Progesterone as an Anti-convulsant

BY

Afshin Shahzamani

A thesis submitted to the Department of Pharmacology

in conformity with the requirements for

the degree of Masters of Science

University of Toronto

Toronto, Ontario, Canada

July, 2001

8 Copyright Afshin Shahzamani, 2001 The author has grauted a non- L'auteur a accordé une licence non exclusive licence allowing the exclusive permettant th la Natioaal Library of Canada to Bibliothèque nationale dr? Cmda de reproduce, loan, distriibute or sell reproduire, prêter, distribuer ou copies of this thesis in microform, vendre des copies de cette thése sous paper or electronic formats. la forme de microfiche/fW, de reproduction sur papier ou sur format blectronique .

The author retains ownersbip of the L'auteur conserve la propriété du top-ght in this thesis. Ne&= the bitd'auteur qui protège cette &&se. thesis nor substantid extracts fkom it Ni la thèse ni des extraits substantieis may be printed or otherwise de celle-ci ne doivent être imphés reproduced without the author's ou autrement reproduits sans son permissim. 8utOIIsation. Acknowledgements There are many people that I'd like to thank for helping me in this long journey.

First, 1 would like to thank Dr. Bumham. His love for life is contagious. He opens doors when others close them. All dong, despite cornpeting responsibilities, he has been helpful.

1 would like to thank the following "lab-relates people who have, in many ways, heiped me complete this project. Some of these individuals have helped more than professional duty requires; some have become friends.

Ted Brown Antonio Mendonca Jerome Cheng Brian Scott Sergei Likhodi Nancy Mingo Janet Eric Law Elizabeth Ng David Lai John Misek Heather Edwards Emmanuel Ho Sue Ambrosini

1 would like to thank Alla Vilner. She cheers me up, always.

1 would like to thank Paula Thiessen who never "doodles all-day-long" when fiiends need her.

1 would like to th& David Peiroff who studies Gravity - we cmsleep a sound sleep, some things are still as they should be.

1 would like to thank my aunt's entire family for having open hearts.

1 would like to thank my sister and her farnily, who gave me a second chance.

1 would like to thank my parents, Ali and Azar, and my brother, Ramin, who have always - without equivocation - believed in my innate abilities and my integrity. Most of all, 1 would like to hank my wife, Andrea Griggs, who has already been there for me and with me, in the tough times, and in the good times. She gives meaning to what is othenvise absurd, and 1 could not have done anything meaningfui without her.

III Progesterone as an Anti-convulsant

Master of Science, 200 1

Afshin Shahzamani

Department of Pharmacology

University of Toronto

Abstract

Progesterone (Pd) has been successfully used as an anticonvulsant. Its use, however, has been limited to patients whose seinires are exacerbated by varying hormonal States in the menstrual cycle. Its anticonvulsant mechanism of action is unclear. Therefore, the following studies were designed to increase the understanding of the mechanism of action of P4 and promote its broader use in a clinical setting: 1) Effect of P, on Amygdala-kindled seisure 2) Effeci of chronic low doses of P, in various seizure rnodels 3) Eflect of gender and age on the anticonvu~santactivity of P44) The role of

GABA and the classic intracelizdar receptor in mediating PI S anticonvulsant action.

These studies indicate that P4 is an etrective anticonvulsant in males and infants. Clinical trials in non-catamenial adults and children would be warranted. Its mechanism of action may involve the newly discovered cell-surface P4 receptor and rnay provide the possibility of new targets for future drug development. %-.. . 1 INTRODUCTION

1.1 Epilcpsy 1.1.1 Definitions

1.2 Types of Seizures 1.2.1 Comrnon Seizure Types

1.3 Therapy for Epikpsy 1.3.1 Antiepileptic D~gs(AED's) 1.3.2 Prognosis for Seizure Control

1.4 Steroid Sex Hormones and Epilepsy 1.4.1 Catamenial Seizurcs 1.4.2 Steroid Sex Hormones: General Background 1.4.3 Progesterone and Estrogen as Ncurosteroids 1.4.4 Synthetic and Metabolic Pathways 1.4.5 Comrnercially Available Progesterone, MPA and Ganaxolone 1.4.6 Sex Steroids and the Brain: Reccptors 1.4.7 Somc Other Notable Neuro-active Steroids

1.5 Animal Seizure Models 1 .S. 1 Electroconvulsivc Shock Seizures: Maximal (MES) and Threshold (ECS) 1.5.2 Pentylenetetrazol: Maximal (MMT) and Threshold (MET) 1-53 Kindling

1.6 Fast studies of Progestin EfCects on Seizures 1.6.1 Clinical Studies 1.6.2 Animal Studies

1.7 Objectives

2 GENERAC METHODS

2.1 Animals

2.2 Drugs

2.3 Procedures for the Kindling Experiments 2.3.1 Implantation of Electrodes 2.3.2 Kindling Procedure 2.3.3 Detemination of AAer-discharge Threshold 2.3.4 Determination of Stability 2.3.5 Procedurc for Drug Testing 2.3.6 Verification of Electrode Placement 2.3.7 Procedure for Ovariectomy and Gonadectomy

2.4 Procedurc for Vaginal Smears

2.5 Proccdurc for the MET Test

2.6 Proccdure for t hc MMT Tcsl 2.7 Procedure for the ECS Test

2.8 Proccdure for the MES Test

2.9 Procedurc for Pentylenetetrazol Infusion

2.10 Procedure for Tests of Sedation and Ataxia

2.1 1 Statistical Analysis

3.1 Rationale Experiment la Experimcnt lb Experiment Ic Experiment Id

3.2 Methods Su bjects Kindling and Threshold Tcsting Dose-Response Testing Seimre Scoring Data Analysis

3.3 Results Sedation Experiment la: Progesterone Dose-Response: Route of Administeration and Vehicle Expriment 1b: Progesterone Dose-Response: Estrogen Priming Experiment I c: Progesterone Dose-Response: tntact subjects Experiment Id: Progesterone dose-response, male rats, hormone replacement

3.4 Summary 75

4 ANTICONVULSANT EFFECTS OF CHRONIC LOW-DOSE PROGESTERONE AND MPA IN SIX SElZURE MODELS 78

4. I Rationale 78

4.2 Methods Subjects Drug Administration Testing and Seizure Scoring Data Analysis

4.3 Results Effects on the Estrous Cycle

4.4 Summary 92

5 THE ANTICONVULSANT EFFECTS OF PROGESTERONE IN ADULT AND 15-DAY-OLD MALE AND FEMALE RATS IN THE MMT AND THE MES SEIZURE MOOECS 94

5.1 Rationalc 94 5.2 Mcthods Dnig Preparation and Administration Testing and Seizure Scoring Data Analysis

5.3 Results 5.3.1 Expcriment 3a: Adult Female Subjects 5.3.2 Experiment 3b: Adult Male Subjects 5 -3.3 Experiment 3c: immature Female Subjccts 5.3.4 Experiment 3d: Imrnahm Male Subjects

5.4 Summiry 107

6 MECHANISTIC STUDIES OF THE ANTICONVULSANT EFFECTS OF PROGESTERONE 309

6.1 Rationale 109

6.2 Methods Subjects Drugs and Dnig Administration Testing and Seizure Sconng Data Analysis

6.3 Results 6.3.1 Experiment 4a: lndomethacin 6.3.2 Experimcnt 4b: RU486

6.3 Summary 113

1. DISCUSSION 118

7.1 The Kindling Data: Experiment 1 118

7.2 The "Chronic, Low-Dose" Data: Experiment 2 120

7.3 The Influence ofgender and Age: Studies at 15-Minute Injcction/rcst Interval 121

7.4 Studies OC Mechanism: Experiment 4 124

7.5 How Does Progcstcronc Stop Seizures? :The Role of Allopregnanolone 125

7.6 How Does Progesterone Stop Seizures? : Non-CABkMediatcd Actions 1 ) Non-Specific Effects 2) Binding to the lntracellular Receptor 3) Other Metabolites 4) Binding to the Membrane Receptor

7.7 How Does Progesteronc Stop Seizures? Th reshold Riscs 129

7.8 How Does Progcsteronc Stop Seizurcs: Clinical Actions 1) Contraceptive effects 2) Anti-estrogen Effects 7.9 Clinical Relevance of the Prcsent Study 132

7.1 O Future Studies 133

APPENDIX 2: NEUROSTEROIDS - THREE IMPORTANT METABOUC ENZYMES 142

A2.1 Aromatasc 142

A2.2 5-alpha reductase 144

2. 3-alpha hydroxystcroid dchydrogenase (3-alpha HSD) 147

APPENDIX 3: THE ROLE OF SEX HORMONES IN SOME IMPORTANT ASPECTS OF NEURONAL PUiSTlClTV 150

A3.1 Somc Behavioral Changes During the Fcrtility Cycle 150

AL2 Neuronal Plasticity and the Fertility Cycle: Cellular Changes 152

A3.3 Neuronal Plasticity and the Fcrtility cycle: Somc Molccular Changes A3.3.1 GABA A3.3.2 Glutamate A3.3.3 Connexin A3.3.4 C- OS A3.3.5 BDNF List of Figures

Figure 1 Metabolic Pathways of Sex Hormones

Figure 2a Systemic Plasma Hormone Concentrations During the Menstnial Cycle in the Human

Figure 2b Systemic Plasma Hormone Concentrations During the 4-Day Estrous Cycle in the Rat

