of disorders, with the hope that identifying common themes will lead to better therapeutic control over the episodic signs and symptoms characteristic of ion channelopathies. 26

CHAPTER 2: TOTTERING MOUSE DEVELOPMENT

Chapter Summary

Voltage dependent calcium channels play a critical role during normal development of the mammalian central nervous system. A mutation in the gene encoding the pore-forming subunit of murine P/Q-type calcium channels generates a delayed onset syndrome in mice. The tottering syndrome includes aberrant cerebellar gene expression and attacks of a movement disorder, which are spontaneous or induced by stress in adult mice, but are not apparent in developing animals until sometime during the third or fourth week after birth. A careful developmental analysis of aberrant gene expression and attack onset may reveal the timing of underlying molecular events that give rise to these phenotypes. Since previous studies suggest that L-type calcium channel misregulation may contribute to tottering mouse attacks, the mRNA expression and functional activity of these channels during development was investigated as a potential molecular mechanism responsible for the onset of attacks.

Tottering mice between the ages of twenty to seventy days were subjected to short-term restraint to investigate the age of onset for stress-induced attacks. Phenotypic onset occurred during an extremely short time period between postnatal days twenty-one and twenty-two: tottering mice were largely unresponsive to restraint stress at p21 but by p22 exhibited attack frequencies comparable to adults. The developmental onset of tyrosine hydroxylase (TH) mRNA in control and tottering mouse Purkinje cells was assessed to identify the age at which aberrant

TH gene expression is manifested in the tottering mouse cerebellum. Abnormal expression of TH 27 mRNA was observed in the anterior cerebellum of tottering mice only after the onset of attacks.

Molecular analyses of calcium channel subunit mRNA expression and calcium uptake were conducted before and after the onset of attacks, at p21 and p28. Although calcium uptake was

α reduced in older animals, consistent with a reduction in 1A mRNA, neither changes in calcium channel mRNA expression nor calcium uptake differed between tottering and control mice. The results suggest that a developmental switch between p21 and p22 renders tottering mice susceptible to restraint-induced attacks. Ectopic TH mRNA expression, apparent after the onset of attacks, may be a secondary consequence of the episodic motor phenotype. As changes in calcium channel mRNA expression and activity are not correlated with attack onset, additional factors influencing the developmental onset of the phenotype should be considered. 28

Rationale

α Tottering (tg) mice inherit a missense mutation in the 1A subunit of P/Q-type calcium channels (Fletcher et al., 1996; Doyle et al., 1997), a subtype abundantly expressed in the cerebellum and hippocampus (Hillman et al., 1991). The mutation reduces whole cell calcium current measured in dissociated tottering mouse Purkinje cells by approximately 40%

(Wakamori et al., 1998). Surprisingly, however, several lines of evidence suggest that an increase in calcium flux in the cerebellum may be responsible for some of the tottering phenotypes. Affected adult mice exhibit attacks of a movement disorder that can be initiated by an L-type calcium channel agonist and prevented by L-type calcium channel antagonists

(Campbell and Hess, 1999). Further, the tottering mutation is associated with increased density

α of L-type calcium channel binding sites and 1C mRNA expression in the cerebellum, suggesting that L-type calcium channel upregulation might compensate for the tottering mutation in P/Q- type calcium channels (Campbell and Hess, 1999). Taken together, these results have been interpreted as evidence that increased L-type calcium channel activity in the cerebellum plays a role in generating attacks in tottering mice (Campbell and Hess, 1999). In support of this hypothesis, a genetic lesion of cerebellar Purkinje cells can eliminate attacks (Campbell et al.,

1999). These findings indicate that attacks may result from periodically increased calcium channel activity in the cerebellum indicative of a gain-of-function, rather than the loss-of- function predicted by the recessive mutation. As the exact developmental onset of attacks is not 29 known, it has been difficult to design studies to identify cellular abnormalities before and after the onset of this phenotype.

Adult tottering mice display unusually robust expression of tyrosine hydroxylase (TH) mRNA and protein in subsets of cerebellar Purkinje cells (Hess and Wilson, 1991; Austin et al.,

1992; Abbott et al., 1996). Although control mice express TH in posterior Purkinje cells transiently during development, TH transcription is normally suppressed by postnatal day 35

(Hess and Wilson, 1991) as mature Purkinje cells utilize GABA as a neurotransmitter. In adulthood, tottering mice fail to suppress TH expression in the posterior cerebellum, and ectopically express TH in Purkinje cells of the anterior cerebellum. Although TH protein is present in tottering mouse Purkinje cells, it is not likely that these cells synthesize catecholamines, as a second requisite enzyme, aromatic amino acid decarboxylase, is not expressed in this region (Hess and Wilson, 1991). Therefore, TH expression in tottering mouse

Purkinje cells likely reflects altered TH gene transcription rather than a fundamental change in neurotransmitter phenotype.

TH is a calcium-responsive gene; L-type calcium channels have been implicated in TH gene transcription in vitro (Vidal et al., 1989; Brosenitsch et al., 1998; Cigola et al., 1998). Thus the relationship between increased L-type calcium channel binding sites and aberrant TH mRNA expression in tottering cerebellar Purkinje cells suggests that, in addition to motor attacks, this cellular phenotype may be linked to calcium current through L-type calcium channels. If TH mRNA expression is indeed regulated by calcium influx in the tottering mouse cerebellum, it 30 could be considered a unique marker for those Purkinje cells most affected by altered calcium signaling. Interestingly, the timing of TH expression in the cerebellum of control mice is loosely correlated with the developmental expression profile of L-type calcium channels in this region.

L-type calcium channel expression in the rodent brain is initially low but peaks during the second and third week of postnatal development (Erdman et al., 1983; Kazazoglou et al., 1983;

Mourre et al., 1987; Litzinger et al., 1993) followed by a decline to the moderate levels seen in adults. P/Q-type calcium channels show little activity until sometime during the second week after birth and then predominate in several synapses. For example, P/Q-type calcium channel antagonists have little effect on neurotransmitter release in the cerebellum of normal rodents prior to p13, but by p19 almost completely abolish cerebellar IPSPs (Iwasaki et al., 2000).

Perhaps not coincidentally, tottering mice exhibit few abnormalities during the first two to three weeks of life. A mutation in a calcium channel subtype could conceivably alter the developmental expression of other subtypes and initiate a sequence of molecular events that lead to cellular and behavioral abnormalities.

We have undertaken a study of tottering mouse development to determine the precise onset of the motor attacks as well as their temporal relationship to cerebellar TH mRNA expression. Additionally, the effect of the P/Q-type calcium channel mutation on the expression

α of L-type calcium channel 1C subunit mRNA and potassium-stimulated calcium uptake in the cerebellum was assessed. Developmental investigations may shed light on the issue of whether 31 aberrant cerebellar TH expression and altered calcium channel regulation is a cause or a consequence of the abnormal neurological signals that give rise to attacks of dyskinesia in tottering mice. 32

Materials and Methods

Animals

C57BL/6J +/+, +/tg and Os/+; +/tg mice were obtained from The Jackson Laboratories and bred at the Pennsylvania State University College of Medicine. The Os (oligosyndactyly) gene mutation is a dominant, homozygous lethal mutation tightly linked to the recessive tottering mutation on mouse chromosome 8. Heterozygous Os/+; +/tg mice inherit a paw malformation that causes the third and fourth digits to fuse, enabling immediate identification of heterozygous and homozygous tottering mice at birth. Adult tottering mice (16-24 weeks of age) were identified either by analysis of PCR amplified simple sequence length polymorphisms in

C57BL/6J +/tg x C57BL/6J +/tg cross progeny (Campbell and Hess, 1997), or by the absence of oligosyndactylism in Os +/+ tg x Os +/+ tg cross progeny. The day of birth was designated p0, and all animals were experimentally naïve. Age and gender matched +/+ C57BL/6J mice were used as controls. All procedures were in accordance with the Guide for the Care and Use of

Animals.

Attack induction

To determine the frequency of attacks induced by brief restraint stress, 20, 21, 22, 24, 28, or 70-day old tottering mice were immobilized in plastic syringes (all ages but p70 were restrained in 30cc syringes, p70 mice were placed in 60cc syringes). Restraint lasted for 10 minutes and was followed by release to a novel cage for 30 min. The mice were returned to the home cage and observed for another 10 min. Mice were scored for the presence or absence of 33 attacks every 5 min following release. Statistical analyses on the frequency of attacks were carried out using the chi-square test.

Calcium uptake assays

Synaptosome Preparation: Synaptosomes were prepared from the cerebellum and forebrain (minus olfactory bulbs and brainstem) of developing control and tottering mice. The mice were deeply anesthetized with carbon dioxide and killed by decapitation. Brains were rapidly removed and dissected on ice. Brain regions were homogenized in 10 volumes (w:v) cold 0.32M sucrose solution using 10 strokes with the homogenizer operating at the slowest speed possible. Homogenates were centrifuged at 1000g at 4oC for 10 min. The supernatant was removed and centrifuged at 17000g for 30 min at 4oC. The synaptosomal pellet was

o resuspended in 37 C HEPES buffer (1mM CaCl2, 3.5mM KCl, 0.4mM KH2PO4, 1.2mM MgSO4,

125mM NaCl, 10mM glucose, 20mM HEPES, pH 7.4 adjusted with 1M NaOH) at approximately 15 mg wet weight tissue/ml. Synaptosomes were incubated for 15 min at 37oC before use in uptake assays. Protein was measured using the Pierce BCA Protein Assay Kit

(Rockford, IL) according to the manufacturer’s protocol.

Calcium uptake: For measurements of calcium uptake, 100µl of the pre-warmed synaptosome suspension was added to tubes containing 300µl basal (4mM KCl) or stimulation

(25mM or 60mM KCl, with isomolar salt replacement) buffer and 2µCi 45Ca2+. After 15 sec, uptake was halted by the addition of 3ml ice cold HEPES buffer and rapid filtration over 34

Whatman GF/B filters (Brandel), and washed three times with an equal volume of ice-cold buffer. Filters were air-dried, then placed in ScintiVerse (Fisher) scintillation fluid for 24 hr to allow the release of 45Ca2+ from synaptosomes. Radioactivity was measured by liquid scintillation spectroscopy at an efficiency of 100%. Data are expressed as nMol 45Ca2+/mg protein. Statistical analyses were performed by two-way analysis of variance with genotype and age as factors.

In situ hybridization

Control and tottering mice were deeply anesthetized with carbon dioxide and killed by decapitation. Brains were rapidly removed and frozen in isopentane at -40oC and stored at -70oC.

Twenty micrometer sagittal sections were cut using a cryostat and thaw mounted on Superfrost

Plus glass slides (Fisher, Pittsburgh, PA). After drying, the slide-mounted sections were stored at

-70oC.

α An 1A-specific cDNA was generated by reverse transcription-PCR. A 554bp fragment

α corresponding to base pairs 5736 to 6289 of the mouse calcium channel 1A cDNA (GenBank

α U76716) was amplified and inserted into pBluescript II SK(-). The template for 1C sense and antisense cRNA probes was a unique 418bp fragment (basepairs 8596-9014) from the 3' untranslated region of the gene, subcloned from the I.M.A.G.E. consortium clone #AA462894

(Lawrence Livermore National Laboratory, Livermore, CA) at EcoRI restriction sites into pBlueScriptII KS(-). A 1.7kb fragment of the mouse TH cDNA in pBluescript KS+ (gift from T. 35

Nagatsu, Fujita Health University, Japan) was the template for TH sense and antisense transcription reactions. In vitro transcription was performed for approximately 2 hr at 37oC in a

25µl volume containing 40mM Tris, pH 7.9, 6mM MgCl2, 2mM dithiothreitol (DTT), 40U

RNase inhibitor (Promega, Madison, WI), 400µM each ATP, GTP and UTP, 10µM [ 35S]CTP

(800Ci/mmol), 1µg linearized plasmid and 20U RNA polymerase (Promega, Madison, WI).

Following transcription reactions, DNA templates were removed by RNase-free DNase

(Promega, Madison, WI) digestion for 30 min at 37oC. Transcripts were size-reduced by alkalai treatment with 0.2M NaOH for 45 min on ice. Probes were purified with phenol:chloroform:isoamyl alcohol (25:24:1) and separated on a G50 Sephadex Nick column

(Pharmacia, Piscataway, NJ) to remove unincorporated nucleotides.

Slide-mounted sections were pretreated by fixation in buffered 4% formaldehyde for 15 min at room temperature followed by a 5 min rinse in 0.1M phosphate buffered saline (PBS).

Slides were treated with 0.25% acetic anhydride in 0.1M triethanolamine-HCl/0.15M NaCl (pH

8.0) for 10 min and rinsed in 2X standard sodium citrate (1X SSC; 0.15M NaCl, 0.015M sodium citrate). Sections were dehydrated in graded for 1 min each followed by a 5 min incubation in chloroform. One min incubations in 100% and 95% ethanol were followed by air drying.

Slides were hybridized with 100µl of hybridization buffer containing 7.5ng cRNA probe in 50% formamide, 0.75M NaCl, 20mM 1,4-piperazine diethane sulfonic acid, pH 6.8, 10mM

EDTA, 10% dextran sulfate, 5X Denhardt’s solution (0.02% bovine serum albumin, 0.02% 36 ficoll, 0.02% polyvinylpyrolidone), 50mM DTT, 0.2% sodium dodecyl sulfate and 100µg/ml each salmon sperm DNA and yeast tRNA. Slides were coverslipped, sealed with Royalbond Grip

o α contact cement (Columbia Aluminum Products, CA) and hybridized for 16 hr at 56 C ( 1A and

o α TH) or 58 C ( 1C).

Following overnight hybridization, coverslips were removed in 4X SSC with 300mM 2- mercaptoethanol at room temperature. Slides were incubated in this solution for 15 min followed by 15 min in 4X SSC alone. The slides were treated with 50µg/ml pancreatic RNase A in 0.5M

NaCl, 50mM Tris, pH 8.0, 5mM EDTA for 30 min at 37oC, washed in graded salt solutions (2X,

1X, and 0.5X SSC each for 5 min at 56oC), and in 0.1X SSC at 65oC for 30 min. Slides were dipped in 60% ethanol with 0.33M ammonium acetate and air dried. Sections were exposed to x- ray film (DuPont Cronex) and subsequently dipped in Kodak NTB-2 photographic emulsion

o (diluted 1:2 with dH2O), exposed at 4 C and developed in Kodak D19 developer.

The ontogeny of TH mRNA expression in the cerebellum was determined by visual

α inspection of autoradiographic film images and darkfield views from emulsion-dipped slides. 1A

α and 1C mRNA expression data was recorded from autoradiographic film images using a

Microcomputer Imaging Device (MCID) system (Imaging Research, Inc., Ontario, Canada) using M5+ software version 4.0. Autoradiographic film images were captured using a high- resolution camera and digitally transferred to computer. Densitometric data were recorded as raw optical density (R.O.D.) values using various automated selection tools in discrete brain regions. 37

Measures were taken from a minimum of four slides from each animal (n = 3 animals per group).

Data were analyzed using two-way analysis of variance with age and genotype as factors. 38

Results

The onset of restraint-induced attacks in developing tottering mice

Anecdotal reports suggest that tottering mice do not display attacks of dyskinesia until sometime during the third or fourth week of age. Naïve tottering mice varying in age from p20 to p70 were tested using the restraint paradigm to determine the precise onset of the attacks.

Attack progression in developing tottering mice was similar to that reported for adults. The onset of an attack began with contraction of the hindlimbs and flattening of the trunk, causing a waddling gait. The attack progressively involved the entire body, severely impairing mobility.

As the lower body recovers,

the upper body (forelimbs,

neck and jaw) was involved

in extension/contraction

movements. Upon recovery,

the mice returned to baseline

motor behavior.

No attacks were

observed in animals tested at

p20 (0/9), and only 11%

(1/9) animals displayed an 39 attack after restraint at p21. However, at p22, a large increase in the percent of affected tottering mice was observed. The percentage of mice exhibiting an attack after restraint at p22 (6/11, or

55%) was similar to adults, and there was little change in the response at p24 (4/8, or 50%), p28

(5/8, or 63%) or p70 (5/10, or 50%).

TH mRNA expression during development

Dopaminergic cell groups in the substantia nigra, locus coeruleus, and olfactory bulb normally express the catecholamine synthesis enzyme, tyrosine hydroxylase. In sagittal brain sections from developing control and tottering mice, TH mRNA expression in these regions was apparent on P1 (data not shown), consistent with previous reports. However, transient TH mRNA expression is observed in the cerebral cortex and in a subset of cerebellar Purkinje cells during the first weeks of postnatal development (Satoh and Suzuki, 1990; Hess and Wilson,

1991). As expected, TH mRNA was expressed in scattered cells throughout the cerebral cortex at p14 and 21 and appeared to diminish with increasing age in mice of both genotypes (Figure

2.2). Cerebellar TH mRNA expression was not apparent at p7 in either tottering or control mice

(data not shown) but was detected in the posterior cerebellum by p14 in mice of both genotypes

(Figure 2.2). In control mice, TH mRNA was not detected in the anterior cerebellum at any age studied, but tottering mice expressed TH mRNA in the anterior cerebellum beginning at p28. By p70, control mice had largely suppressed cerebellar TH mRNA expression, while in tottering mice TH mRNA expression persisted in both posterior and anterior cerebellum.

41

Calcium uptake in the developing tottering mouse brain

Calcium uptake was measured in

control and tottering mouse synaptosomes

prepared from cerebellum and forebrain of

developing animals. Limited tissue

availability in developing brains precluded

the addition of specific calcium channel

antagonists to further differentiate calcium

influx through various channel subtypes. No

effect of genotype was observed after

stimulation with either 25mM or 60mM

potassium in either of the brain regions

studied. However, at p28, a significant decrease in 60mM potassium-stimulated calcium uptake in control and tottering synaptosomes was observed in both brain regions. In the cerebellum, control synaptosomes took up 31% less calcium at p28 than at p21; in tottering synaptosomes the difference was 37% (p < 0.01, Figure

2.2B). In the forebrain, 60mM potassium uptake was reduced by 17% and 29%, respectively

(p<0.05, Figure 2.3B). 42

α 1 subunit mRNA expression during

development

mRNA expression for the calcium

α channel subunits encoding P/Q-type ( 1A)

α and L-type ( 1C) was assessed in the

developing brain. Although no effects of

genotype were apparent, the expression of

α 1A mRNA at p70 steadily declined from

p14 values by roughly 25 to 50% in the

cerebellum, cortex and hippocampus (Table

2.1, p<0.05).

α Expression of 1C mRNA was

α Table 2.1 1A mRNA expression in developing tottering and control mice P14 P21 P28 P42 P70 Cerebellum Control 0.706 ± 0.001 0.480 ± 0.061 0.476 ± 0.103 0.389 ± 0.088 0.384 ± 0.101 Tottering 0.575 ± 0.128 0.573 ± 0.084 0.362 ± 0.062 0.303 ± 0.039 0.334 ± 0.076 Hippocampus Control 0.568 ± 0.129 0.386 ± 0.059 0.357 ± 0.091 0.327 ± 0.044 0.264 ± 0.068 Tottering 0.438 ± 0.017 0.405 ± 0.036 0.325 ± 0.055 0.295 ± 0.065 0.297 ± 0.078 Cortex Control 0.261 ± 0.072 0.218 ± 0.020 0.172 ± 0.033 0.147 ± 0.022 0.115 ± 0.001 Tottering 0.217 ± 0.027 0.204 ± 0.003 0.195 ± 0.029 0.161 ± 0.031 0.158 ± 0.029 Images captured from autoradiograms of in situ hybridization experiments (n=3 animals/genotype with the exception of p14 control, n=2). Data represent mean ± S.E.M. in raw optical density (ROD) units. No α significant differences were observed among genotypes, however, 1A mRNA expression declined over age in all regions (p<0.05). 43 assessed in the cerebellum of developing control and tottering mice. Previous reports have

α shown an increase in 1C mRNA in adult tottering mouse cerebellar Purkinje cells, however, no significant differences between tottering and control mice of any age were observed in the present study (Table 2.2).

α Table 2.2 1c mRNA expression in developing tottering and control mice P14 P21 P28 P42 P70 Cerebellum Control 0.193 ± 0.029 0.200 ± 0.034 0.247 ± 0.016 0.219 ± 0.037 0.226 ± 0.032 Tottering 0.207 ± 0.030 0.219 ± 0.027 0.237 ± 0.022 0.184 ± 0.015 0.184 ± 0.012 Images captured from autoradiograms of in situ hybridization experiments (n=3 animals/genotype with the exception of p14 control, n=2). Data represent mean ± S.E.M. in raw optical density (ROD) units. α There were no significant effects of age or genotype on 1c mRNA expression. 44

Discussion

The present study documents an extremely narrow time window for the development of attacks of dyskinesia in tottering mice, suggesting a precipitous change in the neural network mediating attacks. Restraint stress, which reliably induces attacks in adults (Campbell and Hess,

1998, 1999; Campbell et al., 1999), was an ineffective trigger at p21. Remarkably, tottering mice exhibited restraint-induced attacks at frequencies comparable to adults just twenty-four hours later. Ectopic TH mRNA expression was observed in the anterior cerebellum of tottering mice after the onset of restraint-induced attacks. Calcium channel misregulation in the cerebellum has been associated with attacks in adult tottering mice (Campbell and Hess, 1999); therefore, calcium channel activity was investigated during tottering mouse development. There was little

α α difference in calcium channel 1A or 1C subunit mRNA expression between tottering and control mice at any age studied. Likewise, calcium uptake experiments did not reveal differences between tottering and control mice either before or after attack onset. These results suggest that other factors influencing calcium channel activity or neuronal excitability may play a role in the dramatic onset of restraint-induced attacks in developing tottering mice.

Central nervous system development in tottering mice presumably occurs in the context of abnormal calcium channel activity due to the P/Q-type calcium channel mutation and misregulation of cerebellar L-type calcium channels. It was hypothesized that changes in the development of the calcium channel system plays a role in the onset of attacks and aberrant gene expression. However, no significant differences were observed in the mRNA expression profiles 45

α α of 1A, the subunit affected by the tottering mutation, or 1C, a subunit predicted to underlie L-

α type calcium channel upregulation in tottering mouse cerebellum. Decreases in 1A mRNA expression from p14 to p70 were consistent with the decrease in 60mM potassium-stimultated calcium influx between p21 and p28 in mice of both genotypes. However, neither changes in calcium channel expression nor activity appear to account for the dramatic onset of restraint- induced attacks during tottering mouse development. As slight changes in L-type calcium channel density, demonstrated in adult tottering mouse cerebellum (Chapter 3 and Campbell and

Hess, 1999), may be obscured in the calcium uptake assays, the ontogeny of radioligand binding sites for L-type calcium channels should also be investigated in developing tottering mice.

The temporal relationship between aberrant cerebellar TH expression and motor dysfunction has been an open question in tottering mouse development. In the present study, TH mRNA expression in the posterior cerebellum was found in both control and tottering mouse

Purkinje cells beginning at p14. Cerebellar TH mRNA expression does not appear qualitatively different in tottering mice until after the onset of restraint-induced attacks. At p28 and later ages,

TH mRNA expression was found in the tottering mouse anterior cerebellum, a region devoid of

TH expression in control mice at all ages studied. This finding demonstrates that cerebellar TH mRNA expression is present at p14, earlier than previously predicted (Hess and Wilson, 1991;

Austin et al., 1992). However, ectopic TH mRNA expression followed the onset of attacks, suggesting that abnormal activity in the anterior cerebellum may contribute to this tottering mouse phenotype. In light of this result, it is interesting that anterior, but not posterior, cerebellar 46 lesions reduce the incidence and duration of tottering mouse attacks (Abbott et al., 2000). By p70, cerebellar TH mRNA expression was largely eliminated in control mice while expression remained abundant in posterior and anterior cerebellum in tottering mice, consistent with previous reports (Austin et al., 1992; Abbott et al., 1996). Compared to the previous developmental study (Hess and Wilson, 1991), the present study describes a slightly earlier onset and longer duration of cerebellar TH mRNA expression in control mice, as well as ectopic expression in the anterior cerebellum of tottering mice. These differences are likely due to increased specificity of the TH probe, permitting a lower limit of detection in the present study.

Overall, the appearance of ectopic TH mRNA expression in anterior cerebellar Purkinje cells suggests that this phenotype may be a secondary consequence of abnormal cerebellar signals associated with motor attacks.

The developmental onset of attacks may be related to the postnatal development of afferent synaptic connections within the cerebellar cortex. Climbing fibers, which originate in the inferior olive and provide powerful excitatory input to Purkinje cells, contact Purkinje cells during the first week after birth and initially multiple climbing fibers innvervate a single Purkinje cell. Over the next weeks of development, however, elimination of climbing fiber synapses occurs so that by approximately p21, individual Purkinje cells are innervated by a single climbing fiber (reviewed by (Sotelo, 1989). Recently, a very small critical period for multiple climbing fiber elimination was established in the mouse. NMDA receptor blockade during p15-

16, but not before or after this time, prevented climbing fiber synapse elimination and produced 47 chronic motor incoordination (Kakizawa et al., 2000). This study suggests that NMDA receptor activation during a very small time frame can have long-term consequences on cerebellar synapses. As excitatory neurotransmission is reduced in the tottering mouse brain (Caddick et al.,

1999; Ayata et al., 2000) and the cerebellum has been implicated in attacks of dyskinesia

(Campbell and Hess, 1998; Campbell et al., 1999; Abbott et al., 2000), it seems likely that impaired glutamatergic signaling during the maturation of cerebellar synapses may produce synapses that generate abnormal signals in tottering mice.

Alternatively, it is possible that the cerebellar circuitry is not sufficiently mature to allow the expression of motor attacks in tottering mice until p22. Several molecules required for membrane excitability, such as sodium channels and Ca2+-dependent potassium channels, are not abundantly expressed in the cerebellum until the second to third week after birth (Mourre et al.,

1987). Normal mice injected with the L-type calcium channel agonist, BAY K8644 are capable of displaying dyskinesia similar to tottering mouse attacks as early as p20 (personal observation), arguing against interpretations solely based on the developmental integrity of cerebellar circuitry.

However, synaptic maturation may be delayed in the tottering mouse cerebellum as a result of the P/Q-type calcium channel mutation. Further study to assess other components of the ion channel system in developing tottering mice is required to critically evaluate this possibility.

Last, it is possible that the expression of the restraint-induced phenotype may depend on hormonal or neurotransmitter systems reactive to stress. In rodents, restraint produces a stress- like activation of the hypothalamic-pituitary-adrenal (HPA) axis, initiating a cascade of stress 48 hormones that can have effects on the central nervous system (Armario and Castellanos, 1984;

Hauger et al., 1988; Berridge and Dunn, 1989), and on ion channels in particular (Xie and

McCobb, 1998; Yu and Shinnick-Gallagher, 1998). However, it has been suggested that developing animals exhibit a so-called stress-hyporesponsive period, during which stressful stimuli are less effective at initiating HPA axis activation (reviewed by Dallman, 2000). The hyporesponsive period is predicted to end sometime during the second to third week after birth, roughly corresponding to the timeframe of attack onset in tottering mice. It is tempting to speculate that the period from p21 to 22 describes the end of the stress-hyporesponsive period in tottering mice, rendering them susceptible to effects of restraint stress. Hormonal systems in tottering mice have not been studied previously, but comprise a class of neuromodulators that may significantly influence the expression of tottering mouse attacks.

Mutations in genes encoding neuronal voltage dependent calcium channels and subsequent disruptions in calcium signaling processes are likely to disturb the complex orchestration of neural development, resulting in phenotypic abnormalities. Tottering mice become susceptible to a stress-induced movement disorder over an extremely short period, providing an in vivo system to study factors that influence neural development in the context of an ion channel mutation. Ectopic gene expression occurs after the onset of motor attacks, suggesting that this cellular phenotype is a consequence of the cerebellar activation associated with the episodic movement disorder. Ongoing study into this critical period may lead to novel 49 strategies to intervene in the development of tottering mouse phenotypes, and in similar genetic disorders in humans. 50

CHAPTER 3: L-TYPE CALCIUM CHANNELS IN TOTTERING MOUSE

PHENOTYPES

Chapter Summary

It has been suggested that the increased density of cerebellar L-type calcium channel binding sites reflects functional compensation for the P/Q-type calcium channel mutation, and that cerebellar L-type calcium channel misregulation contributes to the attacks in tottering mice.

Potassium-stimulated calcium uptake was measured in control and tottering mouse synaptosomes in the presence of the L-type calcium channel antagonist, verapamil, to estimate the relative contribution of L-type calcium channels to total calcium uptake in several regions of the tottering mouse brain. Additionally, L-type calcium channel regulation in vivo was investigated in animals chronically challenged with the L-type calcium channel antagonist, nimodipine. These experiments were conducted to determine the effects of chronically reduced calcium influx on the density of L-type calcium channels in control and tottering mice, and to analyze the potential link between L-type calcium channels and the aberrant expression of the calcium-responsive gene, TH, in tottering cerebellar Purkinje cells. The increase in tottering mouse cerebellar L-type calcium channel binding sites was confirmed using the dihydropyridine ligand, [3H]PN200-110, however, there was little change in the contribution of L-type calcium channels to total calcium influx in any region of the tottering mouse brain studied. Repeated exposure to the L-type calcium channel antagonist, nimodipine, did not alter the number of L-type calcium channel binding sites in mice of either genotype. The aberrant TH mRNA expression in tottering mouse 51

Purkinje cells was significantly decreased following chronic nimodipine treatment, suggesting that calcium influx promotes this cellular phenotype. Taken together, these data suggest that altered calcium influx influences aberrant cerebellar TH expression. 52

Rationale

Voltage-dependent calcium channels regulate several key events in the central nervous system, including neurotransmitter release, activation of intracellular second messengers, and transcription. As calcium channel mutations have recently been linked to several disorders in both humans (Ophoff et al., 1996; Jodice et al., 1997; Jen et al., 1999) and mice (Fletcher et al.,

1996; Burgess et al., 1997; Doyle et al., 1997; Letts et al., 1998; Mori et al., 2000), there is growing interest in the mechanisms by which calcium channels are regulated in vivo. A mutation in one of the six subtypes of voltage-dependent calcium channels (L-, N-, P-, Q-, R-, and T-type) may dramatically impact the neuronal processes associated with that subtype. It has been suggested that a mutation in one calcium channel subtype alters the expression or function of other subtypes (McEnery et al., 1998; Campbell and Hess, 1999; Jun et al., 1999), but the extent to which these changes influence the neurological abnormalities generated by a calcium channel mutation is not known.

The tottering mutation affects P/Q-type calcium channels (Wakamori et al., 1998), while

L-type calcium channel binding sites are increased in tottering mouse cerebella and are implicated in motor attacks (Campbell and Hess, 1999). Additionally, the aberrant expression of

TH in cerebellar Purkinje cells suggests that calcium influx through L-type calcium channels is perturbed, as calcium positively regulates TH transcription (Vidal et al., 1989; Brosenitsch et al.,

1998; Cigola et al., 1998). 53

Several questions regarding the role of L-type calcium channels in the tottering mouse cerebellum were addressed in the present studies. First, does increased calcium influx through cerebellar L-type calcium channels compensate for the effects of the P/Q-type calcium channel mutation? Second, do tottering mice regulate L-type calcium channels differently than control mice? Last, is aberrant TH mRNA expression in Purkinje cells a result of increased calcium influx in the tottering mouse cerebellum? The following experiments provide evidence regarding the regulation and function of L-type calcium channels in response to the initial tottering mutation, and to subsequent long-term administration of a calcium channel antagonist. 54

Materials and Methods

Animals

Mice were maintained and genotyped as described (Chapter 2). Adult tottering mice (16-

24 weeks of age) were used, while age- and gender-matched C57BL/6J +/+, +/tg or Os +/+ tg mice served as controls, as indicated.

Calcium uptake assays

Synaptosome Preparation: Synaptosomes were prepared from the cerebellum, hippocampus, cortex and striatum of adult control (C57BL/6 +/+) and tottering mice. The mice were deeply anesthetized with carbon dioxide and killed by decapitation. Brains were rapidly removed and dissected on ice. Brain regions were homogenized in 10 volumes (w:v) cold 0.32M sucrose solution using 10 strokes with the homogenizer operating at the slowest speed possible.

Homogenates were centrifuged at 1000g at 4oC for 10 min. The supernatant was removed and centrifuged at 17000g for 30 min at 4oC. The synaptosomal pellet was resuspended in 37oC

HEPES buffer (1mM CaCl2, 3.5mM KCl, 0.4mM KH2PO4, 1.2mM MgSO4, 125mM NaCl,

10mM glucose, 20mM HEPES, pH 7.4 adjusted with 1M NaOH) at approximately 15 mg wet weight tissue/ml. Verapamil (Research Biochemicals International, Natick, MA) or saline vehicle was added to the reaction tubes and synaptosomes to a final concentration of 100µM.

The phenylalkylamine verapamil binds to a non-dihydropyridine site on the L-type calcium channel, and consistently antagonized calcium influx in preliminary experiments. Synaptosomes 55 were incubated for 15 min at 37oC before use in uptake assays. Protein was measured using the

Pierce BCA Protein Assay Kit (Rockford, IL) according to the manufacturer’s protocol.

Calcium uptake: For measurements of calcium uptake, 100µl of the pre-warmed synaptosome suspension was added to tubes containing 300µl basal (4mM KCl) or stimulation

(25mM or 60mM KCl, with isomolar salt replacement) buffer and 2µCi 45Ca2+. After 15 sec, uptake was halted by the addition of 3ml ice cold HEPES buffer and rapid filtration over

Whatman GF/B filters (Brandel), and washed three times with an equal volume of ice-cold buffer. Filters were air-dried, then placed in ScintiVerse (Fisher) scintillation fluid for 24 hr to allow the release of 45Ca2 from synaptosomes. Radioactivity was measured by liquid scintillation spectroscopy at an efficiency of 100%. Data are expressed as nMol 45Ca2+/mg protein. Statistical analyses were performed using one-way repeated measures analysis of variance.

Nimodipine administration and behavioral assessment

Nimodipine (Research Biochemicals International, Natick, MA) was prepared at 2mg/ml in 14.5% ethanol/ 2.25% Tween80 in 0.9% Saline. The mice received a 5ml/kg volume of nimodipine or vehicle twice daily (20mg/kg/day) at 12 hr intervals for a total of 14 days. The mice were transported from the virarium to the laboratory daily; this procedure reliably induces attacks in tottering mice (see Chapter 5). Attacks following transport were recorded immediately upon arrival to the laboratory in the morning throughout the course of the study to determine effects of repeated nimodipine exposure on attack frequency. 56

[3H]PN200-110 Saturation Binding Assays

Following the fourteenth day of nimodipine administration, age- and gender-matched tottering and control (C57BL/6J +/+, +/tg or Os +/+ tg) mice were allowed a 24hr clearance period to eliminate residual nimodipine. The mice were then deeply anesthetized with carbon dioxide and killed by decapitation. Brains were rapidly removed, dissected into cerebellum and forebrain (including all tissue rostral to the inferior colliculus), frozen in isopentane at -40oC and stored at -70oC. In most cases, saturation analysis was performed using tissue from a single brain; in some cases, it was necessary to pool tissue from two brains to obtain enough tissue for analysis. Tissue was homogenized in 100 volumes of ice-cold 50mM Tris-HCl buffer (pH 7.5) with a Tekmar tissue homogenizer (Cincinnati, OH) at setting 70 for 15 sec. The homogenates were centrifuged at 30,000 X g for 12 min and the supernatant discarded. Pellets were resuspended by homogenization in ice-cold buffer and centrifuged at 30,000 X g for 12 min.

Pellets were resuspended at 10mg wet weight tissue/ml buffer (forebrain) and 20mg wet weight tissue/ml buffer (cerebellum). Binding assays were performed in a total volume of 2ml in 50mM

Tris-HCl buffer for 45 min at 37oC. Each reaction contained either 1.0mg/ml (forebrain) or

2.0mg/ml (cerebellum) tissue. Concentrations of [3H]PN200-110 (82 Ci/mMol, Amersham,

Piscataway, NJ) ranging from 0.0125 to 0.4nM defined total binding; nonspecific binding was determined in the presence of 1µM nifedipine. Saturation analyses consisted of five or six concentrations of [3H]PN200-110 depending on tissue availability. Reactions were terminated by rapid filtration over Whatman GF/C filters and washed three times with an equal volume of ice- 57 cold buffer. Filters were air-dried, then placed in ScintiVerse scintillation fluid. Radioactivity was measured by liquid scintillation spectroscopy at an efficiency of ~55%. Data were analyzed using Ligand software (Munson and Robard, 1980). Statistical analyses were performed using two-way analysis of variance with genotype and nimodipine treatment as factors.

In situ hybridization

In situ hybridization was performed as described (Chapter 2), with the following modifications. Control (C57BL/6J Os +/+ tg mice) and tottering mice were chronically treated with nimodipine or vehicle and killed twenty-four hours after the last drug injection. TH mRNA expression in anterior and posterior cerebellum was compared in darkfield views to assess potential regional differences in the nimodipine treatment effect. Analysis of mean grain density was performed using the public domain NIH Image program (National Technical Information

Service, Springfield, Virginia) from emulsion-dipped slides. Darkfield views of anterior and posterior cerebellum were image-captured (using the primary fissure as the boundary between anterior and posterior), inverted, and grain density over the entire anterior or posterior cerebellum counted. Measures were taken from a minimum of four slides from each animal (n =

4-5 animals per treatment group). Data were analyzed using the Student’s t-test. 58

Results

Potassium stimulated calcium uptake in tottering synaptosomes.