Figure 2c Feedback Control of the Hypothalamus and the Pituitary by Progesterone and Estrogen a Different Times of the Fertility Cycle

Figure 3 The Anti-convulsant Effects of Progesterone in the Kindling Model Progesterone Was Administered IP or SC

Figure 4 The Anti-convulsant Effects of Progesterone in the Kindling Model Progesterone Was Administered in Cyclodextrin and Corn Oil

Figure Sa The EfTect of Estrogen on Aftu-discharge Threshold

Figure 5b The Anti-convulsant Effects of Progesterone in the Kindling Model Subejects Were Estrogen-Primed

Figure 6 The Anti-convulsant Effects of Progesterone in the Kindling Model Intact Subjects Were Used

Figure 7 The Anti-convulsant Effects of Progesterone in the Kindling Model Subjects Were Gonadectomized and Hormone-Replaced Males

Figure 8 The Effect of Chronic Progesterone and MPA on the Genetalized Convulsive Threshold in Amygdala-Kindled Rats

Figure 9 The Effect of Chronic Progesterone and MPA on the Threshold for ECS Seizures

Figure 10 The Effect of Chmnic Progesterone and MPA on the Threshold for MES Seinues

Figure 11 The Effect of Chronic Progesterone and MPA on the Latencies to the Onset of MET Clonic Seizures

Figure 12 The Effect of Chronic Progesterone and MPA on the Latencies to the Onset of MMT Tonic (Forelimb Extension) Seizures Figure 1 3 The Effect of Chronic Progesterone and MPA on the Threshold Dose 92 of Penty lenetetrazol Required to lnduce Clonic Seizures (FLC)

Figure 14 The Anti-convulsant Effects of Progesterone in Adult, Female Wistar 99 Rats

Figure 15 The Anti-convulsant Effects of Progesterone in Adult, Male Wistar 102 Rats

Figure 16 The Anti-convulsant Effects of Progesterone in 1 May-old, Female 104 Wistar Rats

Figure 17 The Anti-convulsant Effects of Progesterone in 1 May-old, Male 105 Wistar Rats

Figure 18 The Influence of Indomethacin (1 Smg/kg) on the Anticonvulsant 116 Effects of Progesterone (1 00mgkg)

Figure 19 The Influence of RU486 (20mg/kg) on the Anticonvulsant Effects of 1 17 Progesterone (1 OOmgkg) List of Tables

Table t Brain Distribution of Sex Hormones and Related Enzymes 23

Table 2 Past Studies on the Anticonvulsant Effects of Progesterone 40

Table 3 The Anticonwlsant EDSOs of Progesterone for the Suppression of 77 Kindled Seizures

Table 4 The Anticonvulsant EDSOs of Progestemne for the Suppression of 107 MMT and MES Seizures List of Abbreviations

micro-Amps

Pl3 microgram 3-alpha HSD 3-alpha hydroxy-steroid dehydrogenase ADT afterdischarge threshold ADT afterdischarge threshold AED antiepileptic dnig AED antiepileptic drugs aMPN anterior medial preoptic nucleus CA ammon's hom CAmg central arnygdaloid nucleus CAMP cyclic adenosine mono-phosphate clMPN caudal parts of the lateral periphery of the preoptic nucleus Cyclodextrin 2 hydroxypropyl-pcyclodextrin Cyclodextrin hydroxypropy 1-beta-cyclodextrin DHEA Dihydroepiandrosterone ElB estradiol benzoate ECS threshold electroconvulsive shock seizure test EDa effective dose (50%) EEG electroencephalogram ER estrogen receptor EUE estrogen receptor elements ERa estrogen receptor alpha

ERP estrogen receptor beta FEBP fetoheonatal estrogen binding protein FLC forelimb clonus FLE forelimb extension FSH Follicular stimulating hormone GABA-A gamma-aminobutyric acid

XII GCT general ized convulsive threshold GD gestational day GDX gonadectomized GnRH generalized releasing hormone GRIP I glucocorticoid receptor interacting protein 1 HLE hindlimb extension HPOA Hypothalamus ICV intracerbralventricular ILS lateral septal nucleus IP intraperitoneal kg kilograrn LH Lutenzing hormone mA milli-Amps MES maximal electroconvulsive shock seizure test MET threshold pentylenetetrazol seizure test mg milligrarn MMT maximal pentylenetetrazol seinire test MPA medroxy progesterone acetate mRNA messenger RNA NDv neurotoxic dose (50%) ovBNST oval nucleus of the bed nucleus of the stria terminalis OVX ovariectomized PdMAmg posterodonal part of the medial amygdaloid nucleus Pa Progcsterone PND Post-natal day PO oral PR progesterone receptors PRA progesterone receptor A PRB progesterone receptor B prBNST principle nucleus of the bed nucleus of the stria terminalis R. maximal response rMPN rostoral portion ofthe medial preoptic nucleus SC subcutaneous SMRT1 silencing mediator for retinoid and thyroid hormone receptors SRCl steroid receptor coactivator 1 StPA stria1 part of the preoptic area TERP- I truncated estrogen receptor product- 1

XIV Chapter 1

Introduction 1 INTRODUCTION

1.1 Epilepsy

1. Definitions

The "epilepsies" are a group of neurological disorciers characterized by spontaneous, recurrent seizures (Engel and Starkman, 1994). ''Seizures" are episodes of excessive neuronal activity that can be identified by characteristic 'spike' or 'spike and wave' abnormalities in electroencephalographs (EEGs) (Burnharn, 1997).

Epilepsy is a very common disorder of the cenaal nervous system (CNS), affecting up to 1% of the population (Berg et al., 1996; Hauser et al., 1991).

Seizures occur in epileptic patients due to the abnonnally low seizure thresholds seen in these patients (Scher, 1997). Al1 brains have the circuitry necessary to produce seizures, and hgsthat biock inhibition or enhance excitation will induce a seizure in anyone (Engel and Starkman, 1994). In non-epileptic people, seizure thresholds are high, however, and spontaneous seizures do not occur. In epileptic patients enhanced excitation - or insufficient inhibition - leads to lowered thresholds and spontaneous seizures (Engel, 1996). 1.2 Types of Seizures

1.2.1 Common Seizure Types

Seizures cm be classified as "generalized, involving the entire brain, or "partial" involving only a part of the brain (Proposal for revised clùùcal and electroencephalographic classification of epileptic seizwes. From the Commission on

Classification and Terminology of the International League Against Epilepsy ,198 1).

There are a number of types of generalized seinires. The two most commonly seen are

"tonic-clonic" and "absence seizures". "Tonic-clonic" seizures involve a loss of consciousness with tonic-clonic convulsions. Constant "spiking" is seen in the EEG

(Swoboda and Drislane, 1994). "Absence" seinires involve a loss of consciousness without convulsions. A pattern of three per second "spike and wave" is seen in the EEG

(Mirsky et al., 1986).

There are only two categones of partial seizures: "simple" partial and "cornplex" partial (So, 1995). "Simple" partial seinves involve abnormal sensations, emotions or movements. They do not involve an impairment of consciousness. EEG "spiking" is seen in neo-cortical or limbic regions (Devinsky et al., 1988). "Cornplex" partial seizures are episodes of confused behavior, during which the patient is out of contact with the environment. Consciousness is said to be "impaired". "Spiking" is unilateral or more commonly bilateral (Gastaut, 1970). Partial seinires of any sort may "generalize" to the whole brain, producing a loss of consciousness and tonicîlonic convulsions. The generalized seizure is then tenned a

"secondarily" generalized seinire (So, 1995).

1.3 Tberapy for Epilepsy

1.3.1 Antiepileptie Drugs (AED's)

Currently, pharmacological treatment is the first line of defense against seizures.

In most cases, it is the best option available (Aiken and Brown, 2000).

Traditional drugs (such as phenytoin, carbarnazepine and phenobarbital) stop seizures by either acting as GABA-A receptor agonists or by blocking calcium or sodium channels (Macdonald and McLean, 1986). These drugs are called "first generation"

AED's in contrast to the mwer drugs which have brrn introduced recentfy. The newer drugs are called "second generation" AED's (Loscher, 1998).

A number of "Second generation" AED's have been introduced in the past decade. In general, they have the same mechanisms of action as the old AED's (Loscher,

1998) (see Appendix 1). Perhaps for this reason, the new AED's have failed to provide significantly greater control of dmg-resistant seizures than the first generation AED's (Loscher, 1998) - although they do have fewer side effects (Rogvi-Hansen and Gram,

1995). Clinical success with non-drug therapies (such as the ketogenic diet) has proved that drug - resistant seizures can be suppressed, and has provided hope that dmgs working through novel mechanisms could be effective against drug-resistant seizures.

A discussion of both the older and newer dmgs that are on the market in Canada is provided in Appendix 1.

1.33 Prognosis for Seizure Control

The majority of patients (about 60%) suffering from epilepsy can control their seinires adequately with current medications (Shorvon, 1996). Approximately 20% of patients attain partial control, and about 20% completely resist drug control (Shorvon,

1996). Patients with hg-resistant seinire are said to have "intractable" epilepsy.

Anticonvulsants that act through novel mechmisms rnay be more effective against intractable seizures, and may lead to fewer and less severe side effects. It is possible that drugs related to steroid sex hormones might provide anticonvulsant compounds that work by a novel mechanism. 1.4 Steroid Scx Hormones and Epikpsy

1.4.1 Catamenial Seizures

The possibility that sex hormones may affect seizure thresholds is suggested by

the phenomenon of catamenial epilepsy.