Calcium uptake was measured in control and tottering mouse synaptosomes prepared from cerebellum, hippocampus and striatum. Preliminary results were also obtained using synaptosomes prepared from cortex. Two concentrations of potassium buffer (25mM and

60mM) were used to stimulate depolarization-induced calcium uptake, since previous reports 59 have suggested that lower potassium concentrations may preferentially activate L-type calcium channels (Dunn, 1988), while higher potassium concentrations activate P/Q-type and N-type calcium channels. No effect of genotype was observed after stimulation with either 25mM or

60mM potassium in any of the brain regions studied. The L-type calcium channel antagonist, verapamil, antagonized 25mM potassium stimulated calcium uptake in both control and tottering cerebellar synaptosomes by 23% and 35%, respectively (p < 0.05, Figure 3.1A). Calcium uptake 60 was reduced by 28% (control) and 34% (tottering), in synaptosomes exposed to 60mM potassium stimulation buffer (p < 0.01, Figure 3.1B). In hippocampal synaptosomes stimulated with 60mM potassium, calcium uptake was reduced by 29% (control) and 25 % (tottering) in the presence of verapamil (p < 0.05, Figure 3.2B). A trend toward an interaction effect of genotype and drug treatment was observed at the lower potassium concentration, as uptake in control synaptosomes was reduced by 27% while verapamil had no effect on calcium tottering synaptosomes (p = 0.1, Figure 3.2A). Preliminary results with cortical synaptosomes revealed a verapamil-sensitive component in both control and tottering mice, with calcium uptake reduced by 37% in control and 47% in tottering synaptosomes in 25mM potassium stimulation buffer (p 61

< 0.05, Fig 3.3A). Similar results were observed with cortical synaptosomes in 60mM potassium stimulation buffer (p < 0.01, Figure 3.3B). In the striatum, limited tissue availability precluded the addition of verapamil, but no differences in total calcium uptake were apparent between tottering and control mouse synaptosomes (Fig 3.4).

Attack frequency during the course of chronic nimodipine treatment

Because attacks have been linked to L-type calcium influx, it was hypothesized that a change in L-type calcium channel regulation in response to chronic drug treatment would alter the behavioral expression of attacks. Therefore, attacks induced by transporting experimental animals to the laboratory for the morning injections were recorded to measure the effects of 62 chronic nimodipine administration on the frequency of this behavior. At the onset of the experiment, tottering mice assigned to the vehicle (n=9) or nimodipine group (n=10) exhibited similar attack frequencies (vehicle = 78%, nimodipine = 80%). On the fourteenth day of nimodipine administration, no change in attack frequency was observed (vehicle = 88%, nimodipine = 80%).

Effect of chronic nimodipine treatment on L-type calcium channel binding site density

The density of L-type

calcium channels was assessed in the

forebrain and cerebellum of control

and tottering mice using [3H]PN200-

110 radioligand binding. The density

of L-type channel binding sites in

the cerebellum of tottering mice was

significantly greater than control

mice (p < 0.05, Figure 3.5A),

consistent with previous reports

(Campbell and Hess, 1999). In

contrast, L-type calcium channel

binding density was comparable in 63 tottering and control mouse forebrain (Figure 3.5B). Chronic nimodipine administration had no effect on L-type calcium channel binding density in either tottering or control mice. Binding affinities for [3H]PN200-110 did not differ between genotypes (data not shown). Chronic treatment with nimodipine did not affect L-type calcium channel binding affinity in either brain region tested, suggesting that little residual nimodipine remained in the tissue following the drug clearance period.

α Effect of chronic nimodipine administration on 1C mRNA expression

α L-type calcium channel 1C subunit mRNA expression was assessed in brains of control and tottering mice chronically treated with nimodipine. No effects of genotype or drug treatment

α Table 3.1 1C mRNA expression following chronic nimodipine treatment in control and tottering mice

Control Control Tottering Tottering Vehicle Nimodipine Vehicle Nimodipine

Cerebellum1 0.156 ± 0.002 0.138 ± 0.004 0.144 ± 0.003 0.143 ± 0.001

Striatum 0.098 ± 0.001 0.095 ± 0.002 0.098 ± 0.001 0.097 ± 0.001

Thalamus 0.107 ± 0.001 0.101 ± 0.002 0.103 ± 0.002 0.104 ± 0.001

Dentate gyrus2 0.190 ± 0.005 0.170 ± 0.004 0.178 ± 0.007 0.177 ± 0.004

Hippocampus- 0.121 ± 0.002 0.118 ± 0.002 0.117 ± 0.003 0.115 ± 0.002 CA1 Hippocampus- 0.142 ± 0.003 0.134 ± 0.005 0.142 ± 0.009 0.139 ± 0.004 CA3 Olfactory bulb 0.180 ± 0.002 0.169 ± 0.005 0.169 ± 0.004 0.171 ± 0.008

1) interaction effect (p < 0.01) 2) treatment effect (p < 0.05) Data are expressed as mean ± S.E.M in raw optical density (R.O.D.) units. n = 4-5 per group, minimum 3 slides per animal 64 were observed in the olfactory bulb, striatum, CA1 or CA3 regions of the hippocampus.

However, a significant interaction effect was observed in the cerebellum, whereby chronic

α nimodipine treatment reduced 1C mRNA expression in control but not tottering mice (p < 0.01,

Table 3.1). Similar trends were observed in the dentate gyrus and thalamus.

Effect of chronic nimodipine treatment on tyrosine hydroxylase mRNA expression.

The effect of chronic L-type calcium channel blockade on TH mRNA expression was assessed in brains of control and tottering mice. Tottering mice exhibited normal TH mRNA expression in the substantia nigra, locus coeruleus and olfactory bulb (data not shown), consistent with previous reports. Chronic nimodipine treatment had no effect on TH mRNA expression in these catecholaminergic regions in mice of either genotype (data not shown).

Control mice were eliminated from the analysis of cerebellar TH mRNA expression, as the population of TH-positive Purkinje cells was too small to provide an accurate measurement of the drug effect. However, TH mRNA was abundantly expressed in tottering mouse Purkinje cells

(Figure 3.6). Tottering mice chronically exposed to nimodipine expressed significantly less TH mRNA in Purkinje cells than vehicle-treated tottering mice (p < 0.05, Figure 3.7). Further, the cerebellum was divided into anterior and posterior regions with the primary fissure defining the border between regions. TH mRNA was differentially expressed in these regions, with significantly greater expression in the posterior cerebellum than anterior cerebellum (p<0.0001).

After chronic treatment with nimodipine, TH mRNA expression in tottering mouse Purkinje cells was significantly reduced in both anterior and posterior cerebellum (p<0.001) by ~ 60% while 65 the rostral-caudal gradient of the TH mRNA expression was still present after chronic treatment

(Figure 3.8).

67

Discussion

The tottering mutation is predicted to decrease calcium current through P/Q-type calcium channels (Wakamori et al., 1998) while initiating secondary changes in cerebellar L-type calcium channel regulation (Campbell and Hess, 1999). In order to assess the impact of the tottering mutation on P/Q-type calcium channel function and subsequent L-type calcium channel upregulation, calcium uptake was stimulated by chemical depolarization of synaptosomes from various regions of the tottering mouse brain with the addition of an L-type calcium channel antagonist. A similar protocol revealed an approximately 25% decrease in total calcium uptake in cortical synaptosomes from lethargic mice (Lin et al., 1999), due to a different P/Q-type calcium channel subunit gene mutation (Burgess et al., 1997).

The results demonstrate that there is little difference in total calcium uptake between genotypes, as might be expected if L-type calcium channels compensate for the effects of the

P/Q-type calcium channel mutation in the tottering mouse brain. However, no differences in the relative contribution of L-type calcium channels were apparent in synaptosomes depolarized in the presence of the L-type calcium channel antagonist, verapamil. Several methodological considerations may affect the interpretation of the results from the assays performed in this study. First, although it was a necessary initial investigation into calcium channel activity in tottering mice, an assay that measures the magnitude of calcium influx in cerebellar homogenates may not detect meaningful alterations in calcium channels on select cerebellar cell types such as

Purkinje cells. In addition, the use of synaptosome preparations, which are sheared-off nerve 68 terminals rather than a complete cellular system, is optimal for P/Q-type calcium channels but introduces bias against L-type calcium channels expressed at the soma and proximal dendrites.

Although a sizable verapamil-sensitive component was detected, in agreement with previous reports (Turner and Goldin, 1985; Dunn, 1988), synaptosome preparations may not accurately represent total cellular L-type calcium channel activity simply by virtue of their predominant localization in more proximal parts of the neuron (Hell et al., 1993). It is likely that L-type calcium channel upregulation has less impact at the nerve terminal than at the axon hillock, where the excitability of the cell is translated into an all-or-none action potential from the sum of synaptic inputs. Finally, the contributions of other voltage-dependent or ligand-gated calcium channel subtypes (such as N- and R-type calcium channels, NMDA, AMPA and kainate glutamate receptor subtypes) have not been measured in the present study but could provide a compensatory increase in calcium influx to reach normal levels. The calcium uptake experiments described here demonstrate that the tottering mutation does not dramatically alter the capacity for calcium influx in synaptic terminals from multiple brain regions, but do not exclude the possibility that several compensatory changes could account for this result.

L-type calcium channel activity is regulated by subunit interactions (reviewed by

Trimmer, 1998), G proteins (reviewed by Zamponi and Snutch, 1998), phosphorylation state

(Yang and Tsien, 1993), calmodulin binding (Zuhlke et al., 1999) and hormonal interactions

(Pang et al., 1990; Mantegazza et al., 1995). Regulation at the transcriptional level is less well-

α studied, however. Increases in L-type calcium channel 1C subunit mRNA have been found in 69 cardiac myocytes in response to norepinephrine exposure (Maki et al., 1996) or elevated extracellular calcium concentrations (Davidoff et al., 1997), demonstrating that calcium channel expression can be regulated by changes in calcium concentration. The present study documents

α similar effects on 1C mRNA expression in the mammalian central nervous system.

Upregulation of L-type calcium channels in reactive astrocytes has been proposed as a general mechanism to promote the release of neurotrophic factors necessary for neuronal repair, since several brain injuries alter L-type calcium channel expression (Westenbroek et al., 1998).

Hyperexcitability due to elevated extracellular K+ is a common theme among models of brain injury, including kainic acid-induced epilepsy, hypomyelination, forebrain lesion and ischemia

(Westenbroek et al., 1998). In the present study, the converse situation (decreased excitability)

α using chronic treatment with the dihydropyridine nimodipine decreased 1C subunit mRNA expression in control but not tottering mouse cerebellum. This result is consistent with the notion that overall excitability may be increased in the tottering mouse cerebellum, rendering

α nimodipine a less effective regulator of 1C subunit mRNA in these mice. Together, these results suggest that hyperexcitability in the cerebellum could account for differences in L-type calcium channel regulation in tottering mice, but identifying the precise molecular mechanisms requires further study.

The aberrant expression of a calcium-responsive gene, TH, in tottering mouse Purkinje cells is a cellular phenotype that suggests that calcium influx is increased in the tottering mouse 70 cerebellum. TH mRNA expression is responsive to changes in intracellular calcium concentration (Vidal et al., 1989; Brosenitsch et al., 1998; Cigola et al., 1998), as increased calcium influx through L-type calcium channels promotes the expression of TH. L-type calcium channels are mainly localized on the cell soma and proximal dendrites; these channels directly interact with the calcium binding protein calmodulin (Zuhlke et al., 1999), which serves as an important effector molecule in the regulation of calcium-responsive gene transcription. Given the apparent increase in the density of L-type calcium channels in the tottering mouse cerebellum, it seemed plausible that abnormal cerebellar TH expression may be linked to increased calcium influx through L-type calcium channels.

Expression of TH mRNA in tottering mouse Purkinje cells is significantly reduced following L-type calcium channel blockade. The most parsimonious explanation for the nimodipine-induced reduction in TH mRNA expression is blockade of L-type calcium channels directly on Purkinje cells, thereby decreasing the calcium signals regulating TH gene expression.

However, L-type calcium channels are also expressed by cerebellar granule cells (Tanaka et al.,

1995) and cells in the inferior olivary nucleus (Hillman et al., 1991), which both provide powerful excitatory input to Purkinje cells. Blockade of L-type calcium channels on cerebellar granule cells and olivary climbing fibers would decrease excitatory input to Purkinje cells.

Regardless of the site of action, the net effect of nimodipine administration is a decrease in

Purkinje cell excitability, and consequently a decrease in calcium influx. These data support the hypothesis that some tottering mouse phenotypes, such as cerebellar TH expression and motor 71 attacks, are due to excess calcium signaling despite the initial reduction in P/Q-type calcium channel current caused by the mutation.

The data presented in this chapter suggest that the P/Q-type calcium channel defect in tottering mice is associated with changes in the regulation of cerebellar L-type calcium channels.

These alterations include increased binding density of an L-type calcium channel ligand, and an

α impaired ability to regulate 1C subunit mRNA in response to chronic drug challenge. Although

L-type calcium channels do not appear to contribute significantly more calcium in tottering mouse synaptosomes, it remains possible that L-type calcium channels, along with other calcium channel subtypes, may participate in a global compensatory mechanism to increase calcium influx in tottering mouse cerebellar cells. The reduction in the aberrant TH mRNA expressed in tottering mouse Purkinje cells following chronic calcium channel blockade supports the hypothesis that increased calcium influx also plays a role in this phenotype.

The paradox emerging from in vivo analyses in the calcium channel mouse mutant tottering is the juxtaposition of reduced current density through P-type calcium channels

(Wakamori et al., 1998) and several lines of evidence suggesting that increased calcium influx promotes the cerebellar phenotypes. The initial mutation in tottering mouse P/Q-type calcium channels appears to have secondary effects on L-type calcium channel regulation, which ultimately participates in phenotypic abnormalities including motor attacks and ectopic gene expression. Investigations into calcium regulation using these in vivo models may prove valuable in understanding the etiology and progression of neurological disease caused by calcium channel 72 mutations in humans, as well as in understanding basic mechanisms controlling calcium homeostasis in the central nervous system. 73

CHAPTER 4: CHRONIC L-TYPE CALCIUM CHANNEL ACTIVATION IN

TOTTERING MICE

Chapter Summary

The identification of several strains of mice carrying mutations in voltage-dependent calcium channel subunits allows the study of calcium channels in vivo, where interactions with multiple cellular systems may have significant influence on channel (mis)regulation and

α (dys)function. Tottering mice inherit a missense mutation in the 1A subunit of P/Q-type calcium channels, but exhibit several features suggesting that L-type calcium channel misregulation may contribute to the phenotypic abnormalities. To assess effects of the P/Q-type calcium channel tottering mutation on L-type calcium channel regulation in vivo, mice were chronically challenged with BAY K8644, an L-type calcium channel agonist known to provoke behavioral tolerance after repeated drug exposure in normal mice.

Animals were injected with the agonist once per day for a total of five days and scored

α for the degree of motor impairment. L-type calcium channel 1C mRNA expression and potassium-stimulated calcium uptake was also measured in animals repeatedly exposed to BAY

K8644 to determine the effect of chronic activation on calcium channel regulation. Last, TH mRNA expression in tottering mouse Purkinje cells was assessed to determine effects of chronic calcium channel activation on the aberrant expression of this calcium-responsive gene in the tottering mouse cerebellum. 74

Consistent with previous reports, motor behavior in control and tottering mice was affected by BAY K8644 administration; the most prominent feature was prolonged involuntary twisting movements and abnormal postures reminiscent of generalized dyskinesia in humans.

Tottering mice were more severely affected by BAY K8644. After repeated exposure, dyskinesia was significantly reduced in both control and tottering mice, however, the impairment in tottering mice was only slightly ameliorated while control mice developed greater tolerance to the drug. Following chronic administration of BAY K8644, the component of potassium- stimulated calcium influx sensitive to the L-type calcium channel antagonist verapamil was significantly reduced in tottering, but not control cerebellar synaptosomes. It appears that control mice do not develop tolerance via downregulation of L-type calcium channels. Tottering mice were more sensitive to increases in calcium influx, demonstrated by drastically impaired motor behavior and downregulation of L-type calcium channels in response to chronic calcium channel activation. These results suggest that L-type calcium channel regulation is altered in tottering mice, perhaps as a secondary effect of the mutation in the P/Q-type calcium channel. 75

Rationale

Although voltage-dependent calcium channels are responsible for critical neuronal functions, including membrane excitability and neurotransmitter release, calcium channel regulatory mechanisms in the mammalian central nervous system are not well understood. The recent identification of several human neurological disorders (Ophoff et al., 1996; Jodice et al.,

1997; Jen et al., 1999) and mouse phenotypes (Fletcher et al., 1996; Burgess et al., 1997; Doyle et al., 1997; Letts et al., 1998; Mori et al., 2000) caused by calcium channel mutation underscores the importance of understanding the dynamics of calcium channel systems in the intact organism.

Tottering mice inherit a mutation in a gene encoding P/Q-type calcium channels (Fletcher et al., 1996; Doyle et al., 1997), and exhibit attacks of motor impairment (Green and Sidman,

1962) and ectopic gene expression (Hess and Wilson, 1991; Austin et al., 1992; Abbott et al.,

1996) in the cerebellum. However, changes in cerebellar L-type calcium channels may be central to the tottering syndrome. The density of L-type calcium channel binding sites is increased in the tottering mouse cerebellum (Campbell and Hess, 1999), a region required for the expression of motor attacks (Campbell et al., 1999; Abbott et al., 2000). L-type calcium channel blockade reduces the aberrant expression of TH in tottering mouse Purkinje cells (Chapter 2). Further, L- type calcium channel antagonists are capable of preventing motor attacks in tottering mice, and the L-type calcium channel agonist, BAY K8644 can elicit attacks in tottering mice at doses having little to no effect on control mice (Campbell and Hess, 1999). Overall, these studies 76 suggest that, contrary to the effects predicted by the P/Q-type calcium channel mutation, increases in calcium influx promote several of the tottering mouse phenotypes. Surprisingly, calcium uptake studies revealed little change in the contribution of L-type calcium channels to calcium influx in tottering mouse cerebellum (Chapter 3). However, even small changes in L- type calcium channel physiology may be magnified by subsequent calcium signaling to generate abnormal activity in this region, as evidenced by the decrease in cerebellar TH mRNA expression after chronic L-type calcium channel antagonism.