Seizure fkquency in many wornen is related to the phases of their fertility cycle-

and. thus, to the levels of progesterone and estrogen (Herzog, 1999). Seinires in these

patients are called "catamenial". The prevalence of catamenial seizures is unclear, but, estimates suggest that they occur in ten to seventy percent of the female epileptic

population (Backstmm, 1976; Bauer et al., 1995). The majority of women suffenng fiom catamenial epilepsy have seinves of the complex partial variety (Curnmings et al., 1995).

Until recently, it was thought that catamenial exacerbation was due simply to a high ratio of estrogen to progesterone, estrogen king thought to be proconvulsant and progesterone king though to be anticonvulsant (Backstrorn, 1976). Henog (Herzog et al., 1997a), however, has recently proposed that there are three distinct patterns of catamenial seizures with different hormonal profiles: 1) Pattern One is seen in women whose seizure frequency increases around the time of ovulation. This correlates with a surge in estrogen levels (Herzog et a!., 1997b). 2) Pattern Two is seen in women who have increased seizures during the entire luteal (pst ovulation) phase. During the luteal phase, progesterone (and estrogen) levels are normally high. Women displaying this pattern of seizure frequency. however, rnay have chronically elevated levels of estrogen or may have deficient levels of progesterone (Bonuccelli et al., 1989; Murri and Galli,

1997; Narbone et al., 1990). 3) Pattern Theis seen in women whose seinues get worse during the perimenstd period. This occurs at the end of the luteal phase when the levels of both estrogen and progesterone are are dropping. This pattern may correspond to a period of "progesterone withdrawal" (Herzog et al., 1997b).

Relative to progesterone withdrawai, Smith's group has recently shown that progesterone withdrawal enhances excitation by rendering GABA-A receptors less insensitive to the effects of endogenous GABA (Smith et al., 1998~).

Further evidence that seizure incidence is affected by sex hormones cornes from studies that show seinire onset, or an increase in seizure fiequency, at puberty (Herzog,

199 1). Other studies have shown that seizure fiequency is reduced, or that seimres lose their catamenial pattem, der rnenopause (Harden et al., 1999).

The phenornenon of catamenial epilepsy suggests that estrogen is proconvulsant, and that progesterone is anticonvulsant. If progesterone is anticonvulsant, conceivably it could be used as a treatment for epilepsy. The focus of this thesis is an in-depth pharmacological study of progesterone as an anticonvulsant. 1.4.2 Steroid Sex Hormones: General Background

The steroidal rnetabolic pathway is illustnited in Figure 1. The best-known steroid sex hormones are progesterone and estrogen, which are secreted primarily fkom the ovaries in females, and testosterone, which is secreted primarily fiom the testicles in males. Estrogen and progesterone are not exclusively found in females, however, and testosterone is not exclusively found in males. Low levels of testosterone are found in females (McNatty, 198 1). Likewise, males produce low levels of progesterone and estrogen (Bemti, 1998).

Peptide hormones released from the hypothalamus and the pituitary control the release of steroid sex hormones in the periphery. The hypothalamus releases the peptide gonadotropin-releasing hormone (GnRH) (Reindollar et al., 1985). GnRH controls the release of the peptides follicle-stimulating hormone (FSH) and luteinizing hormone (LH) from the anterior pituitary. FSH and LH control spematogenesis in males, which involves the production of testosterone (Franz, 1988). They also control the development of follicular growth and the corpus luteurn in fernales, which results in the production of progesterone and estrogen (Baird and McNeilly, 198 1).

Steroid sex hormones released in the periphery enter the brain and provide positive and negative feedback to both the hypothalamus and the anterior pituitary

(McNeilly, 1988). Since they have a high partition coefficient, they easily cross the b lood-b rain barrier . - Metabolic Patliways of Sex Hormones 20a hydrocortisol

20a, 22 dihydrocholesterol

20,Pd-lu (Pt- SU)

pregnenolone

Sa rcducuK MI

androsterone

P4SOcI 1 Q (Z Fm) stenediol Ir i P450e11P3(2 FP)

Sa nductasc 2 -5a DH corticosterone P450,I 103 3a hydroxystcroid dehydrogtnasc

corticosterone

P450,I 1 AS corticosterone

aldosterone 1. 1. : 3a hydroxystcroiddchydmgcnase 111111 Hormonal release is relatively constant in males, but it is cyclical in fernales.

Figure 2 demonstrates the variation in levels of progesterone and estrogen during the

female fertility cycle. The fertility cycle in humans is divided into the follicular phase,

ovulation, and the luteal phase. A primary follicle contains an ovum surrounded by

several layen of granulosa and theca cells, which produce estrogen (and some

progesterone) (Baird and McNeilly, 198 1). In the follicular phase, a primordial follicle

grows to maturity - mainly under the control of FSH (Baird and McNeilly, 198 1). Just

prior to mid cycle, increasing LH and FSH secretion cause rapid follicular growth and

begin to alter the function of the granulosa and theca cells fiom the production of

estrogen to the production of progesterone (Fr= 1988). This culminates in a surge in

LH and FSH release, which causes ovulation (Franz, 1988). Ovulation is the expulsion of the ovum fiom the follicle and the conversion of the granulosa cells to lutein cells.

Together, these cells make up the corpus luteum. The conversion is mainly under the control of LH (Cooke, 1988). AAer ovulation, the luteal phase begins. The corpus luteum produces large quantities of progesterone and estrogen (Cooke, 1988). Negative

feedback fkorn progesterone ont0 the hypothalamas leads to Iow levels of LH and FSH.

The end of the luteal phase is marked by the degeneration of the corpus luteum, which results in a sharp drop in the levels of progesterone and estrogen. The degradation of the corpus luteum also causes menstration, and the start of a new ovarian cycle (Franz, 1988). 1.4.3 Progesterone and Estrogen as Neurosteroids

Progesterone and estrogen are gonadai hormones that act peripherally, and also enter the brain to provide feedback. In addition, they are "neurosteroids", which means that they are both produced and catabolized within the brain (McEwen, 1992). The enzymes that synthesize and metabolize sex hormones are widespread in the brain, and are found in many structures outside of those involved in sexual function.

1.4.4 Synthetie and Metabolie Pathways

The synthetic pathways involved in steroidogenesis are shown in figure 1. They are similar in the body and the brain. Progesterone is derived from pregnenolone, which is derived fiom cholesterol (figure 1). Progesterone is metabolized to allopregnanolone

(figure 1 ). Through another pathway , progesterone is also the precursor for testosterone and estrogen. Aromatase converts testosterone to estrogen.

The enzymes involved in the catabolism of progesterone to allopregnanolone are two Cyp450 enzymes: 5-al pha reductase and 3-alpha dehyàrogenase. Allopreganaolone is a potent agonist and an allosteric modulator of the alpha subunit of the OABA-A receptor (Maitra and Reynolds, 1998). Brain distribution and other pertinent details of these enzymes are provided in Appendix 2.

In the periphery, it is primarily the gonads, which synthesize sex steroids.

Catabolism occm primarily in the liver (Feuer, 1983). In the brain, both glia and neurons contribute to steroidogenesis (Zwain and Yen, 1999). Astrocytes are the most active steroidogenic cells. They produce both progesterone and dehydroepiandrosterone

(DHEA). Oligodendrocytes predominantly produce pregnenolones. Neurons predominantly produce estrogens (Zwain and Yen, 1999).

1.4.5 Commercirlly Available Progesterone, MPA and Ganaxolone

Progestins are synthetic compounds that mimic progesterone. Progesterone and a number of progestins are cornmercially available and are used clinically to treat a variety of conditions. Indications include contraception, hormone replacement therapy, amenorrhea, dysfunctional uterine bleeding and endometrial hyperplasia (Apgar and

Greenberg, 2000).

Progestins are categorized according to the time of market introduction or according to structural derivation. Ethynoldiol diacetate and norethindrone are examples of "estranes". Norgestrel, levonorgestrel, desogestrel, gestodene, and norgestimate are examples of "gonanes". Medroxyprogesterone acetate (MPA) is an example of a

"pregnane" (Apgar and Greenberg, 2000). MPA and natural progesterone have been used successfully in clinical trials to

treat catarnenial seimes (Herzog, 1995;Mattson et al., 1984). MPA is available in oral

tablets (Provera@) and in injectable form (Depo-Provera@). Natural progesterone is

orally administered in micronized progesterone tablets (PrometriumQ) and as a vaginal

gel (CrinoneO) (Apgar and Greenberg, 2000).

In the present studies, nadprogesterone and MPA were used. An important

di fference between these is that natural progesterone is metabol ized to allopregnanolone,

which enhances GABAergic activity, whereas MPA is not.

Ganaxolone is a synthetic steroid whose structure mimics allopregnanolone, the progesterone metabolite. Like allopregnanolone, ganaxolone binds to the steroid site of the GABA-A receptor, and increases chloride flux (Carter et al., 1997). Like most

GABA-A-acting drugs, ganaxolone exacerbates absence seizures (Snead, 1998). In clinical trials for epilepsy, ganaxolone has had only mild beneficial effects on intractable seizures (Kemgan et al., 2000;Laxer et al., 2000). Allopregnanolone itself may be usefbl in pnventing the withdrawat effects of progesterone in catarnenial seizures (Reddy and

Rogawski, 2000).

1.4.6 Sex Steroids and the Brain: Receptors Many lines of evidence suggest that sex hormones affect brain excitability

(McEwen, 1999) and seizure thresholds (Woolley and Schwartzkroin, 1998). Most

studies have found that estrogen is excitatory (and proconvulsant) and that progesterone

is inhibitory (and anticonvulsant) (Woolley and Schwaiukroin, 1998).

While the main focus of this thesis is progesterone and its receptors, estrogen and

its receptors will be discussed as well, since the two hormones are intimately related.

Progesterone and estrogen both cause physiological changes in the brain at the molecular

and cellular levels (Smith et al., 1998a;Woolley and Schwartzkroin, 1998) and at the

behaviod level (Steiner, l992), and there is some evidence of functional antagonism

between the two hormones (Brann et al., 1988). There are also instances of synergism

between the two (McNicol, Jr. and Crews, 1979;Williams et al., 198 1). It is difficult to

discuss the effects of one hormone without delving into the effects of the other. It is

possible that some of the anticonvulsant effects of progesterone relate to its functional antagonism of estrogen.

1.4.6.1 Estrogen Receptors

The best-studied estrogen receptors are the intracellular receptors, which mediate gene transcription. There are two types of the "classic" (intercellular) estrogen receptors

(ER'S): alpha estrogen receptors (ERa),and beta estrogen receptors (ERP). The two isofoms are highly homologous in al1 domains except in the amino terminal (Delaunay et al., 2000). Both isofoms are capable of transcriptional activation, although ERP has weaker activity (Delaunay et al., 2000). After binding estrogen, the ER'S homo- or hetro- dimerize and bind to the estrogen receptor response elernents (ERE), enhancing transcription (Delaunay et al., 2000).

Estrogen is involved in the regulation of a multitude of genes, including the induction of progesterone receptors (Blankenstein et al., 1995), c-fos (Giannakopoulou et al., 200 l), glanin (Howard et al., 1997), gonadotropin recepton (Xiong et al., 1994) and oxytocin (Richard and Zingg, 1990). Interestingly, estrogen quickly down-regulates its own receptors (Brown et al., 1996a).

In the rat brain, both ERa and ERP are widely distributed (table 1) (Shughe et al., 1998). In the limbic structures, ERa and ERP mRNA are most abundant in the nuclei of the arnygdala. The lateral nuclei, however, lack both isoforms, and the central nuclei lack ERP. Pirifonn, entorhinal, and isocortex are weakly labeled for ERa, but are more intensely labeled for ERP. The sme pattern is sem in the hippocarnpus, with pater concentration of receptors in the ventral region. It should be noted that Shughure's data contrast with Orikasa's findings that ERa, but not ERB, are present in the hippocampus

(Orikasa et al., 2000). Orikasa, however, used intact rats, whereas Shughure used ovariectomized rats. Within the hippocampus, McEwen's group has found that ER are highly concentrated in the hilus and that they are found only in internewons (Weiland et

al., 1997a).

Gender and age do not seem to affect ER distribution (Orikasa et al., 2000).

Males and females have similar levels of ER'S (in males testosterone is metabolized into

estrogen, which then binds the ER'S) (Brown et al., 1996b). Sex differences in ER

distribution, however, are found in the hypothalamus (Brown et al., 1996b), with females

having higher levels than males. Raab et al. found that in the rnidbrain ERa is present in

both the pre and postnatal periods, whereas ERP is only expressed postnatally (Raab et

al., 1999). ERa and ERP are both found in the brains of infants and children, although

levels of stimulation are low until puberty.

Significant interspecies differences do exist in ER distribution (Osterlund et al.,

2000b). As compared to the rat, primates lack ERa mRNA in the media1 nuclei of the

arnygdala. Primates also lack ERa mRNA in the amygdoloidal areas related to fear

(central and basolateral nuclei). The highest expression of ER in the arnygdala of

primates (including humans) is in the amygdala-hippocarnpal area. the posterior cortex

nucleus, the accessory basal, and the periamygdala cortex. There are significant species differences in the hypothalamus and cerebral cortex also. In the hypothalamus, ERa mRNA levels are very high in the paraventricular and supraoptic regions of primates as compared to rats. In rats, ERP is highly expressed in these regions. In the cerebral cortex, there is no larninar pattern of ERa distribution in rats. In the human, there are

Mem. Mem. Brain Region ER ERa ERP ER PR PRA PRB Aromatase Reâieiase ( HSD 1 Subtbalamic nucleus +" +" Habeaula Media1 (Habenula) -a -u -a Lateral (Habenula) .tu +a +' Hypothalamus +teo +++" *~l.lb +II,IU.~ ++li,m +l 0Lr)t4*1J +++a'1' +l itia.T3 +l J4*4*14 Preoptic area ++a*n ++-ta ++lb ++l

Lateral +& fa +IV

Media1 ++++" +++-ta ++-++a +++" +IY Periventricular +++" ++-ka Pefiventricular nucleus ++"*'o -H-+~ +' -ulb Pqaventricular nucleus ++lu 0 +tu*'L Su)mchiasmatic nucleus ttm +tIo a S9raoptic nucleus +$-tu ++++"em Arpuate nucleus ++' +-+-kW +' +++l ' +Iy Dqrsalmedial nucleus +' +" +" Feral hypothalamus +" ++agir Vqtralmedial nucleus tu -6 utii

Pr(mammil1ary nuclei ++O ++O -t+" Su~ramammillarynuclei +-ka +" uU Tu~romammillary nuc lei -a ++@*IL Mammillary nuclei -O -II Subcomissural organ wu +& +++O

Pituitary .t+r +fU ++l l ++l Basil ganglia Cerebellum ++I

Medulla A

'George, Ojeda, 1982 Laubcr, 1994 14 Compagnone, Mellon, 2000 ' Lephari, 1992 l5 Camacho-Arroyo et al. 1998 ' MacLusky, 1985 " Lauber et al. 1991 ' Sanghera, 1 99 1 " Szabo et al. 2000 'Pelletier, 1994 " Cammancho-Arroyo, 1 998 'ducna- rai ta, et al. ZOO0 'O Shinoda, 1994 :Shughrue et al. 1997 l9Brown et al., 1987 Shughnic et al. 1998 "Lauber, Lichensteiger, 1994 'O brikasa et al. 2000 " MacLusky et al. 1994 " hagihara, et al. 1992 Lauber, Lichtensteiger, 1996 l2 $hughure, Lane, 1997 " Poletti, Martini, 1999 " kiland, Orikasa, 1997 '' Cheng, 1994 low levels of ERa in al1 cortical areas, and layer five of the temporal cortex expresses

higher levels of ERa rnRNA. In the human hippocampus, both ERa and ERP are present

(with ERP being more prominent in the hippocampus than in any other region of the

brain) (Osterlund et al., 2000a). In the rat, however, only ERa is present in the

hippocampus (Orikasa et al., 2000). Osterlund notes that ER distribution in human brains

is in areas involved in emotion, memory and cognition (and also endocrine fiction of

course) and may reflect estrogen's role in these fùnctions. Furthemore, estrogen levels

are significantly different between primates (including the human) and rodents - primates

having much higher levels than rodents. Therefore, estrogens or anti-estrogens may have

differential effects on behavior in rats and humans.

Recently, cell-surface (membrane) receptors for estrogen have also been described

(Ramirez et al., 1996). These receptors do not have transcriptional activity and

presumably mediate changes in membrane excitability (Ramirez et al., 1996). The "fast"

or "non-genomic" effects of estrogen have ken attributed to these membrane receptors

(Wong et al., 1996).

Functionally, at a molecular level. cell-surface estrogen recepton have been found

to: 1 ) activate the adenylate cyclase system (Morozova et al., 1WO), 2) induce the

phosphorylation of the CAMP response element (Zhou et al., 1996), and 3) stimulate

nitnc oxide synthase activity by increasing intracellular calcium concentrations (Baldi et al., 2000;Stefano et al., 2000). Their net effect on ion channels is to: 1) decrease L-type calcium currents (Mermelstein et al., 1996), and 2) increase kainate-induced currents (Gu

et al., 1999). These effects are excitatory in nature (Joels, 1997). The relevance of these

recepton to the effects of estrogen on behavior, or brain excitability, is not well

understood yet.

1A6.2 Progesterone Receptors

Like the ER'S, the "classic" progesterone receptors (PR) are intracellular. They

also corne in two fonns: progesterone receptor A (PR-A) and progesterone receptor B

(PR-B). PR-A is identical to PR-Bexcept that PR-A is amino-terminal muicated

(Giangrande et al.. 1997). PR-A is missing the first 164 amino acids thai are present in

PR-B. The two isoforms are obtained fiom the same copy of the PR gene, but they have distinct estrogen-inducible promoters (Giangrande and McDonnell, 1999).

PR-B is a transcriptional activator. PR-A, on the other hand, lùnctions as a transcriptional repressor (Giangrande and McDonneII, 1999). PR-A inhibits the transcriptional activity of al1 steroid hormone receptor genes. The opposing transcriptional activities of the two isoforms are mediated through distinct signaling pathways (Wen et al., 1994). This is due to differential cofactor binding. PR-A has a higher affinity for the corepressor Silencing Mediator for Retinoid and Thyroid hormone receptors (SMRT). PR-B has a higher afinity for the coactivators Glucocorticoid Receptor Interacting Protein 1 (GRIPI) and Steroid Receptor Coactivator 1 (SRC 1)

(Giangrande et al., 2000).