One approach to studying calcium channel regulation and the potential effects of misregulation in vivo is to chronically challenge the calcium channel system with a specific agonist. Acute administration of the L-type calcium channel agonist BAY K8644 elicits a syndrome of behavioral impairment in normal mice, including reduced activity, postural changes and involuntary movements. These effects can be antagonized by dihydropyridine antagonists and are mediated centrally (Bourson, 1989; Shelton, 1987; O'Neill et al., 1990). Repeated administration of BAY K8644 has been shown to evoke behavioral tolerance to subsequent doses of the drug (O'Neill and Bolger, 1988; Nikodijevic et al., 1994; Jinnah et al., 2000), within a few exposures. PC12 cells incubated with BAY K8644 for 12 hours in vitro downregulate binding sites for an L-type calcium channel dihydropyridine ligand (Skattebol et al., 1989 and

Nikodijevic et al., 1994). Thus, there is evidence to suggest that manipulation of the calcium channel system using BAY K8644 may provide insight into L-type calcium channel-responsive phenotypes in tottering mice. 77

Materials and Methods

Animals

Mice were maintained and genotyped as described previously (Chapter 2). All tottering animals used in these experiments were between the ages of 2-6 months, and age- and gender- matched C57BL/6J +/+ mice were used as controls.

Drug Administration

BAY K8644 (Research Biochemicals International, Natick, MA) was prepared at

0.8mg/ml in 14.5% ethanol/ 2.25% Tween80 in 0.9% Saline. The mice were injected subcutaneously with a 10ml/kg volume of BAY K8644 or vehicle once per day. The initial experiment was designed for five days of BAY K8644 injection, however, behavioral observations revealed that maximal tolerance was reached by the fourth day. Therefore, subsequent experiments utilized four days of BAY K8644 treatment.

Behavioral Scores

BAY K8644-induced attacks: Administration of a high dose of BAY K8644 (8mg/kg) causes behavioral abnormalities including dyskinesia in tottering and control mice. To determine the severity of the impairment over successive treatment days, control and tottering mice were scored using a motor disability rating system adapted from (Jinnah et al., 2000) (Table

4.1). Tottering mice were more sensitive to BAY K8644, consistent with previous reports

(Campbell and Hess, 1999), resulting in a longer duration of dyskinesia than observed in control 78 mice. Therefore, mean disability scores were calculated for controls over two hours, and for tottering mice over three hours to better reflect the actual severity of impairment caused by BAY

K8644 administration for each genotype.

Table 4.1 Rating system for motor impairment (Adapted from Jinnah et al., 2000) Disability score Behavioral signs D0 Normal motor behavior D1 No impairment; slightly slowed or abnormal movements D2 Mild impairment; limited ambulation unless disturbed, transient abnormal postures, infrequent falls D3 Moderate impairment; limited ambulation even when disturbed. Frequent abnormal postures, frequent falls but upright for the majority of the time D4 Severe impairment; almost no ambulation, sustained abnormal postures, not upright for the majority of the time D5 Prolonged immobility in abnormal postures

Restraint-induced attacks: If L-type calcium channels were downregulated by chronic

BAY K8644 administration, the frequency or severity of restraint-induced attacks may be reduced in tottering mice. Therefore, twenty-four hours after the final drug exposure, experimental animals were subjected to restraint. This paradigm consists of restraining the mice in plastic 60cc syringes for 10 min followed by release to a novel cage for 30 min. The mice are returned to the home cage and observed for another 10 min. Mice are scored for the presence or absence of attacks every 10 min following release. 79

Calcium uptake assays

Calcium uptake assays were performed as previously described (Chapter 2), with the following modifications. Twenty-four hours after the final BAY K8644 injection, synaptosomes were prepared from the cerebellum and striatum of adult control and tottering mice treated with

BAY K8644 or vehicle. Statistical analyses were performed using two-way repeated measures analysis of variance with genotype and treatment group as factors.

In situ hybridization

The procedure for in situ hybridization was performed as described (Chapter 2), with the following modifications. Mice were deeply anesthetized with carbon dioxide and killed by decapitation twenty-four hours after the last injection of BAY K8644. Subsequently, densitometric data were recorded from a minimum of four slides from each animal (n = 3 animals per treatment group). Data were analyzed using two-way repeated measures analysis of variance with genotype and treatment as factors. An exception was TH mRNA expression in tottering mouse cerebellum; values from vehicle and BAY K8644 treated mice were compared using a paired Student’s t-test. 80

Results

Chronic administration of 8 mg/kg BAY K8644

Control and tottering mice were injected with 8 mg/kg BAY K8644 once per day over five days to assess the behavioral deficits elicited by repeated exposure to this calcium channel agonist.

Both tottering and control mice responded to BAY K8644 administration with behavioral abnormalities, including decreased activity, increased vocalizations, abnormal postures and

involuntary movements, with

occasional self-biting. However,

tottering mice showed more

severe impairment, often

immobile for up to thirty minutes

in a sustained abnormal posture.

While control mice showed

abnormal postures such as

hunched back, Straub tail, and

hindlimb extension, controls

were seldom unable to move,

albeit slowly, if provoked.

Although the latency to the first 81 behavioral sign was similar between genotypes, tottering mice were affected for the entire three hours of behavioral observation, while control mice generally recovered within two hours. With repeated injection with 8 mg/kg BAY K8644, both control and tottering mice developed some tolerance to the drug. However, control mice developed significantly greater tolerance to the effects of the drug than did tottering mice (Figure 4.1 interaction effect, p < 0.05).

Restraint-induced attacks following chronic BAY K8644 exposure

Following the last day of BAY K8644 administration, the number of tottering mice

exhibiting an attack after

restraint was not affected

by repeated BAY K8644

administration (vehicle =

33%, 2/6; BAY K8644 =

40%, 2/5).

Calcium uptake

To determine

physiological consequences

of repeated BAY K8644

administration on voltage-

dependent calcium channels, 82

calcium uptake was measured in

control and tottering mice

following repeated L-type

calcium channel agonist

administration. The cerebellum

and the striatum were chosen for

study since previous work has

implicated these regions in

tottering mouse attacks

(Campbell and Hess, 1999) or

movement disorders in general.

Two concentrations of potassium (25mM and 60 mM) were used to stimulate depolarization-induced calcium uptake to preferentially activate L-type calcium channels or P/Q-type and N-type calcium channels.

In the cerebellum, an interaction effect was observed among the groups for both 25mM

(Figure 4.2, p < 0.01) and 60mM (p < 0.05, data not shown) potassium conditions, such that the verapamil-sensitive component of total calcium uptake was eliminated in the cerebellum of tottering mice treated with BAY K8644. In the striatum, however, regardless of genotype or treatment no significant differences were found among the groups at 25mM potassium 83 concentration (Figure 4.3), however, verapamil significantly antagonized 60 mM potassium- stimulated calcium uptake (p<0.05, data not shown).

α L-type calcium channel 1C mRNA expression following chronic BAY K8644 exposure

α The mRNA expression profile of the L-type calcium channel subunit 1C was investigated in the cerebellum of animals repeatedly exposed to BAY K8644. No statistically significant differences in

α cerebellar 1C mRNA expression were observed among the groups (Table 4.2).

α Table 4.2. Cerebellar L-type calcium channel 1C subunit mRNA expression following chronic BAY K8644 administration VEHICLE BAY K8644 (R.O.D.) (R.O.D.) Control 0.370 ± 0.072 0.412 ± 0.108 Tottering 0.391 ± 0.083 0.396 ± 0.088 Data represent mean ± S.E.M. raw optical density (R.O.D.) units taken from a minimum of four slides per animal (n=3 animals/group). No significant differences were observed among the groups.

Tyrosine hydroxylase mRNA expression in tottering mouse Purkinje cells

TH mRNA expression in tottering mouse Purkinje cells was assessed to determine if chronic BAY K8644 administration and subsequent changes in L-type calcium channels affected this cerebellar phenotype in tottering mice. No statistically significant differences in cerebellar

ΤΗ mRNA expression was observed between the groups (vehicle-treated tottering = 0.250 ±

0.021 ROD; BAY K8644-treated tottering = 0.219 ± 0.026 ROD, p = 0.07). 84

Discussion

Repeated administration of BAY K8644 produced a reduction in the severity of dyskinesia in tottering and control mice by the final exposure to BAY K8644. Both tottering and control mice initially respond to 8 mg/kg BAY K8644 administration with abnormal body postures and involuntary jaw and limb movements. Tottering mice were much more sensitive to the activating properties of the drug, even to the point of mortality, as 25% (2/8) of tottering mice injected with 8 mg/kg BAY K8644 died from tonic-clonic seizures while no deaths occurred in the group of control mice. Tottering mice were unresponsive to external stimuli for long periods but control mice could be provoked to move even when most severely affected.

These observations are consistent with a previous study demonstrating that tottering mice exhibit dyskinesia in response to doses of BAY K8644 that do not affect control mice (Campbell and

Hess, 1999). By the final day of BAY K8644 injection, control mice showed reduced locomotion, but abnormal postures and involuntary movements were largely eliminated. During the same timeframe, tottering mice were still disabled by involuntary movements, although less severely than at earlier timepoints. Overall, control mice, initially less sensitive to the behavioral effects of BAY K8644, developed tolerance to the drug while severe impairments in tottering mice were only slightly ameliorated after multiple days of drug exposure. Further, tottering mice chronically injected with BAY K8644 showed a dramatic reduction in calcium influx through L- type calcium channels, suggesting that L-type calcium channels are abnormally regulated in response to repeated agonist administration. Collectively, these data indicate that tottering mice 85 are highly vulnerable to effects of calcium influx, possibly a consequence of impaired calcium channel regulation in response to the initial P/Q-type calcium channel mutation.

Potassium-stimulated calcium uptake assays revealed downregulation in tottering mouse cerebellar L-type calcium channels in response to repeated BAY K8644 administration. Control mice showed no change in the verapamil-sensitive component of calcium uptake after BAY

K8644 treatment in the cerebellum or striatum, suggesting little change in the activity of L-type calcium channels in these regions in normal mice. Repeated BAY K8644 injection had no effect on L-type calcium channel-mediated calcium influx in the tottering mouse striatum. This result supports the notion that the cerebellum is a primary site of abnormal calcium channel regulation in tottering mice, consistent with studies documenting the central role of the cerebellum in tottering mouse dyskinesia (Campbell and Hess, 1998; Campbell et al., 1999; Abbott et al.,

2000). Since control mice developed tolerance after repeated exposure to BAY K8644, it appears that changes in calcium channel activity are potentially less important in the development of

BAY K8644 tolerance in control mice than secondary changes, such as postsynaptic receptor downregulation. This conclusion is consistent with previous studies performed in vivo (O'Neill and Bolger, 1988). Tottering mice, however, appear to differentially regulate L-type calcium channels following repeated BAY K8644 exposure to L-type calcium channel activation.

Radioligand binding studies will help to differentiate whether this result is due to L-type calcium channel insensitivity or to downregulation of L-type calcium channel binding sites. 86

Prolonged BAY K8644 administration to PC12 cells in vitro results in dramatic reductions in the subsequent ability to stimulate calcium uptake and norepinephrine release with

BAY K8644. Dihydropyridine binding density decreases in these cells, with no change in potassium-stimulated calcium uptake or in the ability to stimulate calcium-responsive c-fos gene expression, suggesting that L-type calcium channels remain functionally intact and that secondary mechanisms lead to BAY K8644 insensitivity (Nikodijevic et al., 1994). Studies in vivo have demonstrated behavioral tolerance to the motor impairing effects with no change in L- type calcium channel ligand binding (O'Neill and Bolger, 1988). While the results from control mice used in the present study revealed little change in L-type calcium channels after BAY

K8644 administration, the decrease in L-type calcium channel-mediated calcium influx in tottering mice appears to be the first demonstration of chronic BAY K8644 effects on L-type calcium channel regulation in vivo. Overall, these observations suggest that tottering mice are a novel and valuable tool to study calcium channel regulation in vivo, as the P/Q-type calcium channel mutation appears to differentially affect mechanisms of calcium channel regulation.

The abnormal expression of TH mRNA in tottering mouse Purkinje cells is responsive to chronic L-type calcium channel blockade (Chapter 3), thus repeated calcium channel activation might be expected to increase TH mRNA expression in this region. In the present study, the results demonstrate that TH mRNA expression in tottering mouse Purkinje cells was unchanged by repeated BAY K8644 injection. Although changes in gene transcription are rapid and the half-life of TH mRNA is approximately thirty hours (Tank et al., 1986), the abundant TH mRNA 87 expression in tottering mouse Purkinje cells may preclude seeing slight changes within short time periods. Additionally, cerebellar TH mRNA expression may already be maximal in tottering mouse Purkinje cells so that any further increase in calcium influx is transcriptionally ineffectual.

On the other hand, cerebellar L-type calcium channel activity was reduced in tottering mice chronically injected with BAY K8644; TH mRNA expression may actually decrease in tottering mouse Purkinje cells if L-type calcium channels are downregulated following repeated BAY

K8644 administration. Perhaps a longer regimen of BAY K8644 administration would produce greater decreases in TH mRNA expression in this region. The previous study using chronic nimodipine blockade of L-type calcium channels was conducted over two weeks, while the present study spanned five days. Last, it is possible that changes in TH mRNA expression included both an increase during the first days of BAY K8644 injection along with a decrease during the last days, generating little net change in the TH mRNA detected in cerebellar Purkinje cells.

The present study with the L-type calcium channel specific agonist, BAY K8644, suggests that L-type calcium channel activity is perturbed in tottering mutant mice. Tottering mice are more sensitive to the motor impairing effects of the agonist, and less able to develop tolerance to subsequent drug exposures. Although control mice do not appear to require L-type calcium channel downregulation to develop behavioral tolerance, tottering mice exhibit a large reduction in L-type calcium channel-mediated calcium influx in the cerebellum. These results suggest that disruptions to the fragile balance of calcium homeostasis have greater impact on 88 tottering mice. In addition to directly altering PQ-type calcium channel kinetics, the tottering mutation also appears to disrupt the regulation of the calcium channel system in discrete regions of the mouse central nervous system. 89

CHAPTER 5: TRIGGERS AND POTENTIAL THERAPEUTICS IN TOTTERING

MOUSE ATTACKS

Chapter Summary

Mutations in ion channels can lead to disorders in which symptoms are episodically expressed, such as migraine, weakness and ataxia. Though the clinical features in these disorders vary widely, attacks are often precipitated by similar stimuli, including stress, caffeine, alcohol, exercise or fatigue, although an understanding of the triggering mechanisms are lacking.

The tottering mutant mouse exhibits attacks of dyskinesia due to a mutation in the gene encoding

α the 1 subunit of P/Q-type calcium channels. Tottering mice were challenged with various environmental stimuli to assess whether increased arousal and stress precipitate attacks.

Additionally, tottering mice were administered caffeine (15 mg/kg) or ethanol (1.5 g/kg) to determine their response to agents known to trigger episodic disorders in humans. Stress, caffeine and alcohol all reliably induced attacks in tottering mice. Caffeine- and ethanol-induced attacks were blocked by the L-type calcium channel antagonist, nimodipine. In addition, the

NMDA receptor antagonist, MK-801 blocked restraint-induced attacks in a dose-dependent manner. Anticonvulsants including ethosuximide, phenytoin, and carbamazepine were ineffective against tottering mouse attacks.

As stress, caffeine and ethanol all activate the hypothalamic-pituitary-adrenal axis, it was hypothesized that stress hormones, such as corticotropin-releasing hormone (CRF) and corticosterone (CORT), play a role in generating attacks in tottering mice. Tottering mice were 90 challenged with restraint or caffeine after surgical removal of the adrenal glands. Adrenalectomy failed to prevent attacks, suggesting that the adrenal stress hormone corticosterone is not a requisite factor. Consistent with this interpretation, the synthetic glucocorticoid dexamethasone failed to induce attacks in tottering mice. Additionally, a specific CRF type 1 receptor antagonist, antalarmin, was ineffective against restraint-induced attacks, suggesting that CRF-1 receptors are are unlikely to control the expression of stress-induced attacks. Overall, the results suggest that conditions that increase calcium influx and heighten neuronal excitability, caused by multiple factors, can precipitate attacks in tottering mice. This study demonstrates that the tottering mouse is an excellent model to investigate triggers of episodic neural dysfunction arising from ion channelopathies. 91

Rationale

Several neurological conditions share a unique feature; episodic expression of neuronal dysfunction superimposed on a normal baseline. Such disorders include paroxysmal dyskinesia, periodic paralysis, episodic ataxia, epilepsy, hemiplegic migraine, and the common migraine, among others (Demirkiran and Jankovic, 1995; Ptacek, 1997; Gardner and Hoffman, 1998;

Ptacek, 1998; Bhatia, 1999; Cooper and Jan, 1999; Jen, 1999; Ptacek, 1999). Since many of these disorders have been linked to genes encoding ion channels, it has been suggested that the episodic nature of impairment in such disorders results from transient dysfunction of ion channels. Channelopathies as a group share genetic etiology and transient expression of neurological dysfunction. Once considered clinically distinct disorders, more recently it has been realized that co-occurrence of various features, such as migraine, epilepsy and movement disorders, can exist in subsets of channelopathy patients (Neville et al., 1998; Singh et al.,

1999b). It has been proposed that understanding the pathogenesis of attacks in these rare but genetically simple channelopathies may provide the groundwork for unraveling more complex episodic disorders, including migraine headache and epilepsy (Ptacek, 1998).

Though the symptoms of different channelopathies are quite diverse, they share common triggers. Psychological or emotional stress, fatigue, exercise, caffeine, alcohol, or hormonal fluctuations are commonly noted precipitants (Mount and Reback, 1940; Richards and Barnett,

1968; Lance, 1977; Boel and Casaer, 1988; Bressman et al., 1988; Ptacek, 1998; Battistini et al.,

1999; Cooper and Jan, 1999; Jen, 1999; Ptacek, 1999; Jarman et al., 2000; Pittock et al., 2000). 92

Unfortunately, there is little understanding of the mechanisms by which these triggers precipitate symptoms in a patient who is normal between attacks. Unraveling the triggering mechanisms may be a key step in understanding the pathogenesis of attacks and in developing effective prevention strategies.

An animal model of paroxysmal neuronal dysfunction could provide a powerful tool to examine the cellular and neuroanatomical bases of episodic neurological disorders, and may prove helpful in testing candidate drug therapies. The tottering mutant mouse displays an episodic movement disorder similar to paroxysmal dyskinesia (Campbell and Hess, 1999;

Campbell et al., 1999) in addition to absence epilepsy (Kaplan et al., 1979; Noebels and Sidman,

1979). Like many human episodic disorders, the tottering syndrome is a channelopathy, resulting

α from a mutation within the gene encoding the 1 subunit of P/Q-type calcium channels (Fletcher et al., 1996). Attacks of dyskinesia impair motor function in tottering mice approximately once or twice per day; between attacks the mice are mildly ataxic but otherwise exhibit normal motor behaviors. Attacks typically last from 20-40 minutes and can occur in succession with little to no refractory period. Attacks are differentiated from seizures in that the mice do not lose consciousness, and the EEG appears normal during an attack (Kaplan et al., 1979; Noebels,

1984). Because the tottering mouse may be an important model for understanding episodic disorders, the behavioral profile of episodic dyskinesia in response to various stimuli reported to induce attacks in human channelopathy patients was studied. Further, given the convergence of 93 several trigger factors in activating a hormonal stress response, investigations into the role of stress hormones in triggering tottering mouse attacks were conducted. 94

Materials and Methods

Animals

Tottering mice were identified and maintained as previously described (Chapter 2). All animals used in these experiments were between the ages of 2-6 months.

Drug Administration

Drugs and components of vehicle solutions were obtained from Sigma RBI (Natick,

MA). Most drugs were dissolved in 0.9% saline. Exceptions were phenytoin and carbamazepine, which were dissolved in dimethylsulfoxide (DMSO), then diluted to 50%

DMSO with 0.9% saline, and nimodipine, which was dissolved in a small volume of 100% ethanol and diluted with 5% ethanol, 2.5% Tween 80, 0.9% saline. Dexamethasone was dissolved in 2-hydroxypropyl-β-cyclodextrin at a concentration of 0.1 mg/ml. Antalarmin was a kind gift from Dr. G. Chrousos (NIH, Bethesda MD) and was dissolved in 1:1 v:v 100%

ethanol/Emulphor solution and diluted with dH2O to 2 mg/ml or 4 mg/ml. All injections were given subcutaneously in a volume of 10 ml/kg, except ethanol (10 ml/kg, intraperitoneal).

Triggers

Environmental stimuli: Attack frequency and severity was recorded after subjecting tottering mice to various changes in their environment, in view of previous reports that anecdotally described increased attacks after environmental disturbance (Kaplan et al., 1979;

Syapin, 1983a). To determine a baseline attack frequency at 2 PM (seven hours into the light cycle) under quiet conditions in the vivarium, tottering animals were observed in their home 95 cages for 40 min and scored every 10 min for the presence or absence of attacks. Attacks induced by a change in the light cycle were recorded similarly from tottering mice observed in the vivarium in their home cages over 40 min immediately following the onset of the light cycle

(7AM). The effect of changing the home cage and transporting the animals to the laboratory was recorded during the 40 min immediately following transport. After a minimum of two hours of acclimation, the mice were placed in novel cylindrical cages measuring approximately 4 inches in diameter and 5 inches in length, with wire mesh walls and wooden ends, for 10 min. The mice were scored for the presence of an attack in the novel cage and during the 40 min following release. After repeatedly exposing the mice to the cylindrical cages once per day for four days, the cages were set in motion at a speed of 2 RPM for 10 min to assess the effect of exercise on the frequency of attacks in tottering mice. To examine the effect of short-term restraint on tottering mouse attacks, mice were restrained as previously described (Chapter 4).

Caffeine, alcohol, and dexamethasone challenge: For drug testing, mice were placed in clean cages, transported to the laboratory, weighed and then acclimated to the laboratory for a minimum of two hours prior to the experiment. The mice were scored over 40 min for the presence of an attack following the injection of vehicle, caffeine, alcohol or dexamethasone.

Potential therapeutics

To determine if drugs effective against some human paroxysmal disorders might be effective against tottering attacks, animals were pre-treated with ethosuximide, phenytoin, carbamazepine, nimodipine, MK-801 or antalarmin and then exposed to a trigger. Mice were 96 injected with the drug and observed for 30 min to allow time for the drug to reach the brain.

After 30 min, the mice were subjected to various conditions known to trigger attacks. However, if the drug itself was observed to cause attacks, the animals were not exposed to the trigger.

Surgical adrenalectomy

The bilateral adrenalectomy procedure was performed on mice anesthetized using pentobarbital sodium (Nembutol, 70 mg/kg, Abbott Laboratories, Chicago, IL) prepared in 10% ethanol. A small incision was made in the overlying muscle tissue to expose the kidney; adrenal glands were removed and verified by visual inspection. Incisions were repaired using sutures and surgical staples. The entire procedure minus the removal of the adrenal glands was performed on sham-adrenalectomized animals. Subsequently all animals were given one injection of 2 ml 0.9% saline once per day for three days after surgery and were maintained on

0.9% saline drinking water.

Statistical analysis

During the environmental observations, mice were scored once every 10 min for 40 min for the presence or absence of an attack following the environmental disturbance. Each mouse was assigned a total score, reflecting the number of times it received a positive score during the observation period (maximum score = 4). These data were analyzed using the Mann-Whitney U- test to compare attacks following an environmental trigger with attacks at the baseline control conditions. Similar methods were used to compare the effect of caffeine, alcohol and dexamethasone on tottering mouse attacks. For drug experiments, total scores among several 97 doses were compared using Kruskal-Wallis nonparametric test for multiple independent groups.

Statistically significant differences (α = 0.05) among groups are noted in the figure legend. For ease of presentation, data are shown as percent of tottering mice exhibiting an attack. 98

Results

Attacks

Attacks in tottering mice

are often preceded by increased

activity. The onset of an attack is

signaled by contraction of the

hindlimbs, causing a waddling gait

but little change in mobility. As

the attack progresses, the mouse

loses mobility and begins to slowly

extend and contract the hindlimbs.

Gradually the upper limbs, head

Table 5.1 Effect of caffeine on tottering attacks. and neck exhibit severe flexion in Treatment # Attacks Total % Attacks addition to paddling movements of Vehicle 0 9 0 both hindlimbs. As the lower body 15 mg/kg Caffeine 9 9 100 recovers, the mouse assumes a Tottering mice were injected with caffeine (15 mg/kg, s.c.) or saline. Data are expressed as the percentage of mice tripod position on the hindlimbs exhibiting an attack within 40 min of exposure to the trigger factor. Caffeine triggered attacks in tottering mice (p < 0.0001). and tail while the upper body is

involved in extension/contraction 99

Table 5.2 Effect of ethanol on tottering attacks. movements, with frequent falls. These Treatment # Attacks Total % Attacks episodes occur over 20-40 minutes Vehicle 3 9 33 and are best characterized as 1.5 g/kg Ethanol 8 9 89 paroxysmal dyskinesia. Tottering mice were injected with ethanol (1.5 g/kg, i.p.) or saline. Data are expressed as the percentage of mice Environmental stimuli exhibiting an attack within 40 min of exposure to the trigger factor. Ethanol triggered attacks in tottering mice (p < 0.01). None of the tottering animals

had an attack during the undisturbed afternoon observation period (Figure 5.1, n=38). However,

51.4% (19/37) of the tottering animals had an attack during the 40 min after the onset of the light

cycle in the morning. Following transport to the laboratory, 90% (18/20) of tottering animals

had an attack. After acclimation to the laboratory, 50% (10 /20) tottering mice had an attack in

response to exposure to the novel environment (cylindrical cages) for 10 min. Restraining

tottering mice for 10 min precipitated attacks in 85% (23/27) of the mice, consistent with

previous reports (Campbell and Hess, 1998, 1999; Campbell et al., 1999). All environmental

disturbances significantly triggered attacks (p<0.0001). 100

Exercise: During four days of

acclimation to the cylindrical cage,

tottering mice exhibited attack

frequencies ranging from 55-85%,

suggesting that the mice did not

habituate to the procedure during

this time. On the fifth day, the

same tottering mice were

subjected to 10 min of exercise as

the cylindrical cages were set in

motion and the rotation of the cage

forced the mice to maintain a slow

but steady pace. Attacks were observed in 60% (12/20) of the mice following exercise, a frequency not significantly different from that observed when mice were simply exposed to the motionless cage for 10 minutes (Fig

5.1).

Chemical triggers

Caffeine (15 mg/kg) caused 100% (9/9) of tottering mice to exhibit an attack, while none

(0/9) of the vehicle-injected mice had an attack within the observation period (Table 5.1, p

<0.0001). A dose of 1.5 g/kg ethanol also caused 89% (8/9) of tottering mice to exhibit an attack, 101 compared to 33% (3/9) of vehicle-injected tottering mice (Table 5.2, p <0.01). The attacks induced by caffeine and ethanol were indistinguishable from those induced by other methods.

Potential therapeutics

Table 5.3 Effect of nimodipine pre-treatment on caffeine-induced Tottering mice tottering attacks. Treatment # Attacks Total % Attacks were pre-treated with

Vehicle + Caffeine 6 7 86 drugs known to be

20 mg/kg 2922effective against certain Nimodipine + Caffeine episodic disorders to

Tottering mice were pre-treated with nimodipine (20 mg/kg) or vehicle 30 min prior to caffeine injection (15 mg/kg). Data are determine if they would expressed as the total number and percentage of mice exhibiting an attack within 40 min of exposure to the trigger. Nimodipine pre- prevent attacks. treatment significantly prevented attacks triggered by caffeine (p < 0.01). Ethosuximide and

phenytoin tended to Table 5.4 Effect of nimodipine pre-treatment on ethanol-induced tottering attacks. increase the frequency Treatment # Attacks Total % Attacks of attacks (Fig 5.2), Vehicle + Ethanol 5 5 100 while attacks were 20 mg/kg 080 Nimodipine + significantly induced by Ethanol carbamazepine (Fig 5.2, Tottering mice were pre-treated with nimodipine (20 mg/kg) or saline 30 min prior to ethanol injection (1.5 g/kg). Data are expressed as p < 0.01). Therefore, the total number and percentage of mice exhibiting an attack within 40 min of exposure to the trigger. Nimodipine pre-treatment significantly prevented attacks triggered by ethanol (p < 0.001). mice were not exposed 102 to a subsequent trigger factor. L-type calcium channel blockers, including the dihydropyridine nimodipine, are effective at blocking tottering mouse attacks triggered by restraint (Campbell and Hess, 1999). Nimodipine also blocked attacks provoked by caffeine (2/9, Table 3, p < 0.004) and ethanol (0/8, Table 4, p < 0.0002).

Effect of MK-801 on restraint-induced attacks

Tottering mice were injected with the non-competitive NMDA receptor antagonist, MK-

801 to determine effects of ligand-gated calcium channel blockade on tottering mouse attacks. 103

MK-801 significantly prevented attacks induced by restraint (Figure 5.3, p<0.05). At the 0.1

mg/kg dose, the mice displayed increased ataxia and hyperactivity that became more severe at

0.2 mg/kg MK-801, however, these mice did not progress into typical tottering attacks.

Effect of adrenalectomy on attacks in tottering mice

Overall, evidence from the primary literature and our initial experiments suggested that

activation of the HPA axis may be a common mechanism among attack triggers in tottering

mice. Therefore, the adrenal Table 5.5 Efficacy of triggers following surgical adrenalectomy in tottering mice glands were surgically removed Gender Treatment Transport Restraint Caffeine Female Sham - - + from tottering mice (n=6) to Male Sham + + + Male Sham + + + determine if adrenal hormones are

Female ADX + - + required for the expression of Male ADX - - + Male ADX - - - Male ADX - + + attacks. In all but one case, attacks Male ADX - + + Male ADX + - + could still be elicited by at least

one trigger after adrenalectomy Table 5.6 Effect of exogenous glucocorticoid on the frequency of tottering attacks (Table 5.5). Treatment # Attacks Total % Attacks

Vehicle 2 8 25 Dexamethasone administration 1 mg/kg 1813 The synthetic glucocorticoid Dexamethasone

Tottering mice were injected with dexamethasone (1 mg/kg) or dexamethasone was administered to vehicle. Data are expressed as the total number and percentage of mice exhibiting an attack within 40 min of exposure to the potential trigger. Dexamethasone did not induce attacks. 104 intact tottering mice to determine if increased corticosterone levels could initiate attacks in the absence of another trigger. Exogenous administration of dexamethasone, did not increase the frequency of attacks in intact tottering mice (Table 5.6).

Effect of CRF-1 blockade on restraint-induced attacks

Tottering mice were injected with the CRF-1 receptor antagonist, antalarmin, to determine if this receptor subtype is involved in mediating tottering mouse attacks induced by restraint.

Antalarmin failed to prevent attacks induced by restraint in tottering mice (Table 5.7).

Table 5.7 Effect of CRF-1 receptor blockade on the frequency of tottering attacks Treatment # Attacks Total % Attacks

Vehicle 8 8 100

20 mg/kg Antalarmin 6 8 75

40 mg/kg Antalarmin 6 8 75

Tottering mice were pre-treated with antalarmin or vehicle 30 min prior to restraint. Antalarmin did not significantly induce attacks. Data are expressed as the total number and percentage of mice exhibiting an attack within 40 min of exposure to the trigger. Antalarmin pre-treatment had no effect on restraint-induced attacks. 105

Discussion

Many human disorders in which patients exhibit episodic attacks between periods of relatively normal behavior have been linked to mutations in genes encoding ion channels

(Ptacek, 1998). Interestingly, the features of these seemingly diverse disorders are often triggered by similar phenomena, such as psychological or emotional stress, exercise, fatigue, caffeine, alcohol, and hormonal cycles in women. Although these stimuli are well-known triggers, there is little understanding of the mechanism by which such agents precipitate attacks.

A newly emerging concept is that despite apparent differences in the features of individual disorders, channelopathies as a group may share a common disease mechanism, particularly with regard to precipitating factors. The present results show that the tottering mouse provides an excellent model of episodic symptomology with which to study the action of trigger factors on perturbed ion channel systems.

Several of the most common stimuli reported to trigger attacks in human episodic disorders, specifically caffeine, alcohol and stress, also reliably induce attacks in tottering mice.

The biological actions of the precipitants themselves are currently the only clues to the events underlying the onset of an attack. Caffeine, alcohol and stress exert a wide variety of effects on the central nervous system, ranging from behavioral deficits in fine motor control with acute alcohol ingestion to changes in gene expression mediated by stress hormones. Additionally, each of these trigger factors has direct effects on neuronal excitability. At physiologic levels, caffeine

acts as an antagonist at A1 and A2A adenosine receptors to block the inhibitory effects of 106 adenosine (reviewed by (Fredholm et al., 1999). Other actions of caffeine, particularly the

release from intracellular calcium stores via ryanodine receptors, blockade of GABAA receptors, and the inhibition of cyclic-nucleotide phosphodiesterases, are unlikely to occur at the doses attained during normal human consumption. Caffeine exerts its stimulatory effects via inhibitory

G-coupled A1 receptors, present in almost all brain regions, and stimulatory G-protein coupled

A2A receptors, found within the striatum, nucleus accumbens and olfactory tubercle. In contrast, acute intoxicating levels of alcohol depress synaptic transmission, likely by reducing excitatory inputs and enhancing inhibitory inputs in selected ion channel subtypes (Nestoros, 1980;

Lovinger et al., 1990; Lewohl et al., 1999). The disparity in the acute effects of physiological doses of caffeine and alcohol suggest that these substances should have opposite effects on the initiation of symptomatic attacks in channelopathy patients. However, there is a common biological mechanism shared by caffeine and alcohol. Despite their opposing action on direct measures of neuronal excitability, administration of both caffeine and alcohol generates a neuroendocrine response analogous to a stress-like activation of the hypothalamic-pituitary- adrenal (HPA) axis (Ellis, 1966; Rivier et al., 1984; Pollard, 1988; Nicholson, 1989).

HPA activation in response to a stressor is mediated by the release of a 41-amino acid peptide, corticotropin-releasing factor (CRF)(Vale et al., 1981), from the median eminence of the hypothalamus. Although CRF receptors are distributed throughout the brain (De Souza et al.,

1984; Wynn et al., 1984; De Souza et al., 1985), the second phase of HPA axis activation specifically requires release of CRF to corticotrophic cells in the anterior pituitary. These 107 pituitary cells in turn secrete adrenocorticotropic hormone (ACTH) to the bloodstream. ACTH activates the release of glucocorticoid (GC) hormones from the adrenal glands; corticosterone (in rodents) and cortisol (in humans) act via negative feedback to regulate the release of CRF and

ACTH. Acute or chronic restraint of rats and mice is a typical experimental paradigm for studying stress, as rodents demonstrate increased activation of the HPA axis after restraint, even for short time periods (Barlow et al., 1975; Armario and Castellanos, 1984; Hauger et al., 1988;

Berridge and Dunn, 1989). Hauger and colleagues have demonstrated that in the restrained rat,

ACTH levels increase approximately 20-fold within 5 min, while corticosterone is increased 10- fold during the same time period (Hauger et al., 1988). Both hormones reach peak levels within

15 minutes after the animals are restrained. Similarly, rats given 30 mg/kg caffeine exhibit elevated corticosterone levels within 3 minutes after injection (Pollard, 1988); rats given a single dose of 1 g/kg EtOH show peak levels of corticosterone within 15 minutes, reflecting the central release of CRF (Rivier et al., 1984). This convergence suggests that a hormonal trigger may play a role in the final common pathway to attack initiation.

The present results suggest that tottering mouse attacks can be initiated in the absence of the adrenal stress hormone, corticosterone. Supporting this conclusion is the fact that exogenous administration of the synthetic glucocorticoid, dexamethasone, did not trigger attacks in intact mice. However, since corticosterone is an end product of HPA axis activation, these results do not eliminate potential effects of the central stress hormone, corticotropin-releasing hormone.

CRF itself has activating properties in other regions of the central nervous system, mediated 108 through G-protein coupled CRF receptor types 1 (Perrin et al., 1993) and 2 (Lovenberg et al.,

1995). Blockade of CRF-1 receptors using specific antagonists like antalarmin (Webster et al.,

1996), reduces behavioral effects of stress such as increased locomotion and suppression of feeding behavior (reviewed by (Koob, 1999). However, at a similar dose antalarmin had no effect on restraint-induced attacks in tottering mice. These results suggest that CRF-1 receptors are not required for the initiation of attacks, but do not rule out effects mediated by CRF-2 receptors. Further study is required to elucidate the role of CRF in stress-induced tottering mouse attacks.

In this study, treatment with the anticonvulsants ethosuximide, phenytoin and carbamazepine offered no therapeutic value in preventing attacks in tottering mice. In fact, at higher doses carbamazepine actually promoted attacks, while phenytoin and ethosuximide exhibited similar trends. These results support the idea that attacks of dyskinesia in tottering mice truly represent a movement disorder not related to motor seizures. Secondly, these results demonstrate that the pharmacological profile of attacks in tottering mice is similar to that of human paroxysmal non-kinesiogenic dyskinesia, the subtype of paroxysmal dyskinesia that the tottering mouse phenotype most closely resembles. Patients with this disorder are often unresponsive to typical anticonvulsant therapies (Demirkiran and Jankovic, 1995).

In contrast, the L-type calcium channel antagonist, nimodipine, was extremely effective in preventing attacks precipitated by caffeine or alcohol administration. The L-type calcium channel has previously been implicated in the pathophysiology of attacks in tottering mice, since 109 administration of L-type calcium channel antagonists inhibits restraint-induced attacks in a dose- dependent manner (Campbell and Hess, 1999). The noncompetitive NMDA receptor antagonist

MK-801 was also effective at preventing restraint-induced attacks. NMDA receptors are ligand- gated calcium channels responsive to glutamate; these receptors, along with other ionotropic glutamate receptors such as the AMPA and delta subtypes, are postsynaptically expressed at excitatory synapses in the cerebellum (Takayama et al., 1996; Kakizawa et al., 2000). The inhibition of attacks with MK-801 demonstrates that decreasing neuronal excitability and calcium influx through a second calcium channel type can also prevent attacks. The data shown here demonstrate that restraint, caffeine and alcohol initiate attacks via mechanisms dependent on altered neuronal excitability and increased calcium influx, further supporting the idea that a final common pathway exists between diverse triggers in this model of episodic neurological dysfunction.