PR-A and PR-Bare estrogen-inducible, but they also engage in 'cross-talk' by affecting ER transcriptional activity (Katzenellenbogen, 2000). PR-A is a trans-dominant repressor of the transcriptional activity of other sex and stress hormones. This effect involves a noncornpetitive interaction through distinct cellular targets or distinct contact sites within the same target (Wen et al., 1994). The repressor effects of PR-A, including its anti-estrogenic effects, can be induced by progesterone receptor antagonists such as

RU486 (mifepristone) (McDonnell and Goldman, 1994) or ZK98299 (onapristone)

(Vegeto et al., 1993). Hence it appears that these cornpounds have agonist activity with

PR-A.

PR'S are involved in the regulation of many genes. These include the GnRH

(Kepa et ai., 1996) and FSH genes (OIConneret al., 1997), as well as the genes related to

ER'S.

Similar to the ER, PR-A and PR-B have a pattern of brain distribution that extends far beyond the hypothaiamic nuclei involved in sexual function. Within the isocortex, layers two and four express high levels of PR. The cortical nucleus of the amygdala also expresses high levels of PR (Hagihara et al., 1992) (note: Hagihara did not distinguish between receptor sub-types). Both PR-A and PR-B are equally expressed in tne hippocampus (Guerra-Araiza et al., 2000). Within the hippocampus, the pyramidal layers of CA 1 and CA3 fields express high levels of PR (Hagihara et al., 1992). The

normal fiction of these widespread receptors is not clear. They may, however, play a

role in progesterone's anticonvulsant effects.

PR levels change during the estrous cycle. This is probably due to the estrogen- inducible nature of the PR. Therefore, the question arises as to whether estrogen synergizes progesterone's anticonvulsant effects or whether progestemne antagonizes estrogen's proconvulsant effect? The interaction between estmgen and progesterone may be important in progesterone is actions as an anticonvulsant. It is important to note that both PR-A and PR43 expression can be estrogen sensitive or insensitive (Camacho-

Arroyo et al., 1998b). This sensitivity is determined by the region of the brain and not necessarily (but possibly) by the receptor isiform. The PR levels Vary during the estrous cycle in the hypothalamus adthe frontal cortex, but the hippocampal receptors are not affected (Guerra-Araiza et al., 2000). In the hypothalamus, both isoforms are induced by estrogen and down-regulated by progesterone(Camacho-Arroyo et al., 1998a). Estrogen treatrnent does not affect PR mRNA in the amygdala of either male or female rats

(Lauber et al., 1991).

PR'S are found in the brains of both females and males. There are some male/female differences in the distribution and levels, however. These differences are most evident in regions of the brain related to sexual fùnction such as the hypothalamus

(Khisti et al., 1998). There is no significant gender-related difference in estrogen-induced

PR in cortical regions. PR's are also found in pre-pubertal brains. In the cortex, the intracellular PR is detectable at post-natal day one. PR levels rapidly increase, starting on day seven, and reach maximum levels (higher than adult) at day ten. Thereafter, PR expression drops gradually to adult levels. A sirnilar early pattern of expression is seen in the hypothalamic preoptic area, except that levels remain constant after day fourteen (Gee et al., 1988). The ontogenic development of PR's is sirnilar in both sexes (with the exception of particular nuclei mentioned above).

Recently, cell-surface PR's have been discovered. These are distinct fiom the

"classic" intracellular PR's. Little is known as yet about the function of these ce11 surface receptors, although the "fast" effects of progesterone have been attributed to them (Joels,

1997). These include inhibition of purkinje neurons (Smith, 199 1) and stimulation of sperm functions (Calogero et al., 2000). Recently, Momson's group has show that membrane PR'S stimulate tyrosine kinase activity, which results in phospholipase C activation (Momson et al., 2000). In sperm, progesterone enhances calcium currents through its membrane receptor (Foresta et al., 1993). Mile these effects should enhance excitability, membrane PR's are curiously associated with decrease excitability in neurons (Joels, 1997).

1.4.7 Some Other Notable Neuro-active Steroids Progesterone, allopregnanolone and estrogen are not the only neuro-active

steroids involved in the normal and abnormal hction of the brain. Stress hormones,

androgens and pregnenalone cm also alter neuronal excitability (Rupprecht and Holsboer,

1999).

Administration of stress hormones (Le., corticosterone) generally increases

seinire thresholds in both animals (Stitt and Kinnard, 1968) and in humans (Aird 195 1).

Paradoxically, stress can trigger seizws in some epileptic patients (Frucht et al., 2000;

Spector et al., 2000).

Testosterone appears to be proconvulsant (Rosse et al.. 1990). This may be due to

the conversion of testosterone to estrogen (Edwards et al., 1999a).

Pregnenolone sulfate is proconvulsant and dec~asesseinire thresholds (Kokate et

al., 1999; Maione et al., 1992; Reddy and Kulkarni, 1998). Pregnenolone does this in two

ways. Fint, pregnenolone antagonizes GAB A-A receptoa (Majewska et al., 1988;

Majewska and Schwartz, 1987; Mienville and Vicini, t 989). Second, it allosterically potentiates NMDA receptors (Wu et al., 1991).

1.5 Animal Seizure Models The fact that sex hormones affect excitability in many regions of the brain - and

that some of the effects are inhibitory - suggests that sex hormones might be used

pharmacologically to control seizures. The focus of the present thesis was the

pharmacological antagonism of seizures with progesterone. The work was done using

animal seizure models. It was done to supplement past clinical studies, and to provide a basis for fbture clinical work.

Animal seizure models are used in drug development because unproven potential

AED's cannot be tested legally or ethically on epileptic patients. These models should reliably produce seizure activity. They should also be easy to use and have clinical relevance (Fisher, 1989).

A number of animal models are used to study seizures and the epilepsies (Fisher,

1989). Most cornmonly, acute "seizure" models are used. These involve electrical brain stimulation or the systemic administration of various chernical convulsants. In these models, seinws do not occur spontaneously. They occur only when the experimenter appks the epiteptogenic stimuius (Fisher, 1989).

Some investigators have argued that these acute preparations do not fùlly modei clinical epilepsy since the seizures are not spontaneous (Avanzini, 1995). A number of chronic "epilepsy" models - where the seizures occur spontaneously - do exist (Fisher,

1989). Some of these are genetic models, whereas others involve post-status

"spontaneous recurring seizures" (SRS)(Niuinen et al., 2000). While these spontaneous models have some attractive features, they have never been validated pharmacologically. To test the anticonvulsant properties of new compounds, models with known and valid phannacological profiles are necessary.

Therefore, acute, validated models have been used in the present studies. These models are discussed below.

1.5.1 Electroconvulsive Sbock Seizures: Maximal (MES) and Tb reshold

@CS)

The maximal electroconvulsive shock seinw (MES)model is the standard animal model for human tonic-clonic seizures (Woodbury, 1972). High-intensity electrical stimulation is passed through the subject's brain using comeal or pinneal electrodes.

Usually, mice or rats are used. Standard current parameters are 60 Hz sine-wave current for 0.2 seconds at 50 mA in mice or 150 mA in rats (Swinyard, 1972). The seizures begin with a brief period of tonic flexion, followed by forelimb and hind limb tonic extension. Anticonvuhnts that suppress the tonic hindtimb extension in rodents - such as phenytoin and phenobarbital - also suppress tonic-clonic seizures in man (Woodbury,

1972).

Lower-intensity electrical currents - applied in the same manner - produce seizures that are exclusively clonic (Woodbury, 1972). These submaximal seizures are called "threshold" electroconvulsive shock seizures (ECS). Anticonvulsants that suppress forelimb clonus in the "threshold" ECS mode! - such as ethosuximide and valproic acid - also suppress absence seizures in man (Woodbury, 1972).

In administering ECS or MES, investigators should be aware of species, gender and age differences. With regard to species, the threshold for ECS and MES is higher in rats than in mice (Swinyard, 1972).

With regard to gender, females tend to have lower seinire thresholds than males

(Kokka et al., 1992; Sackeim et al., 1987). The ECS threshold for female rats is 16 rnA, for instance, and for males it is 19 rnA (Kokka et al., 1992).

With regard to age, developmental variaiions are particularly dramatic (London and Buterbaugh, 1978). Immediately afler birth, rats have remarkably high ECS and

MES seizure thresholds (over 40 mA for the clonic threshold, over 160 rnA for the tonic threshold). The threshold drops rapidly to base-line levels over the first three weeks of li fe. At day 2 1, the threshold for clonic seizures is approximately 15 mA and for tonic seizures, appximateiy 30 mA. Thereafter, thRsholds rise slightly into adul thood

(London and Buterbaugh, 1978).

1.5.2 Pentylenetetrazol: Maximal (MMT)and Threshold (MET)

Pentylenetetrazol (also called "Metrazol" and "PTZ") is a chemical convulsant, which antagonizes GABA-A receptoa non-selectively (Pellmar and Wilson, 1977). A subcutaneous injection of a high dose of pentylenetetarazol(8Smg/kg in rat) produces forelirnb and hindlimb tonic fiexion and extension in rats. This is the "maximal metrazol seizure" model (MMT)(Fisher, 1989). Like the tonic seizures induced by electrical currents, MMT seizures mimic tonic-clonic seinves in man (Woodbury, 1972).