Although the clinical presentation has historically divided most channelopathies into movement disorders, migraine headache disorders, or muscle disorders, more recently the clinical boundaries between different episodic disorders has been called into question. There is a growing awareness that co-occurrence rates of epilepsy, migraine, periodic paralyses and paroxysmal movement disorders are somewhat higher than might be expected due to chance

(Demirkiran and Jankovic, 1995; Hofele et al., 1997; Gardner and Hoffman, 1998; Ptacek, 1998,

1999; Singh et al., 1999b). Susceptibility to multiple types of episodic disorders appears to be true of tottering mice as well. The tottering mouse is widely studied as a model of absence 110 epilepsy, owing to periodic polyspike bursts in the EEG accompanied by behavioral arrest

(Kaplan et al., 1979; Noebels and Sidman, 1979). Although absence seizures and paroxysmal dyskinesia in tottering mice are most often regarded as independent phenotypes, the co- occurance of epilepsy and a paroxysmal movement disorder in mice is remarkably consistent with recent reports of subsets of channelopathy patients exhibiting more than one type of episodic disorder (Demirkiran and Jankovic, 1995; Neville et al., 1998; Singh et al., 1999a).

Overall, the similarities in genetic etiology, precipitating factors, and overlap in symptomology in both mouse and man suggest that ion channelopathies may present a heterogeneous group of disorders with a common underlying triggering mechanism. Tottering mice, arising from a mutation in a neuronal calcium channel, are one of few genetic animal models that recapitulate the transient nature of episodic disorders with symptoms triggered by stress, caffeine and alcohol. As such, tottering mice offer a new avenue of research into the link between ion channelopathies and episodic disorders. 111

CHAPTER 6: CONCLUSIONS

Mice with mutations in key cellular proteins have traditionally been used to gain insight into protein function via a loss-of-function approach. Behavioral deficits in mutant mice can result directly from the absence or abnormal function of a particular gene product. Alternatively, mutant mice generated either by genetic manipulation or spontaneous mutation may show few overt phenotypes because of compensation for the targeted protein by redundant members of the protein family. This does not mean that the mutation is without effect, as compensatory mechanisms can seldom fully replicate the original function of the mutated or missing protein.

Thus, the possibility that compensatory changes contribute to behavioral and cellular abnormalities should not be overlooked in investigating mutant mouse phenotypes or human genetic disorders.

This caveat may be especially applicable to mutations that have the potential to affect highly conserved processes, such as synaptic transmission and neuronal excitability. The tottering mutation in P/Q-type voltage dependent calcium channels falls into this category. The central paradox emerging from the study of these mice is that while the tottering mutation reduces calcium influx through P/Q-type calcium channels (Wakamori et al., 1998), increased calcium influx has been associated with both behavioral (motor attacks) and cellular (cerebellar

TH expression) tottering mouse phenotypes. The identification of compensatory changes in the central nervous system of tottering mice may be one way this paradox is resolved. As human episodic disorders resulting from ion channel mutations exhibit a number of paradoxical features 112 similar to the tottering syndrome, the results of studies using tottering mice may have broad applicability with clinically predictive value.

The rationale for the studies reported here was based on evidence demonstrating central roles for abnormal cerebellar activity and L-type calcium channel misregulation in tottering mouse phenotypes. Following the onset of a motor attack in tottering mice, polysynaptic activation reflected by c-fos immediate early gene expression first appears in the cerebellum

(Campbell and Hess, 1998). In addition, eliminating output from the cerebellar cortex with a genetic lesion of Purkinje cells abolishes motor attacks (Campbell et al., 1999).

Neurotransmission in cerebellar granule and Purkinje neurons is heavily dependent on calcium influx through P/Q-type calcium channels; the missense tottering mutation is reported to reduce calcium influx by ~40% in murine Purkinje cells (Wakamori et al., 1998) and reduce neurotransmission in several brain areas (Caddick et al., 1999; Ayata et al., 2000; Qian and

Noebels, 2000). However, the density of binding sites for an L-type calcium channel ligand is specifically increased in the tottering mouse cerebellum and attacks can be prevented by L-type calcium channel antagonists or initiated with an L-type calcium channel agonist (Campbell and

Hess, 1999). Tyrosine hydroxylase gene expression, linked to calcium influx through L-type calcium channels, is aberrantly expressed in tottering mouse cerebellar Purkinje cells (Hess and

Wilson, 1991; Austin et al., 1992). Taken together, these data suggested that increased calcium influx through L-type calcium channels in the cerebellum could generate abnormal signals that cause the periodic motor attacks and abnormal gene expression in tottering mice. 113

The present studies have attempted to test the hypothesis that the abnormal regulation of cerebellar L-type calcium channels is responsible for the cellular and behavioral phenotypes in tottering mice. The developmental profile of restraint-induced attacks and cerebellar TH mRNA expression was determined and compared to calcium channel expression and activity in the developing tottering mouse brain. Total calcium influx was assessed in several brain regions, including the cerebellum, where the contribution of L-type calcium channels was estimated using a specific antagonist. Aberrant cerebellar TH mRNA expression in tottering mice was measured after chronic blockade of L-type calcium channels. Behavioral responses and L-type calcium channel activity was compared in animals challenged with repeated L-type calcium channel activation. Finally, multiple trigger factors and potential therapeutics were tested in tottering mice to determine whether a common pathway could be established for this model of episodic neurological dysfunction. While overall the results support a role for increased calcium influx in tottering mouse phenotypes, there are indications that the initial hypothesis focusing on L-type calcium channels is too narrow. Coupled with evidence from previous literature on the tottering mouse and episodic disorders, the results of these investigations support the more general idea that hyperexcitability in the cerebellum is the underlying physiological defect responsible for tottering mouse attacks and aberrant gene expression.

Development of the tottering mouse phenotypes

Previous to these studies, the developmental onset of the motor attacks has been described as beginning sometime within the third to fourth week of age in tottering mice (Green 114 and Sidman, 1962), roughly parallel to the reported onset of tyrosine hydroxylase (TH) expression in the cerebellum (Hess and Wilson, 1991). However, the temporal relationship between cerebellar TH expression and tottering mouse motor attacks was not known. The onset of the motor attacks and aberrant TH mRNA expression was more precisely determined in the present study. The results suggest that the initiating event that renders tottering mice susceptible to restraint-induced dyskinesia occurs over a very short time period, one to two days, providing a discrete timeframe for future developmental analyses of this intriguing phenotype. Further, the pattern TH mRNA expression in the tottering mouse cerebellum differs from controls only after the onset of attacks. The timing of phenotypic onset suggest that aberrant TH mRNA expression in tottering mouse Purkinje cells may be a consequence of the intense cerebellar activation associated with motor attacks. Although functional consequences of excess cerebellar TH have not been documented, future studies could benefit from this epiphenomenon by expressing flourescent markers such as the enhanced green flourescent protein (EGFP) under the control of the TH gene promoter in tottering mice. These transgenic mice could enable more precise analyses of TH-positive tottering mouse Purkinje cell electrophysiology, since these cells are likely to participate in abnormal signaling.

As several of the calcium channel mutant mice express TH in cerebellar Purkinje cells, this phenotype could indicate that similar pathophysiological mechanisms are at play despite differences in the site of mutation. The severity of behavioral phenotypes varies among these mice, but seems to correlate with the abundance of TH mRNA expression (see Table 1.4). 115

Rocker mice (Zwingman et al., 2001) are a notable exception: these mice display neither attacks of dyskinesia nor persistent cerebellar TH mRNA expression. These characteristics are further evidence that attacks and aberrant gene expression are linked phenomena. A preliminary experiment demonstrated that rocker mice exhibit motor attacks almost indistinguishable from tottering mice when exposed to caffeine (personal observation). If cerebellar TH expression is truly a consequence of motor attacks, it should be possible to induce multiple attacks in rocker mice and produce aberrant TH mRNA expression in cerebellar Purkinje cells. Future studies will be required to conclusively demonstrate that persistent and ectopic TH expression in Purkinje cells is due to motor attacks.

Perturbed L-type calcium channel regulation in tottering mice

Studies using [3H]PN200-110 radioligand binding have confirmed a previous study using

[3H]nitrendipine in demonstrating that the density of binding sites for L-type calcium channel ligands is significantly greater in tottering mouse cerebellum (Chapter 3). Differences in binding density were confined to the cerebellum, a region previously implicated in the neurocircuitry of tottering mouse attacks. Cerebellar L-type calcium channels were differentially regulated in tottering mice chronically challenged with BAY K8644 (Chapter 4). In addition, blockade of L- type calcium channels decreased the abnormal TH mRNA expression in tottering mouse Purkinje cells (Chapter 3). Together, these results support a role for altered L-type calcium channel activity in the tottering mouse, consistent with the initial hypothesis. In addition, these results 116 confirm a pivotal role for the cerebellum as the site of abnormal L-type calcium channel regulation.

However, estimates of calcium uptake in tottering mouse cerebellum revealed little change in the total amount of calcium taken up by chemically-stimulated synaptosomes or in the relative contribution of L-type calcium channels to calcium influx as a result of the calcium channel mutation (Chapter 3 and 4). This result was consistent in several experimental paradigms and brain regions. In addition, the onset of restraint-induced attacks in developing tottering mice was not associated with a change in total calcium uptake. It is possible that changes in calcium channel regulation are cell-specific within the cerebellum; calcium influx

α through L-type and N-type calcium channels is augmented in Purkinje cells from 1A null mutant mice, while these calcium channel subtypes are unchanged in cerebellar granule cells (Jun et al.,

1999). It is not likely that such a discrete increase in tottering mouse cerebellar cells would be detected in the calcium uptake assay as performed in these studies. Altered regulation of L-type calcium channels could generate subtle changes in membrane excitability or L-type calcium channel second messenger systems that ultimately participate in the phenotypic abnormalities.

Thus even small increases in the number of L-type calcium channels could have a major impact on firing rates in cerebellar cells. Since the normal contribution of L-type calcium channels is modest, small changes may disproportionately affect the physiology of the cell.

In fact, L-type calcium channel activation can not be separated from membrane excitability; that is, calcium influx through L-type calcium channels is both a cause and a 117 consequence of membrane depolarization. Multiple factors can influence L-type calcium channel activity on a moment-to-moment basis, including G protein inhibition, phosphorylation and modulatory subunit interactions (reviewed by (Trimmer, 1998; Walker and De Waard, 1998;

Zamponi and Snutch, 1998). These confounding features are reflected in the somewhat disparate results from Chapters 2, 3 and 4. Increased ligand binding, for example, may result from either increased density of L-type calcium channels, or from greater association of equivalent numbers

of L-type calcium channels with modulatory β and α2δ subunits, which have been shown to increase dihydropyridine binding to the channel. The ability to initiate or prevent tottering mouse attacks with L-type calcium channel pharmacology may not rest solely on direct effects on the L- type calcium channel as much as on the subsequent change in membrane excitability produced by these agents. L-type calcium channel pharmacology would have the same effects if the cerebellar phenotypes in tottering mice are due to overexcitation, instead of specifically due to increased L-type calcium channels, as an agonist would augment and antagonists dampen excitability. Results from Chapter 5 using blockade of a ligand-gated ion channel support the more general conclusion that decreasing excitatory neurotransmission and subsequent calcium influx prevents tottering mouse attacks. This effect was also seen in a previous study using the

benzodiadepine diazepam (Syapin, 1983b), which acts as an agonist at GABAA receptors to increase synaptic inhibition. Conversely, an antagonist at GABA receptors, pentalenetetrazol, provokes attacks in tottering mice at doses that have little effect on control mice (Syapin, 1983b). 118

Thus, it appears that calcium influx through L-type calcium channels may be less important than net changes in membrane excitability.

Cerebellar hyperexcitability could explain why tottering mice are more sensitive to BAY

K8644 administration. Low doses of BAY K8644 that have little effect on control mice induce prolonged episodes of dyskinesia in tottering mice (Campbell and Hess, 1999) and personal observations). Cerebellar excitation has been linked to dyskinesia; in fact, dyskinesia can be produced in normal mice by direct microinjection of the glutamatergic agonist, kainic acid, into the cerebellum (C. Pizoli, personal communication). These observations support the notion that tottering mice are vulnerable to overexcitation in the cerebellum with attacks of dyskinesia resulting from agents that disrupt the precarious homeostatic mechanisms in this region.

Downregulation of L-type calcium channels after repeated bouts of BAY K8644-mediated excitation (Chapter 4) may be a cellular response unique to tottering mice because of their initial vulnerability to cerebellar activation.

Cerebellar TH mRNA expression in tottering mice

The abnormal TH expression in cerebellar Purkinje cells can also be interpreted in light of cerebellar hyperexcitability. Activity-dependent TH expression is mediated by L-type calcium channels via signaling pathways that converge on calcium-responsive enhancer elements in the TH gene promoter (Kilbourne et al., 1992; Best et al., 1995). Pharmacological blockade of

L-type calcium channels in vitro reduces or prevents the activity-dependent increase in TH expression (Vidal et al., 1989; Brosenitsch et al., 1998; Cigola et al., 1998). Conversely, 119 treatment with the L-type calcium channel agonist Bay K8644 potentiates TH expression in vitro

(Vidal et al., 1989; Cigola et al., 1998). L-type calcium channels appear to be the link between membrane depolarization and calcium-responsive gene-expression (Morgan and Curran, 1986).

This connection implies that abnormal TH gene expression may be due to either calcium influx through increased numbers of L-type calcium channels or a result of abnormally increased

Purkinje cell firing. Nimodipine may have dampened overall excitability in the cerebellum, since

L-type calcium channels are expressed on cerebellar granule cells, and inferior olivary cells, as well as Purkinje cells (Chin et al., 1992). Purkinje cells that receive less normal or abnormal input from their afferent connections will be less active, and this may be the crucial feature leading to decreases in TH mRNA expression. Additionally, TH mRNA expression in ten-week old tottering mice (Chapter 2) appeared to be less intense than in older animals (Chapter 3). This also suggests that TH mRNA expression accumulates as a consequence of repeated attacks over time. As discussed above, because L-type calcium channels and membrane excitability are inexorably linked, the chronic blockade experiment performed in Chapter 3 can not differentiate increased activity from increased L-type calcium channel influx. Thus, perhaps the better interpretation of these results is the less specific conclusion; that TH mRNA expression in cerebellar Purkinje cells is a phenotype linked to increased activity in the tottering mouse cerebellum.

The results of investigations into TH mRNA expression in the developing mouse brain tend to support this interpretation. Prior to the onset of restraint-induced attacks (and 120 presumably the vulnerability to other attack triggers as well), TH mRNA expression in tottering mouse Purkinje cells was not qualitatively different from control mice. One week after tottering mice develop attacks, however, TH mRNA was differentially expressed in Purkinje cells of the anterior cerebellum, a region devoid of TH expression in control mice at every age studied in this and other reports (Hess and Wilson, 1991; Austin et al., 1992). A previous study determined that

TH expression in tottering mouse cerebellar Purkinje cells precisely colocalizes with the banded distribution of another cerebellar marker, Zebrin II (Abbott et al., 1996). Interestingly, zebrin II expression in Purkinje cells corresponds to topographically organized projections from the inferior olive, the source of powerful climbing fiber excitatory afferents (reviewed by (Herrup and Kuemerle, 1997). Along with studies demonstrating a reduction in attack frequency and duration after lesions to the anterior cerebellum (Abbott et al., 2000), the ectopic localization of

TH mRNA in this region after the onset of attacks suggests that hyperexcitability in this cerebellar region promotes phenotypic abnormalities.

Attack triggers and potential therapeutics

The ability to provoke attacks in tottering mice provides control over a behavior historically regarded as “spontaneous”. Moreover, the identification of multiple pharmacological agents and environmental paradigms capable of triggering attacks can be used to identify shared mechanisms involved in attack initiation. Stimuli known to provoke neuronal activation in the form of stress and/or arousal responses, caffeine and alcohol administration, reliably precipitate attacks in tottering mice. To our knowledge, these mice represent the only animal model of an 121 episodic disorder to exhibit attacks in response to the same triggers common to human episodic disorders.

Triggers such as environmental stimuli were capable of inducing attacks in tottering mice, suggesting that a stress response may provide signals that ultimately contribute to attack initiation. Coupled with the observation that administration of caffeine and alcohol also induce

HPA axis activation (Ellis, 1966; Rivier et al., 1984; Pollard, 1988; Nicholson, 1989), it was hypothesized that stress hormones represent a common factor among attack triggers. Although the stress response is a cascade of hormone and neurotransmitter release which ultimately affect autonomic functions and behavioral responses, the predominant hormonal factors in the HPA axis are CRF (released from the hypothalamus), ACTH (secreted from the pituitary) and CORT

(released from the adrenal glands). The results of adrenalectomy and dexamethasone challenge argue against CORT as a necessary or sufficient factor for attack initiation. Likewise, antalarmin blockade of CRF-1 receptors failed to prevent restraint-induced attacks. However, CRF or other neurohormones, such as thyrotropin-releasing hormone (implicated in arousal), may have effects at other central sites that ultimately promote the expression of attacks. Future studies designed to manipulate hormonal systems in tottering mice will address these questions.

Although the initial hormonal candidates for a shared triggering mechanism, CORT and

CRF, do not appear to dramatically influence the attack phenotype, further investigation into shared mechanisms of diverse triggers is likely to provide new insight into conditions that promote attacks of neurological dysfunction. One promising avenue is suggested by the 122 observation that L-type calcium channel blockade prevents attacks induced by caffeine, alcohol and restraint. Further, restraint-induced attacks were also prevented by blockade of a ligand- gated calcium channel. These results suggest that these triggers require either calcium influx or neuronal excitation to produce attacks. This conclusion leads back to the question of whether calcium influx is a cause or consequence of abnormal neuronal activation, which will require further experimental study. Nonetheless, the results of investigations into triggers and potential therapeutics support the hypothesis that cerebellar hyperexcitability is a cause of attacks in tottering mice.

General model of hyperexcitability in the tottering mouse cerebellum

Overall, the results from these experiments suggest that the tottering mutation may lead to a susceptibility to increased excitation in the cerebellum, through compensatory mechanisms including, but probably not limited to, L-type calcium channel upregulation. The unique neurocircuitry of the cerebellum comprises a potential feed-forward pathway that may promote the propagation of abnormal signals. As tottering mouse Purkinje cells have known abnormalities, including ectopic TH mRNA expression, abnormal signals may be produced in

α these cells first. In fact, Purkinje cells express the affected 1A subunit protein more abundantly than almost any other region of the brain (Hillman et al., 1991) and synaptic transmission from these cells can be prevented by P/Q-type calcium channel antagonists (Doroshenko et al., 1997).

Thus, the Purkinje cell is a likely site of secondary changes to compensate for the P/Q-type calcium channel mutation. 123

Purkinje cell activation is accomplished through excitatory connections from the inferior olive via climbing fibers, and from granule cell parallel fibers via mossy fiber connections

(Figure 6.1). The interaction of trigger factors, such as caffeine, alcohol or neurohormones with the Purkinje cell membrane may influence membrane excitability just enough to generate action potentials from EPSPs that normally would not be sufficient to activate the neuron. An increase in L-type calcium channels in the Purkinje cell membrane would accomplish the same net effect.

Thus, a number of potential factors could converge on a cell whose ionic signaling capability is already perturbed by an ion channel abnormality, such as the tottering P/Q-type calcium channel mutation. Once the excitation of the Purkinje cell is accomplished, the cerebellar system may tend to propagate and even enhance the abnormal signal. Purkinje cells inhibit the spontaneously active neurons of the deep cerebellar nuclei (DCN), which send excitatory efferent projections to the red nucleus and the thalamus. Another DCN efferent projection is inhibitory, however, and these neurons synapse on the inferior olivary nucleus. As the inferior olive is the source of excitatory climbing fiber input to the cerebellum, the overall effect of increased Purkinje cell activation is a decrease in the inhibition of the inferior olive. This loop tends to promote another round of Purkinje cell activation, continuing the cycle. 124

Figure 6.1 Model diagram describing neuroanatomical connections that may contribute to hyperexitability in the tottering mouse cerebellar cortex. A potential feedback loop, indicated in black, is formed from the projections between Purkinje cells, the deep cerebellar nuclei and the inferior olive. Movement abnormalities may result from cerebellar hyperexcitability via subsequent activation of the pre-motor cortex and spinal tracts; these projections are indicated in white. Arrowheads indicate excitatory connections; inhibitory connections are depicted as flat lines. PC; Purkinje cell; DCN, deep cerebellar nuclei, IO, inferior olive; RN, red nucleus; SC, spinal cord. 125

In fact, there is some preliminary evidence to support the idea that tottering mouse

Purkinje cell firing rates are intrinsically different than controls. Tottering mouse Purkinje cells recorded in slice preparations appear to receive more excitatory input than controls, suggesting altered activity (J.Netzeband, personal communication). The results presented here are that tottering mice abnormally regulate L-type calcium channels, are more sensitive to calcium channel activation, and exhibit attacks that have a rapid developmental onset which can be blocked by calcium channel antagonists or an inhibitor of excitatory neurotransmission.

Transient hyperexcitability has been proposed as a feature common to episodic neurological disorders caused by ion channel mutations in humans (Ptacek, 1998). As demonstrated in this thesis, the tottering mouse provides an excellent animal model to investigate triggers, potential therapeutics, and the mechanisms generating neuronal hyperexcitability as a secondary consequence of ion channel mutation. 126

REFERENCES

Abbott LC, Isaacs KR, Heckroth JA (1996) Co-localization of tyrosine hydroxylase and zebrin II immunoreactivities in Purkinje cells of the mutant mice, tottering and tottering/leaner. Neuroscience 71:461-475. Abbott LC, Bump M, Brandl A, De Laune S (2000) Investigation of the role of the cerebellum in the myoclonic-like movement disorder exhibited by tottering mice. Mov Disord 15:53- 59. Armario A, Castellanos JM (1984) A comparison of corticoadrenal and gonadal responses to acute immobilization stress in rats and mice. Physiol Behav 32:517-519. Austin MC, Schultzberg M, Abbott LC, Montpied P, Evers JR, Paul SM, Crawley JN (1992) Expression of tyrosine hydroxylase in cerebellar Purkinje neurons of the mutant tottering and leaner mouse. Brain Res Mol Brain Res 15:227-240. Ayata C, Shimizu-Sasamata M, Lo EH, Noebels JL, Moskowitz MA (2000) Impaired neurotransmitter release and elevated threshold for cortical spreading depression in mice with mutations in the alpha1A subunit of P/Q type calcium channels. Neuroscience 95:639-645. Barlow SM, Morrison PJ, Sullivan FM (1975) Effects of acute and chronic stress on plasma corticosterone levels in the pregnant and non-pregnant mouse. J Endocrinol 66:9O-9. Battistini S, Stenirri S, Piatti M, Gelfi C, Righetti PG, Rocchi R, Giannini F, Battistini N, Guazzi GC, Ferrari M, Carrera P (1999) A new CACNA1A gene mutation in acetazolamide- responsive familial hemiplegic migraine and ataxia. Neurology 53:38-43. Berridge CW, Dunn AJ (1989) Restraint-stress-induced changes in exploratory behavior appear to be mediated by norepinephrine-stimulated release of CRF. J Neurosci 9:3513-3521. Best JA, Chen Y, Piech KM, Tank AW (1995) The response of the tyrosine hydroxylase gene to cyclic AMP is mediated by two cyclic AMP-responsive elements. J Neurochem 65:1934- 1943. Bhatia KP (1999) The paroxysmal dyskinesias. J Neurol 246:149-155. Boel M, Casaer P (1988) Familial periodic ataxia responsive to . Neuropediatrics 19:218-220. Borycz J, Bugajski AJ, Gadek-Michalska A, Bugajski J (1993) Calcium channel blockers impair the pituitary-adrenocortical responses to central adrenergic receptors stimulation. J Physiol Pharmacol 44:161-170. Bressman SB, Fahn S, Burke RE (1988) Paroxysmal non-kinesigenic dystonia. Adv Neurol 50:403-413. Brosenitsch TA, Salgado-Commissariat D, Kunze DL, Katz DM (1998) A role for L-type calcium channels in developmental regulation of transmitter phenotype in primary sensory neurons. J Neurosci 18:1047-1055. 127

Browne DL, Gancher ST, Nutt JG, Brunt ER, Smith EA, Kramer P, Litt M (1994) Episodic ataxia/myokymia syndrome is associated with point mutations in the human potassium channel gene, KCNA1. Nat Genet 8:136-140. Burgess DL, Jones JM, Meisler MH, Noebels JL (1997) Mutation of the Ca2+ channel beta subunit gene Cchb4 is associated with ataxia and seizures in the lethargic (lh) mouse. Cell 88:385-392. Caddick SJ, Wang C, Fletcher CF, Jenkins NA, Copeland NG, Hosford DA (1999) Excitatory but not inhibitory synaptic transmission is reduced in lethargic (Cacnb4(lh)) and tottering (Cacna1atg) mouse thalami. J Neurophysiol 81:2066-2074. Campbell DB, Hess EJ (1997) Rapid genotyping of mutant mice using dried blood spots for polymerase chain reaction (PCR) analysis. Brain Res Brain Res Protoc 1:117-123. Campbell DB, Hess EJ (1998) Cerebellar circuitry is activated during convulsive episodes in the tottering (tg/tg) mutant mouse. Neuroscience 85:773-783. Campbell DB, Hess EJ (1999) L-type calcium channels contribute to the tottering mouse dystonic episodes. Mol Pharmacol 55:23-31. Campbell DB, North JB, Hess EJ (1999) Tottering mouse motor dysfunction is abolished on the Purkinje cell degeneration (pcd) mutant background. Exp Neurol 160:268-278. Castellano A, Wei X, Birnbaumer L, Perez-Reyes E (1993) Cloning and expression of a neuronal calcium channel β subunit. J Biol Chem 268:12359-12366. Catterall WA (2000) Structure and regulation of voltage-gated Ca2+ channels. Annu Rev Cell Dev Biol 16:521-555. Charlier C, Singh NA, Ryan SG, Lewis TB, Reus BE, Leach RJ, Leppert M (1998) A pore mutation in a novel KQT-like potassium channel gene in an idiopathic epilepsy family. Nat Genet 18:53-55. Chin H, Smith MA, Kim HL, Kim H (1992) Expression of dihydropyridine-sensitive brain calcium channels in the rat central nervous system. FEBS 299:69-74. Cigola E, Volpe BT, Lee JW, Franzen L, Baker H (1998) Tyrosine hydroxylase expression in primary cultures of olfactory bulb: role of L-type calcium channels. J Neurosci 18:7638- 7649. Cooper EC, Jan LY (1999) Ion channel genes and human neurological disease: recent progress, prospects, and challenges. Proc Natl Acad Sci U S A 96:4759-4766. Dallman MF (2000) Editorial: Moments in time - the neonatal rat hypothalamo-pituitary-adrenal axis. Endocrinology 141:1590-1592. Davidoff AJ, Maki T, Ellingsen O, Marsh JD (1997) Expression of calcium channels in adult cardiac myocytes is regulated by calcium. J Mol Cell Cardiol 29:1791-1803. De Souza EB, Perrin MH, Insel TR, Rivier J, Vale WW, Kuhar MJ (1984) Corticotropin- releasing factor receptors in rat forebrain: autoradiographic identification. Science 224:1449-1451. 128

De Souza EB, Insel TR, Perrin MH, Rivier J, Vale WW, Kuhar MJ (1985) Corticotropin- releasing factor receptors are widely distributed within the rat central nervous system: an autoradiographic study. J Neurosci 5:3189-3203. Demirkiran M, Jankovic J (1995) Paroxysmal dyskinesias: clinical features and classification. Ann Neurol 38:571-579. Doroshenko PA, Woppmann A, Miljanich G, Augustine GJ (1997) Pharmacologically distinct presynaptic calcium channels in cerebellar excitatory and inhibitory synapses. Neuropharmacology 36:865-872. Dove LS, Abbott LC, Griffith WH (1998) Whole-cell and single channel analysis of P-type calcium currents in Purkinje cells of leaner mutant mice. J Neurosci 18:7687-7699. Downen M, Belkowski S, Knowles H, Cardillo M, Prystowsky MB (1999) Developmental expression of voltage-gated potassium channel beta subunits. Brain Res Dev Brain Res 117:71-80. Doyle J, Ren X, Lennon G, Stubbs L (1997) Mutations in the Cacnl1a4 calcium channel gene are associated with seizures, cerebellar degeneration, and ataxia in tottering and leaner mutant mice. Mamm Genome 8:113-120. Dunn S (1988) Multiple calcium channels in synaptosomes: voltage dependence of 1,4- dihydropyridine binding and effects on function. Biochemistry 27:5275-5281. Ellis F (1966) Effect of ethanol on plasma corticosterone levels. J Pharmacol Exp Ther 153:121- 127. Erdman RD, Yamamura HI, Roeske WR (1983) The ontogeny of specific binding sites for the calcium channel antagonist, nitrendipine, in mouse brain and heart. Brain Res 278:327- 331. Felts PA, Yokoyama S, Dib-Hajj S, Black JA, Waxman SG (1997) alpha- subunit mRNAs I, II, III, NaG, Na6 and hNE (PN1): different expression patterns in developing rat nervous system. Brain Res Mol Brain Res 45:71-82. Fletcher CF, Lutz CM, O'Sullivan TN, Shaughnessy JD, Jr., Hawkes R, Frankel WN, Copeland NG, Jenkins NA (1996) Absence epilepsy in tottering mutant mice is associated with calcium channel defects. Cell 87:607-617. Fredholm B, Battig K, Holmen J, Nehlig A, Zvartau E (1999) Actions of caffeine in the brain with special reference to factors that contribute to its widespread use. Pharmacological Reviews 51:83-113. Gardner K, Hoffman EP (1998) Current status of genetic discoveries in migraine: familial hemiplegic migraine and beyond. Curr Opin Neurol 11:211-216. Gaur S, Newcomb R, Rivnay B, Bell JR, Yamashiro D, Ramachandran J, Miljanich GP (1994) Calcium channel antagonist peptides define several components of transmitter release in the hippocampus. Neuropharmacology 33:1211-1219. Green MC, Sidman RL (1962) Tottering-a neuromuscular mutation in the mouse. J Hered 53:233-237. 129

Hauger RL, Millan MA, Lorang M, Harwood JP, Aguilera G (1988) Corticotropin-releasing factor receptors and pituitary adrenal responses during immobilization stress. Endocrinology 123:396-405. Hell JW, Westenbroek RE, Warner C, Ahlijanian MK, Prystay W, Gilbert MM, Snutch TP, Catterall WA (1993) Identification and differential subcellular localization of the neuronal class C and class D L-type calcium channel α1 subunits. J Cell Biol 123:949- 962. Herrup K, Wilczynski SL (1982) Cerebellar cell degeneration in the leaner mutant mouse. Neuroscience 7:2185-2196. Herrup K, Kuemerle B (1997) The compartmentalization of the cerebellum. Ann Rev Neurosci 20:61-90. Hess EJ, Wilson MC (1991) Tottering and leaner mutations perturb transient developmental expression of tyrosine hydroxylase in embryologically distinct Purkinje cells. Neuron 6:123-132. Hillman D, Chen S, Aung TT, Cherksey B, Sugimori M, Llinas RR (1991) Localization of P- type calcium channels in the central nervous system. Proc Natl Acad Sci U S A 88:7076- 7080. Hofele K, Benecke R, Auburger G (1997) Gene locus FPD1 of the dystonic Mount-Reback type of autosomal-dominant paroxysmal choreoathetosis. Neurology 49:1252-1257. Isaacs KR, Abbott LC (1992) Development of the paramedian lobule of the cerebellum in wild- type and tottering mice. Dev Neurosci 14:386-393. Isaacs KR, Abbott LC (1995) Cerebellar volume decreases in the tottering mouse are specific to the molecular layer. Brain Res Bull 36:309-314. Iwasaki S, Momiyama A, Uchitel OD, Takahashi T (2000) Developmental changes in calcium channel types mediating central synaptic transmission. J Neurosci 20:59-65. Janicki PK, Siembab D, Paulo EA, Krzascik P (1988) Single-dose kinetics of nifedipine in rat plasma and brain. Pharmacology 36:183-187. Janis RA, Silver PJ, Triggle DJ (1987) Drug action and cellular calcium regulation. Adv Drug Res 16:309-591. Jarman PR, Bhatia KP, Davie C, Heales SJ, Turjanski N, Taylor-Robinson SD, Marsden CD, Wood NW (2000) Paroxysmal dystonic choreoathetosis: clinical features and investigation of pathophysiology in a large family. Mov Disord 15:648-657. Jen J (1999) Calcium channelopathies in the central nervous system. Curr Opin Neurobiol 9:274- 280. Jen J, Yue Q, Nelson SF, Yu H, Litt M, Nutt J, Baloh RW (1999) A novel nonsense mutation in CACNA1A causes episodic ataxia and hemiplegia [see comments]. Neurology 53:34-37. Jinnah HA, Sepkuty JP, Ho T, Yitta S, Drew T, Rothstein JD, Hess EJ (2000) Calcium channel agonists and dystonia in the mouse. Mov Disord 15:542-551. 130

Jodice C, Mantuano E, Veneziano L, Trettel F, Sabbadini G, Calandriello L, Francia A, Spadaro M, Pierelli F, Salvi F, Ophoff RA, Frants RR, Frontali M (1997) Episodic ataxia type 2 (EA2) and spinocerebellar ataxia type 6 (SCA6) due to CAG repeat expansion in the CACNA1A gene on chromosome 19p. Hum Mol Genet 6:1973-1978. Jones SW (1998) Overview of voltage-dependent calcium channels. J Bioenerg Biomembr 30:299-312. Jun K, Piedras-Renteria ES, Smith SM, Wheeler DB, Lee SB, Lee TG, Chin H, Adams ME, Scheller RH, Tsien RW, Shin HS (1999) Ablation of P/Q-type Ca(2+) channel currents, altered synaptic transmission, and progressive ataxia in mice lacking the alpha(1A)- subunit. Proc Natl Acad Sci U S A 96:15245-15250. Kakizawa S, Yamasaki M, Watanabe M, Kano M (2000) Critical period for activity-dependent synapse elimination in developing cerebellum. J Neurosci 20:4954-4961. Kaplan BJ, Seyfried TN, Glaser GH (1979) Spontaneous polyspike discharges in an epileptic mutant mouse (tottering). Exp Neurol 66:577-586. Kazazoglou T, Schmid A, Renaud JF, Lazdunski M (1983) Ontogenic appearance of Ca2+ channels characterized as binding sites for nitrendipine during development of nervous, skeletal and cardiac muscle systems in the rat. FEBS 164:75-79. Kilbourne EJ, Nankova BB, Lewis EJ, McMahon A, Osaka H, Sabban DB, Sabban EL (1992) Regulated expression of the tyrosine hydroxylase gene by membrane depolarization. J Biol Chem 267:7563-7569. Klugbauer N, Dai S, Specht V, Lacinova L, Marais E, Bohn G, Hofmann F (2000) A family of γ- like calcium channel subunits. FEBS Letters 470:189-197. Koob GF (1999) Corticotropin-releasing factor, norepinephrine, and stress. Biol Psychiatry 46:1167-1180. Laird SM, Hinson JP, Vinson GP, Mallick N, Kapas S, Teja R (1991) Control of steroidogenesis by the calcium messenger system in human adrenocortical cells. J Mol Endocrinol 6:45- 51. Lance JW (1977) Familial paroxysmal dystonic choreoathetosis and its differentiation from related syndromes. Ann Neurol 2:285-293. Larkin JG, Thompson GG, Scobie G, Forrest G, Drennan JE, Brodie MJ (1992) Dihydropyridine calcium antagonists in mice: blood and brain pharmacokinetics and efficacy against pentylenetetrazol seizures. Epilepsia 33:760-769. Letts VA, Felix R, Biddlecome GH, Arikkath J, Mahaffey CL, Valenzuela A, Bartlett FS, 2nd, Mori Y, Campbell KP, Frankel WN (1998) The mouse stargazer gene encodes a neuronal Ca2+-channel gamma subunit. Nat Genet 19:340-347. Levitt P, Noebels JL (1981) Mutant mouse tottering: selective increase of locus ceruleus axons in a defined single-locus mutation. Proc Natl Acad Sci U S A 78:4630-4634. 131