A lower subcutaneous injection of 7Omgkg in rats (8Smgkg in mice) of pentylenetetrazol produces clonic seizures. This is the "threshold metrazol seinin" model (MET)(Stone, 1972). Like the chicseizures produced by electrical stimulation,

MET seizures mimic absence seizures in man (FerrendeIli et al., 1989).

There are also species, gender and age differences in the MMT and MET models.

With regard to species, rats require slightly lower doses of the convulsant than mice to produce clonic or tonic seinires (Fisher, 1989).

With regard to gender, females have higher seizure thresholds than males (Kokka et al., 1992). Twenty-h~omgkg of tail-vein infused pmtyienetetmot produces clonic and tonic seizures in female rats. The sarne response is seen with 17mgkg in male rats

(Kokka et al., 1992).

With regard to age, two conflicting alterations occur. Infant rats have very high seizure thresholds that drop over the fint three weeks of life. After that they rise again. Curiously, although the thresholds are high in infants, the latency to the onset of seizures

is much shorter in infants than in adults (Velisek et al., 1992).

1.5.3 Kindling

Kindling is a technique for producing focal seizures with secondary generalization

(Goddard et al., 1969;Racine, 1972). in the kindling model, an electrode is pemanently implanted in the brain of in an animal subject (oflen a rat). The subjects are then stimulated at intervals (often daily) with a low level of electrical stimulation. At first, only a localized ("focal") seizwe occurs. With repeated stimulations, however, the seizure activity spreads to other sites (or "generalizes"), as measured by the electroencephalograph (EEG). Eventually, afier motor structures have become involved convulsions occur (Racine, 1972).

When the kindling electrode is implanted in limbic structures, such as the amygdala or the hippocarnpus, the kindling preparation models complex partial seims with secondary generalization (Albright and Bumham, 1980). Albright and Bumharn showed that focal limbic discharge in kindled subjects has a pharmacological profile similar to human complex partial seizure (Albright and Bumham, 1980). They found that anticonvulsants successfully suppressed the generalized component of seizures, but were iinable to raise seizure thresholds in the focus at clinically relevant doses (Albright and

Burnharn, 1980; Albright, 1983). In humans, anticonvulsants also suppress secondarily generalized seizures, but fail to suppress lim bic focal (corn plex partial) activity . Hence.

the kindling mode1 is a pharmacologically valid mode1 of human complex partial

seizwes.

Species, gender, and age also affect kindling. Al1 species that have ken tested

(including primates) kindle (Wada et al., 1978), (Wada, 1981). Seinire thresholds and the rate of kindling, however, Vary with the species. Higher animals, including primates. tend to kindle more slowly (Wada, 198 1).

With regard to gender, differences in kindling rate or seizure threshold between males and females have not been studied. There are, however, reports on the effect of the sex hormones on kindling thresholds and kindling rates. With regard to estrogen.

Edwards et al. found that it lowered seizure threshold and increased the kindling rate

(Edwards et al., 1999b). Buterbaugh also saw a similar increase in kindling rate with estrogen (Buterbaugh, 1987). WahnschafTe, however, did not observe variations in seizure threshold across the estrous cycle (WahnschafTe and Loscher, 1992). With regard to progesterom, Edwards et al. (1999) have found that it raises the seizure threshold and decreases the kindling rate.

With regard to age, Moshe (1998) has shown that rat pups kindle at a faster rate and to higher seizure stages than adult subjects. Kindling threshold in rat pups has not been reported. Recently, however, Edwards et al. (unpublished data) have show that rat pups (fourteen days old) have remarkably high seizure thresholds. 1.6 Past studies of Progestin Effects on Seizures

1.6.1 Clinical Studies

Clinicians have used progestin therapy to control intractable catamenial seizures for more than half a century. Table 2a presents a sumrnary of the case studies and open trails that have used progesterone or medtoxyprogesterone acetate, a synthetic progestin.

Positive results have been reported in al1 studies. Although the earlier studies were only case reports, Herzog (1988, 1995) and Matson's (1984) studies were actual clinical trials.

Matson used medroxyprogesterone acetate (MPA), whereas Herzog used natural progesterone. Signifiant and similar anticonvulsant effects were shown in both studies.

Despite this success, progestins have not found widespread use even in the treatment of catamenial epilepsy. Moreover, no clinician has yet ventured to try them on the non-catamenial epileptic population. Part of the reason for this failure may be that no

"blind" (single or double) clinical studies have as yet been performed to evaluate the anticonvulsant effects of progesterone. Equally lacking are convincing animal studies to support the clinical observations.

1.6.2 Animal Studies A number of past animal studies have investigated the anticonwlsant effects of

progesterone. Unfonunately, many of these have used unreasonably high (sedating)

doses. When lower (therapeutic) doses have been used, the data have been inconsistent.

Table 2b and table 2c provide a summary of these studies. Table 2b summarizes studies

involving high doses of progesterone (over 30 mgkg) and table 2c surnrnarizes studies

involving low doses (under 5 mgkg).

The high dose data (Table 2b) are not contradictory. High doses of progesterone

are clearly anticonvulsant (Table 2b). These anticonvulsant effects cm be seen in the

MET,the MES and the kindling models. The anticonvulsant doses used in these studies,

however, are highly sedative (Table 2b), and they produce blood levels that are well

above physiological levels (Finn and Gee, 1994). These studies do not resemble the

clinical trials in which women have received relatively low doses of progesterone and

progestins (Table 2a), which produce plasma levels within the physiological range. In

clinical trials, patients have not suffered from excessive sedation (Table 2a).

The low dose data (Table 2c) an less cornpletc and consistent. In the kindling

model, for instance, Holmes and Weber (1984) found that progesterone retarded quick

kindling in infants, but not in adult male rats. Edwards et al. (1999) reported that

progesterone retards kindling in adult female rats. Also in this model, Mohammad et al.

( 1998) found that progesterone doses below 75 rng/kg are not anticonvulsant in male rats.

Edwards et al. (1 999), however, found significant anticonvulsant activity at 5 mgkg in female rats. These data seem to suggest that progesterone, at low doses, works in female

rats, but not in male rats.

Wooley and Tirnaras (1 962), testing electroshock threshold, found that low dose

progesterone partially countend the small threshold drops caused by estrogen in

ovariectomized adult female rats. Stin and Kinnard (1 968), however, found that low dose

progesterone in female rats was not protective in the MES model. These data might

suggest that progesterone works against threshold, but not maximal seinues. Once again,

no direct cornparisons have been made.

In the threshold pentylenetetrazol model, dose-response studies have show anticonvulsant effects of progesterone at high doses only . Kokate ( 1 998) did these studies in adult male mice, and Craig (1 966) did them in both male and female mice.

In kainic acid seimres, Nicoletti et al. (1985) found anticonvulsant effects using low doses of medroxyprogesterone acetate. In this model, Frye and Bayon (1 999) have also found anticonwlsant effects using low doses of progesterone. These investigaton, however, have used latency measures. Changes in latency generally imply small changes in seizure thresholds.

To summarize, at low doses of progesterone - which resembte the doses used in clinicai studies - the findings are incomplete and inconsistent. Table 2a: Clinical studies of progesterone (P,) as anticonvulsant

Patients Epilepsy

Zimmennan, Case Generalized Primidone 250mg, 1. No effect; 50 mgkg A.W., et al., report Seinires Phenobarbital 30mg & showcd slight efficacy ( 1973) adult (catarnenial) 1. Povera' ( 10,20,50 2. Complete cessation of woman mg/day seizures for 4 months 2. Depo-Provera Adverse cffects: (250,250,150 mg12 wcek amenonhea

1 intervals) case Generalized 1 90mg Phmobarbital Complete seizure control report ton ic-clonic & over 7 months ' adult seizures Progesterone (0.35mglday) Adverse effects: woman (catamenial) axnenorrhea Open 13 complex Prior medication 39% reduction in seizure R.H., et al., dinical partial, & fiequency , (1984) trial 1 absence Provera (in gpatients, IOmg, (3 patients withdrew fkom 14 adult al1 intractable q2=4d, PO) swdy women Depo-Provera (in 6 Adverse effects: patients, 120-1 SOmg, IM) arncnorrtiea, spotting 8ackstrdm, Open Complex partial Prior medication Significant reduction in T., et al., clinical with one distinct & hzquency of epileptic (1 984) study focus, sclected P4(IV in alcohol & Ringers discharges (4/7 patients) 7 adult based on greater Glucose solution, 0.5-3 mg women than 1 epileptic bolus + 4- 12 mghr infusion, 22-4 3 discharge per 5 achieved luteal phase levels, Y cars minutes of EEG 72nmoVL ) old Herzog, A., Open 68% reduction in seizure ( 1 986) clinical frequency trial paroxy smal 50-400 mg natural P, q2d Adverse effects: transient 8 adult of temporal during times of high seinire t iredness, depression (4/8 women origin, frequency (achieved luteal patients) 1 6-41 catamenial) phase levels, 5-25ng/mL) Y =an old Herzog, A., Open Complex partial 200 mg natural P, lozenges 54% reduction in seizure (1 995) clinicat (with secondary q2d during times of high frequcncy trial generalization, seizure fiequency (achieved Adverse effects: asthenia, 25 adult 13/25) luteal phase levels, S- depression (U25 patients) women (catamenial) 25ng/m L) 18-40 years old Herzog, A., Open Complex partial 200 mg natural P4 lozenges 62% reduction in seizure ( 1 999) clinical (catarnenial) q2d during times of high fiequency trial seizure fiequency (achieved 15 adult luteal phase levels, 5- women 25nglmL) (3 year follow- Table 2b: Anixnai studies involving high-dose progesterone (P,)