Lewohl J, Wilson W, Mayfield R, Brozowski S, Morrisett R, Harris R (1999) G-protein-coupled inwardly rectifying potassium channels are targets of alcohol action. Nat Neurosci 2:1084-1089. Lin FH, Barun S, Lutz CM, Wang Y, Hosford DA (1999) Decreased 45Ca2+ uptake in P/Q-type calcium channels in homozygous lethargic (Cacnb4llh) mice is associated with increased β3 and decreased β4 calcium channel subunit mRNA expression. Mol Brain Res 71:1-10. Litzinger MJ, Grover BB, Saderup S, Abbott JR (1993) Voltage sensitive calcium channels mark a critical period in mouse neurodevelopment. Int J Devl Neuroscience 11:17-24. Lorenzon NM, Lutz CM, Frankel WN, Beam KG (1998) Altered calcium channel currents in Purkinje cells of the neurological mutant mouse leaner. J Neurosci 18:4482-4489. Lovenberg TW, Liaw CW, Grigoriadis DE, Clevenger W, Chalmers DT, De Souza EB, Oltersdorf T (1995) Cloning and characterization of a functionally distinct corticotropin- releasing factor receptor subtype from rat brain. Proc Natl Acad Sci U S A 92:836-840. Lovinger DM, White G, Weight FF (1990) NMDA receptor-mediated synaptic excitation selectively inhibited by ethanol in hippocampal slice from adult rat. J Neurosci 10:1372- 1379. Ludwig A, Flockerzi V, Hofmann F (1997) Regional expression and cellular localization of the α β 1 and subunit of high voltage-activated calcium channels in rat brain. Luebke JI, Dunlap K, Turner TJ (1993) Multiple calcium channel types control glutamatergic synaptic transmission in the hippocampus. Neuron 11:895-902. Maki T, Gruver EJ, Davidoff AJ, Izzo N, Toupin D, Colucci W, Marks AR, Marsh JD (1996) Regulation of calcium channel expression in neonatal myocytes by catecholamines. J Clin Invest 97:656-663. Mantegazza M, Fasolato C, Hescheler J, Pietrobon D (1995) Stimulation of single L-type calcium channels in rat pituitary GH3 cells by thyrotropin-releasing hormone. Embo J 14:1075-1083. McEnery MW, Copeland TD, Vance CL (1998) Altered expression and assembly of N-type calcium channel alpha1B and beta subunits in epileptic lethargic (lh/lh) mouse. J Biol Chem 273:21435-21438. Momiyama A, Takahashi T (1994) Calcium channels responsible for potassium-induced transmitter release at rat cerebellar synapses. J Physiol 476:197-202. Morgan JI, Curran T (1986) Role of ion flux in the control of c-fos expression. Nature 322:552- 555. Mori Y, Wakamori M, Oda S, Fletcher CF, Sekiguchi N, Mori E, Copeland NG, Jenkins NA, Matsushita K, Matsuyama Z, Imoto K (2000) Reduced voltage sensitivity of activation of P/Q-type Ca2+ channels is associated with the ataxic mouse mutation rolling Nagoya (tg(rol)). J Neurosci 20:5654-5662. Mount LA, Reback S (1940) Familial Paroxysmal Choreoathetosis: Preliminary report on a hitherto undescribed clinical syndrome. Arch Neurol 44:841-847. 132

Mourre C, Cervera P, Lazdunski M (1987) Autoradiographic analysis in rat brain of the postnatal ontogeny of voltage-dependent Na+ channels, Ca2+-dependent K+ channels, and slow Ca2+ channels identified as receptors for , apamin, and (-)-desmethoxyverapamil. Brain Res 417:21-32. Muramoto O, Kanazawa I, Ando K (1981) Neurotransmitter abnormality in Rolling mouse Nagoya, an ataxic mutant mouse. Brain Res 215:295-304. Muramoto O, Ando K, Kanazawa I (1982) Central noradrenaline metabolism in cerebellar ataxic mice. Brain Res 237:387-395. Nestoros JN (1980) Ethanol specifically potentiates GABA-mediated neurotransmission in feline cerebral cortex. Science 209:708-710. Neville BG, Besag FM, Marsden CD (1998) Exercise induced steroid dependent dystonia, ataxia, and alternating hemiplegia associated with epilepsy. J Neurol Neurosurg Psychiatry 65:241-244. Nicholson SA (1989) Stimulatory effect of caffeine on the hypothalamo-pituitary- adrenocortical axis in the rat. J Endocrinol 122:535-543. Nikodijevic B, Nikodijevic-Kedeva D, Oshima M, Paige B, Guroff G (1994) The long-term down-regulation of dihydropyridine receptors by Bay K 8644 in PC12 cells. J Neurosci Res 37:71-82. Noebels JL (1979) Analysis of inherited epilepsy using single locus mutations in mice. Fed Proc 38:2405-2410. Noebels JL (1984) A single gene error of noradrenergic axon growth synchronizes central neurones. Nature 310:409-411. Noebels JL, Sidman RL (1979) Inherited epilepsy: spike-wave and focal motor seizures in the mutant mouse tottering. Science 204:1334-1336. O'Neill SK, Bolger GT (1988) Enantiomer selectivity and the development of tolerance to the behavioral effects of the calcium channel activator BAY K 8644. Brain Res Bull 21:865- 872. O'Neill SK, McKay DW, Campbell N, Triggle CR, Crowley M, Bolger GT (1990) Comparative behavioral, neurochemical and pharmacological activities of dihydropyridine calcium channel activating drugs. J Pharmacol Exp Ther 253:905-912. Ohno K, Hutchinson DO, Milone M, Brengman JM, Bouzat C, Sine SM, Engel AG (1995) Congenital myasthenic syndrome caused by prolonged acetylcholine receptor channel openings due to a mutation in the M2 domain of the epsilon subunit. Proc Natl Acad Sci U S A 92:758-762. Ophoff RA, Terwindt GM, Vergouwe MN, van Eijk R, Oefner PJ, Hoffman SM, Lamerdin JE, Mohrenweiser HW, Bulman DE, Ferrari M, Haan J, Lindhout D, van Ommen GJ, Hofker MH, Ferrari MD, Frants RR (1996) Familial hemiplegic migraine and episodic ataxia type-2 are caused by mutations in the Ca2+ channel gene CACNL1A4. Cell 87:543-552. 133

Pang PK, Wang R, Shan J, Karpinski E, Benishin CG (1990) Specific inhibition of long-lasting, L-type calcium channels by synthetic parathyroid hormone. Proc Natl Acad Sci U S A 87:623-627. Perrin MH, Donaldson CJ, Chen R, Lewis KA, Vale WW (1993) Cloning and functional expression of a rat brain corticotropin releasing factor (CRF) receptor. Endocrinology 133:3058-3061. Pittock SJ, Joyce C, O'Keane V, Hugle B, Hardiman MO, Brett F, Green AJ, Barton DE, King MD, Webb DW (2000) Rapid-onset dystonia-parkinsonism: a clinical and genetic analysis of a new kindred. Neurology 55:991-995. Pollard I (1988) Increases in plasma concentrations of steroids in the rat after the administration of caffeine: comparison with plasma disposition of caffeine. J Endocrinol 119:275-280. Ptacek LJ (1997) Channelopathies: ion channel disorders of muscle as a paradigm for paroxysmal disorders of the nervous system. Neuromuscul Disord 7:250-255. Ptacek LJ (1998) The place of migraine as a channelopathy. Curr Opin Neurol 11:217-226. Ptacek LJ (1999) Ion channel diseases: episodic disorders of the nervous system. Semin Neurol 19:363-369. Ptacek LJ, George AL, Jr., Barchi RL, Griggs RC, Riggs JE, Robertson M, Leppert MF (1992) Mutations in an S4 segment of the adult skeletal muscle sodium channel cause paramyotonia congenita. Neuron 8:891-897. Ptacek LJ, George AL, Jr., Griggs RC, Tawil R, Kallen RG, Barchi RL, Robertson M, Leppert MF (1991) Identification of a mutation in the gene causing hyperkalemic periodic paralysis. Cell 67:1021-1027. Ptacek LJ, Tawil R, Griggs RC, Meola G, McManis P, Barohn RJ, Mendell JR, Harris C, Spitzer R, Santiago F, et al. (1994a) Sodium channel mutations in acetazolamide-responsive myotonia congenita, paramyotonia congenita, and hyperkalemic periodic paralysis. Neurology 44:1500-1503. Ptacek LJ, Tawil R, Griggs RC, Engel AG, Layzer RB, Kwiecinski H, McManis PG, Santiago L, Moore M, Fouad G, et al. (1994b) Dihydropyridine receptor mutations cause hypokalemic periodic paralysis. Cell 77:863-868. Qian J, Noebels JL (2000) Presynaptic Ca(2+) influx at a mouse central synapse with Ca(2+) channel subunit mutations. J Neurosci 20:163-170. Richards RN, Barnett HJ (1968) Paroxysmal dystonic choreoathetosis. A family study and review of the literature. Neurology 18:461-469. Rivier C, Bruhn T, Vale W (1984) Effect of ethanol on the hypothalamic-pituitary-adrenal axis in the rat: role of corticotropin-releasing factor (CRF). J Pharmacol Exp Ther 229:127- 131. Sashihara S, Oh Y, Black JA, Waxman SG (1995) Na+ channel beta 1 subunit mRNA expression in developing rat central nervous system. Brain Res Mol Brain Res 34:239- 250. 134

Satoh J, Suzuki K (1990) Tyrosine hydroxylase-immunoreactive neurons in the mouse cerebral cortex during the postnatal period. Dev Brain Res 53:1-5. Sawada K, Komatsu S, Haga H, Sun XZ, Hisano S, Fukui Y (1999) Abnormal expression of tyrosine hydroxylase immunoreactivity in cerebellar cortex of ataxic mutant mice. Brain Res 829:107-112. Sheng ZH, Westenbroek RE, Catterall WA (1998) Physical link and functional coupling of presynaptic calcium channels and the synaptic vesicle docking/fusion machinery. J Bioenerg Biomembr 30:335-345. Singh R, Scheffer IE, Crossland K, Berkovic SF (1999a) Generalized epilepsy with febrile seizures plus: a common childhood- onset genetic epilepsy syndrome. Ann Neurol 45:75- 81. Singh R, Macdonell RA, Scheffer IE, Crossland KM, Berkovic SF (1999b) Epilepsy and paroxysmal movement disorders in families: evidence for shared mechanisms. Epileptic Disord 1:93-99. Sotelo C (1989) Formation of cerebellar projection maps: development of olivo- and spinocerebellar projections. In: Developmental Neurobiology (Evrard P, Minkowski A, eds), pp 155-168. New York: Nestec Ltd., Vevey/Raven Press, Ltd. Starr TV, Prystay W, Snutch TP (1991) Primary structure of a calcium channel that is highly expressed in the rat cerebellum. Proc Natl Acad Sci U S A 88:5621-5625. Stea A, Tomlinson WJ, Soong TW, Bourinet E, Dubel SJ, Vincent SR, Snutch TP (1994) α Localization and functional properties of a rat brain 1A calcium channel reflect similarities to neuronal Q- and P-type channels. Proc Natl Acad Sci U S A 91:10576- 10580. Steinlein OK, Mulley JC, Propping P, Wallace RH, Phillips HA, Sutherland GR, Scheffer IE, Berkovic SF (1995) A missense mutation in the neuronal nicotinic acetylcholine receptor alpha 4 subunit is associated with autosomal dominant nocturnal frontal lobe epilepsy. Nat Genet 11:201-203. Syapin P (1983a) Inhibition of pentylenetetrazol induced genetically-determined stereotypic convulsions in tottering mutant mice by diazepam. Pharmacol Biochem Behav 18:389- 394. Syapin PJ (1982) Effects of the tottering mutation in the mouse: multiple neurologic changes. Exp Neurol 76:566-573. Syapin PJ (1983b) Inhibition of pentylenetetrazol induced genetically-determined stereotypic convulsions in tottering mutant mice by diazepam. Pharmacol Biochem Behav 18:389- 394. Takahashi T, Momiyama A (1993) Different types of calcium channels mediate central synaptic transmission. Nature 366:156-158. 135

Takayama C, Nakagawa S, Watanabe M, Mishina M, Inoue Y (1996) Developmental changes in expression and distribution of the glutamate receptor channel delta 2 subunit according to the Purkinje cell maturation. Brain Res Dev Brain Res 92:147-155. Tamaki Y, Oda S, Kameyama Y (1986) Postnatal locomotion development in a neurological mutant of rolling mouse Nagoya. Dev Psychobiol 19:67-77. Tanaka O, Sakagami H, Kondo H (1995) Localization of mRNAs of voltage-dependent Ca(2+)- channels: four subtypes of alpha 1- and beta-subunits in developing and mature rat brain. Brain Res Mol Brain Res 30:1-16. Tank AW, Ham L, Curella P (1986) Induction of tyrosine hydroxylase by cyclic AMP and glucocorticoids in a rat pheochromocytoma cell line: effect of the inducing agents alone or in combination on the enzyme levels and rate of synthesis of tyrosine hydroxylase. Mol Pharmacol 30:486-496. Terwindt GM, Ophoff RA, Haan J, Sandkuijl LA, Frants RR, Ferrari MD (1998) Migraine, ataxia and epilepsy: a challenging spectrum of genetically determined calcium channelopathies. Dutch Migraine Genetics Research Group. Eur J Hum Genet 6:297-307. Trimmer JS (1998) Regulation of ion channel expression by cytoplasmic subunits. Curr Opin Neurobiol 8:370-374. Turner TJ, Goldin SM (1985) Calcium channels in rat brain synaptosomes: identification and pharmacological characterization. J Neurosci Res 5:841-849. Turner TJ, Adams ME, Dunlap K (1993) Multiple Ca2+ channel types coexist to regulate synaptosomal neurotransmitter release. Proc Natl Acad Sci U S A 90:9518-9522. Vale W, Spiess J, Rivier C, Rivier J (1981) Characterization of a 41-residue ovine hypothalamic peptide that stimulates secretion of corticotropin and beta-endorphin. Science 213:1394- 1397. Vidal S, Raynaud B, Weber MJ (1989) The role of Ca2+ channels of the L-type in neurotransmitter plasticity of cultured sympathetic neurons. Brain Res Mol Brain Res 6:187-196. Wakamori M, Yamazaki K, Matsunodaira H, Teramoto T, Tanaka I, Niidome T, Sawada K, Nishizawa Y, Sekiguchi N, Mori E, Mori Y, Imoto K (1998) Single tottering mutations responsible for the neuropathic phenotype of the P-type calcium channel. J Biol Chem 273:34857-34867. Walker D, De Waard M (1998) Subunit interaction sites in voltage-dependent Ca2+ channels: role in channel function. Trends Neurosci 21:148-154. Webster EL, Lewis DB, Torpy DJ, Zachman EK, Rice KC, Chrousos GP (1996) In vitro and in vivo characterization of antalarmin, a nonpeptide corticotropin-releasing hormone (CRH) receptor antagonist: suppression of pituitary ACTH release and peripheral inflammation. Endocrinology 137:5747-5750. 136

Westenbroek RE, Bausch SB, Lin R, Franck JE, Noebels JL, Catterall WA (1998) Upregulation of L-type Ca2+ channels in reactive astrocytes after brain injury, hypomyelination, and ischemia. J Neurosci 18:2321-2334. Williams ME, Marubio LM, Deal CR, Hans M, Brust PF, Philipson LH, Miller RJ, Johnson EC, α Harpold MM, Ellis SB (1994) Structure and functional characterization of neuronal 1E calcium channel subtypes. J Biol Chem 269:22347-22357. Wynn PC, Hauger RL, Holmes MC, Millan MA, Catt KJ, Aguilera G (1984) Brain and pituitary receptors for corticotropin releasing factor: localization and differential regulation after adrenalectomy. Peptides 5:1077-1084. Xie J, McCobb DP (1998) Control of alternative splicing of potassium channels by stress hormones. Science 280:443-446. Yang J, Tsien RW (1993) Enhancement of N- and L-type calcium channel currents by protein kinase C in frog sympathetic neurons. Neuron 10:127-136. Yoshor D, Zhu Z, Burgess DL, Noebels JL (1997) Aberrant expression of tyrosine hydroxylase in Purkinje cells of the mutant mouse lethargic (lh). Soc Neurosci Abst 23:40. Yu B, Shinnick-Gallagher P (1998) Corticotropin-releasing factor increases dihydropyridine- and neurotoxin-resistant calcium currents in neurons of the central amygdala. J Pharmacol Exp Ther 284:170-179. Zamponi GW, Snutch TP (1998) Modulation of voltage dependent calcium channels by G proteins. Curr Opin Neurobiol 8:351-356. Zhang J, George AL, Jr., Griggs RC, Fouad GT, Roberts J, Kwiecinski H, Connolly AM, Ptacek LJ (1996) Mutations in the human skeletal muscle gene (CLCN1) associated with dominant and recessive myotonia congenita. Neurology 47:993-998. Zuhlke RD, Pitt GS, Deisseroth K, Tsien RW, Reuter H (1999) Calmodulin supports both inactivation and facilitation of L-type calcium channels [see comments]. Nature 399:159- 162. Zwingman TA, Neumann PE, Noebels JL, Herrup K (2001) Rocker is a new variant of the voltage-dependent calcium channel gene Cacna1a. J Neurosci 21:1169-1178. VITA BRANDY ELLEN FUREMAN

EDUCATION Cedar Crest College, Allentown, PA, B.S., Genetic Engineering Technology, cum laude, 1996. Pennsylvania State University College of Medicine, Hershey, PA, Ph.D., Neuroscience, 2001, Ellen J. Hess, thesis advisor. Johns Hopkins University School of Medicine, Baltimore, MD, Neurosciences Trainee (completion of Ph.D. dissertation with Dr. Hess), July 2000ÐJuly 2001.

PROFESSIONAL ACTIVITIES Student Member, American Association for the Advancement of Science, 1999-Present. Student Representative, Society for Neuroscience, Susquehanna Valley Chapter, 1998- 2000. Speaker, Neuroscience Day, Derry Township Elementary School, 1998 & 1999. Student Member, Society for Neuroscience, 1997-Present.

HONORS Trainee, Molecular Basis of Cellular Damage and Toxicity, PHS 5 T32 ES07312, 1997-1999. Beta Beta Beta Biology Honor Society, Cedar Crest College, Theta Psi Chapter, 1994-1996. Cedar Crest College Presidential & Cort Scholarships, 1992-1996.

PUBLICATIONS (Research articles and selected abstracts) Fureman, B.E., H.A. Jinnah and E.J. Hess. A new approach for understanding triggers of episodic neurological dysfunction. Submitted. Jinnah, H.A., S.G. Reich, B.E. Fureman, Z. Kahn, H. Shin, K. Jun, S. Oda, Y. Mori and E.J. Hess. Movement disorders in mice carrying mutations of the CACNA1A calcium channel gene. Submitted. Fureman, B.E., D.B. Campbell and E.J. Hess. 1999. L-Type calcium channel regulation of abnormal tyrosine hydroxylase expression in cerebella of tottering mice. Annals of the New York Academy of Sciences: Molecular and Functional Diversity of Ion Channels and Receptors, 868: 217-219. Fureman, B.E. and E.J. Hess. 2000. Developmental analysis of VDCC subunits, syntaxin, and dystonia in the tottering mutant mouse. Society for Neuroscience Abstracts 26:364. Fureman, B.E. and E.J. Hess. 1999. Chronic administration of the L-type calcium channel agonist BAY K8644 reduces the expression of tottering mouse dystonic episodes. Society for Neuroscience Abstracts 25:722. Fureman, B.E., D.B Campbell, V. Canfield, R. Levenson and E.J. Hess. 1997. Transgenic approach to conditional ablation of murine cerebellar Purkinje cells. Society for Neuroscience Abstracts 23:1877.