Subjects Seizure Mode1 Results

Selye, H., White Pentylenetetrazol T Seizure threshold ( 1942) rats Adverse effects: anesthesia Young 8

Spiegel, E., & White Electroshock 1. o Seizure threshold Wycis, H., rats seizure threshold 2. T ( 1945) e Adverse effects: hypnosis, anesthesia Craig, CR., Sw iss- 1. Penty lenetet P, dose-response: (0- T Seizure threshold (1 966) Webster mol85 10oomgn

Craig, C.R., Swiss- 7 Seinire threshold 8t Deason, Webster EDm=200mg/kg J.R., mice Adverse effects: (1 968) 20-30s acute neuro-toxicity, 8&Q NDso=720mg/kg Selye, H., Holtzma Picrotoxin P, (chronic, 100mg/kg, q2d) ? Seizure threshold (1 970) n rats 3.5mgkg Adverse effects: catatoxic lOOg SC 9 Mohammad, White Kindling P, dose-responsc: ? Seizure threshold only @ S., et al. cats (amygdala) (0,10,30,60,75,90mg/kg) 75 mg/k ( 1998) 260- 5 consecutive Adverse effects: 700g stage 5s miId sedat ion @3 Omg/kg, (3 severe sedation @90mg/kg Kokate, T.G., NIH 1. Pentylenetet 1. P, dose-response: 1. T Seizure threshold et al., Swiss raz01 85 (50-200mg/kg) EDm=94mg/kg (1 998) mice mgkg* SC 2. P, dose-response: 2. ? ED,o=250mg/kg 25-50 g 2. Maximal (1 00-400mg/kg) Adverse effects: sedation 6 electroshock seizure 0.2s, 60Hq 50mA Table 2 C : Animal studies invoiving low-dose ptogesterone (P#j

Subjects Seizure Model Results

Woolley, D., Long Electroshock 1. Chronic E,B 1. $4 Seizure threshold & Timaras, Evans seizure threshold (40,l OOpglkglday) 2. 7' P., ( 1962) Rats (Woodbury 2. Chronic P4 ( Smg/kg/day) 3. 4- lmmatur paradigm, 1952) 3. Chronic &B+P, Adverse effects: e& arnenorrhea in intact adultÇ? mature 9 (OVW

Stitt, S.L.,& Wistar Electroshock 1. P4 (acute, 70mgkg) I . t, Seizure threshold Kinnard, Rats seizure threshold 2. P4 (chronic, Smgkg) 2. t, W.J., 125- 150 (Woodbury 3. MPA (chronic, lOmg/kg) 3. O ( 1968) g paradigm, 1952) 4. MPA+E, (chronic, 4. 4 9 1Om!#kg) Adverse effects: amenorrhea Holmes, Sprague Kindling P4 (5,25 mgkg, hourly & 3 Kindling rate in G.L., & -Dawley (am ygdala) daily) immature rats Weber, D.A., raïs Kindled houriy * in mature rats ( 1984) Immatur and daily Adverse effects: sedation e T (15 @2Smg/kg in immature r & 30 days) & Adult $ (36-70 day s) Nicoletti, F., Wistar Kainic acid MPA (chronic, 2.5 rng/kg) 9, 4 Severity &++latency et al., rats 1 5 0. &t, (1985) 200g SC 8&9

Landgren, S., Cat Penicillin focus & Spontaneous interictal et al. 9 (cerebral cortex) spikes (OW Frye, C., & Long 1. Kainic acid, 1. E,B+P4 & withdrawal; 1. 4 Seizure duration Bayon, L., Evans 32 mg/kg, (GB:lOpg/kg@Ohr, P,: 2. 5. ( 1999) rats SC O.Smg/kg @44hrs, kianic Young 2. Kindling acid @48hrs) Adult (Perforant 2. Silastic implants (aprox. path, for 1 (cholesterol, E&B+P,, 60 days) hr, 0.1 ms, intermittent E$+P4) 9 20 Ht, WX) I2.W) Edwards, H., Wistar Kindling ? Seizure threshold et al. rats (amygda ta) ( 1 999) 200- 15 consecutive 225g stage 5s 3 1.7 Objectives

1. Past expenments from this laboratory have suggested that progesterone is an effective anticonvulsant at a low dose (5mg/kg) in the kindling model of complex partial epilepsy (Edwards et al., 1999b). Adult female rats were used and only a single dose of progesterone was tested. The first objective of these snidies was to perfocm a full pharmacological study of progesterone in the kindling model. Dose-response studies were done. These data are described in Experiment 1.

2. The second objective of these studies was to determine whether low dose progesterone was an effective anticonvulsant in other seizure models, especially when applied chronically.

In addition, MPA was tested. MPA has pharmacokinetic advantages over progesterone, and it is not converted to allopregnanolom (Sturm et al ., 199 1). Thercfore, it was hoped that MPA would provide valuable information about the mechanism of the anticonwlsant action of progesterone. Thesa data are reported in Experiment 2.

3. The third objective of these studies was to perfom dose and time-response studies in both immature rats and in adult males. This was to determine whether progesterone could be used in chilcihoocLepilepsies,or in adult men. These chta are reported in Experiment 3.

4. The forth objective of these studies was to determine whether progesterone's anticonvulsant effects were mediated via binding to progesterone receptors (intracellular or cell surface), or whether they mlated to progesterone's neuroactive metabol ite, al lopregnanolone. Progesterone's anticonvulsant effects were assessed in the presence of RU486 - which blocks the "classic" progesterone receptors - and in the presence of indomethach - which blocks progesterone's conversion to allopregnanolone. These data are reported in Experiment 4. Chapter 2:

General Methods 2 General Methods

Wistar rats (Charles River, Canada) served as subjects. The age and ser of the subjects varied and is specified for each expriment. Subjects were housed in transparent, plastic cages (24x24~45cm) in a vivarium maintained at 18OC on a 12-hour light/dark cycle (lights on at 7:00 8.m.). Adult subjects were housed individually, while immature subjects were housed with their dams. All subjects were given fiee access to food (Purina

Rat Cho*) and water. On test days, subjects were transported to the test room in their home cages and were allowed to acclimatize for 30 minutes prior to experimentation.

2.2 Drugs

Medroxyprogesterone acetate and indomethacin were obtained from a local phmacy. Al1 other dmgs including pmgesterone, estrogm, testostezone, and beta- hydroxy cyclodextrin were obtained fiom Sigma-Aldrich.

2.3 Procedures for the Kindling Expcriments 2.3.1 Implantation of Electrodes

Two weeks (minimum) after amval in the animal facility, subjects were

anesthetized with sodium pentobarbital (Somnitol@, 40mgkg for female subjects,

6Omg/kg for male subjects, IP) and their heads were fixed in a stereotaxic instrument.

Heads were sterilized with 70% alcohol, and an iodine solution. A saggitd incision (1.5

cm) was then made in the scalp, and the superficial muscles were separated from the skull

with Q-tipsm. The skull was washed with normal saline solution, and bregma was

located. Three holes were drilled (each 5 mm fiom bregma) in three quadrants of the skull

and jeweler's screws were inserted to act as anchors for the implant. A hole was then

dded at the implant site and a bipolar electrode was lowered into the right amygdala

using the following coordinates: 1 mm caudal fiom bregma, 4.8 mm lateral from the

saggital suture, 8.5 mm ventral from the dura, The incisor bar was set at +5 mm . The

electrode was cemented in place with a mixture of cranioplastic powder and cranioplastic

liquid (Cranioplastics One Inc., Roanoke, VA).

Etectrodes consisted of two pot yunide@-insulated, stainless wi res (0.25 mm diameter) (MS303f1,Plastic One, Roanoke, Virginia). Subjects were allowed 9 days for post-surgical recovery, and then daily handling was begun.

2.3.2 Kindling Procedure Fourteen days afier surgery, kindling was initiated. Kindling was prefonned in a plastic stimulation cage (width: 40cm, length: 50cm, height: 30cm). Daily stimulations were administered between 12:OO and 6:00 PM. The stimulus consisted of a 1s train of lms, 6OHz biphasic (positive - and negative - going), square-wave pulses, 500 UA (peak to peak), generated by Grass Model S88 stimulator coupled to two Grass PSIU6 constant- current stimulus-isolation units (Grass Instruments Co., Quincy, MA). Convulsive responses were scored according to Racine's (1972) 5-stage scale of motor seizures: stage

1) mouth clonus; stage 2) head nodding; stage 3) forelimb clonus; stage 4) rearing; stage

5) loss of postural control. Electroencephalographic (EEG)activity was recorded for the duration of the afier-discharge, using a Grass Model 7D polygraph. Subjects were kindled until 30 stage 5 seizures had been evoked.

2.3.3 Determination of After-discbarge Threshold

AAer-discharge threshold (ADT) was deterrnined using slight variations of the asceding series technique (Pinel et ai., 1976) or half-split technique (Racine, 1972). The ascending series technique was used in Experiment 1. The half-split technique was used in Expenment 2. ADT was detemined before kindling began and also after the subjects had had 5 stage 5 seinires.

The ascending series technique of threshold measurement involves the initial administration of a sub-threshold electrical stimulus followed by subsequent, step-wise increases in the electrical stimulus separated by a short "inter-stimulus interval". In the present study, ADT determination began with a stimulus of 20 UA (peak to peak). The stimulus intensity was then increased in 20 UA steps up to 320 UA; then the intensity was increased in 40 UA steps. A 1 -minute inter~alwas allowed between stimuli. ADT was defined as the minimum stimulus intensity needed to provoke an after-discharge of at least 4 seconds in non-kindled subjects or a generalized seinire in well-kindled subjects.

The half-split technique is also called the "up and dom" method (Kokaia et al.,

1994). The electrical current-interval between a stimulus that produces seizures and one that doesn't is halved to obtain a new current-interval. This procedure is continued until the threshold is deterrnined within an established rnargin of error. Long inter-stimulus intervals are used. In the present shidy, the ADT was detemined over several days. On the first day, a current of 100 UA was used. Depending on whether this current induced a seizure (after-discharge of at least 4 seconds in non-kindled subjects or a generalized seizure in well-kindled subjects), the current intensity was halved or doubled on the second day. This process was continued until the interval between the lowest effective stimulus and the highest ineffmtive stimulus was 20 UA or less. ADT was defined as the halfway point within this interval.

2.3.4 Determination of Stability ABer the ADT was determined, seizure stability at ADT plus 40% was assessed.

Subjects were stimulated daily at a curent intensity equal to ADT plus 40%. 'Stability' was defined as 5 consecutive stage 4 or 5 seizures. If subjects were not stable, the stimulus intensity for kindling was increased by an additionai 40%.

2.3.5 Procedure for Drug Testing

The procedure for drug testing is specified for each expriment in the specific methods section.

2-3.6 Verification of Electrode Placement

Once kindling experiments were cornpiete, subjects were anesthetized with sodium pentobarbital (SomnitolQ, 80mgkg) and perhsed. After loss of consciousness, the site of the implantation was lesioncd by a cwcnt of 8mA,lasting 30 seconds. The thoracic cavity was then opened via a medial incision, exposing the heart. A cut was made in the right auricle and a cannula (gauge 16) was inserted into the lefi ventride.

Perfùsion began with 1OOmL of normal saline. The perfusion rate was 1 SmLlminute.

Perfusion was completed with lOOmL of 4% formaldehyde solution (9g NaCI, 108g 37% formaldehyde, in 1 L of distilled water). Brains were then extracted and stored in 20% sucrose solution containing 4% fomaldehyde. Forty-eight hours later, the brain was withdrawn From the fixative and stored until sectioning at -80°C in a sealed plastic container.

At the time of sectioning, the brains were attached to a specimen disc using ernbedding medium (Tissue Tek@, Sakura Finetechnicd Co., Ltd., Tokyo). They were then transferred to a cryostat (Jung CM3000, Leica Inc., Canada), maintained at -2S°C, and allowed to equilibrate for 20 minutes. Sections (30um)were cut coronally. Every fifth section in the area of the electrode tip was thaw-mounted on poly-L-lysine coated glass slides and allowed to air dry.

Dried sections were stained 6thcresyl violet. Slides were immersed for 3 minutes in each of each of the following media: 100% ethanol, 95% ethanol, 70% ethanol, distilled water, 0.5% cresyl violet, distilled water, and 70% ethanol. Slides were then incubated for 2 minutes in 95% ethanol containing 0.1% acetic acid, and then twice in 100% alcohol. Finally, slides were incubated twice for 5 minutes each in xylene. The slides were then covered using a cover slip and Permount@ (Fisher Chemicals).

The electrode tract was examined with a magnification-projection system

(Research Analysis System mode1 42 125 1, Amersham, MI). Only subjects with verified electrode placements in the targeted region were used in subsequent data analyses.

2.3.7 Procedure for Ovariectomy and Gonadectomy Ovariectorny and gonadectomy were performed on some of the subjects in the kindling experiments.

Twenty-four hours after the 10th stage 5 seinues, female subjects were ovariectomized. Subjects were anesthetized with 4% halothane and then anesthesia was rnaintained using 1.5% halothane. Bilateral flank incisions (1 cm) were made. The uterine horn attached to the ovary and the accompanying fat were pulled âhrough the incision. The ovary and the accompanying fat and vessels were tied (5-0 silk suture) and the oviduct severed and removed with a single cut. The uterine hom was returned to the peritoneal cavity. The muscle and skin incision were closed with one and two sutures, respectively. Subjects were allowed a 14-day postsperative recovery period before funher procedures.

Male subjects were gonadectomized at the time of electrode implantation. Under general anesthesia, a single, saggital incision (1 cm) was made in the scrotal sac. The testes and epididiymes was pulled through the incision. Both tissues were severed and removed by a single cut, following suturing of the tissue above the cut. The muscle and skin incision were closed with one and two sutures respectively. Subjects were allowed

1Cday post-operative recovery period before fùrther procedures.

2.4 frocedure for Vaginal Smears In some experiments, daily vaginal smears were performed (between 10:OO h and

10:30 h) to characterize each day of the 4-day estrous cycle in the intact female rat

(Freeman. 1996). Smears were obtained by placing a blunted, plastic 20 pl Eppendorf tip on the vaginal opening and flushing the orifice twice with 10 pl of 0.9% saline. Tips were rinsed thoroughly with 70% ethanol dereach use. The saline flush was displayed on a clean microscope slide and observed at 20X magnification. Phases of the estrous cycle were classified according to the following criteria: 1) diestrus - smears were characterized by a heterogeneous mixture of polymorphic leukocytes and epithelial cells;

2) proestrus - smears had predominantly large round nucleated epithelial cells; 3) estrus - smears consisted primarily of comified epitheliurn with few or no nuclei visible, and 4) metestrus - smears with exclusively leukocytes in high density.

The threshoid pentylenetetrazol test (MET)was administered according to a modification of the protocol of Kra11 et al. (Wl a al. 1978). This test was used in

Experirnent 2 on adult female Wistar rats. Subjects were injected with 70mgkg of pentylenetetrazol (SC) in the rough of the neck. Pentylenetetrazol was dissolved in normal saline to a concentration of 19 mgkg. The volume of injection was 4mLkg.

Following injection, the subjects were placed in a test chamber and observed for 30 minutes. Seinires were scored as "present" or "absent". The latency to the onset of seinires was also recorded. Seizure absence was defined as the absence of forelimb clonus (FLC).

2.6 Procedure for the MMT Test

The maximal pentylenetetrazol test (MMT)was administered according to a modification to the protocol of Desmedt et al. (Desmedt et al.. 1976). This test was used in Experiments 2,3, and 4. Adult subjects were in injected with 85mg/kg of pentylenetetrazol (SC) in the rough of the neck. Pentylenetetmzol(2lmg/mL) was dissolved in normal saline. The volume of injection was 4mUkg. Rat pups were injected with 150mgkg of pentylenetetrazol (SC) in the rough of the neck. Pentylenetetrazol

(1 5 rng/mL) was dissolved in normal saline. The volume of injection was 1OmLkg.

Following injection, the subjects were placed in a test charnber and observed for

30minutes. The seizures were scored as "present" or "absent". The latency to the onset of seizures was also recorded. Seizure absence was defined as the absence forelirnb extension (FLE). Fotelimb extension was measure as opposed to hind-limb extension

(HLE), because subjects did not produce hind-limb extensions reliably in this model.

2.7 Procedure for the ECS Test The threshold electroconvulsive shock test (ECS) was administered according to a

modification to the protocol of Woodbury (1972). This test was used in Expriment 2 on

adult female Wistar rats. The ECS stimulus was administered via corneal electrodes.

Electrodes were wetted with normal saline prior to application to the comea. The

following parameten were used for stimulation: pulse codiguration - sine wave, pulse

fiequency-60 Hz, train duration-0.2 seconds, pulse intensity - variable (1 0-50 mA).

Variable pulse intensity was used because the threshold for ECS seinires was being

measured. To determine ECS threshold, the half split method (as described previously)

was used. The seinires were scored as "present" or "absent". Seinire absence was

defined as the absence of forelimb clonus.

2.8 Procedure for the MES Test

The maximal electroconvulsive shock test (MES)was administered according to a modification to the protocol of Krall et al. (Kra11 et al. 1978). This test was used in

Experiments 3 and 4. The ECS stimutus was administercd via corneal etectrodes.

Electmdes were wetted with normal saline prior to application to the comea. The

following stimulus parameters were used: pulse configuration-sine wave, pulse fiequency

- 60 Hz, train duration-0.2 seconds, pulse intensity - 150 mA. Seizures wete scored as

"prcsent" or "absent". Seizwe absence was defined as the absence of hindlimb extension (HLE). In Experiment 2 the threshold for MES was detemined in adult female Wistar

nits. The same procedure was followed as in the standard MES test except that variable

pulse intensity (1 0-100mA) was used. To detemine this threshold, the half split method,

as described previously, was used.

2.9 Procedure for Pentylenetetrazol Infusion

The pentylenetetrazol infusion test allows accurate threshold measurements of the clonic or tonic seizures resulting fiom the administration of pentylenetettazol. This test

was used in Experiment 2 in adult female Wistar rats. It was administered according to the protocol of Thavendiranathan et al. (2000). Pentylenetetrazol was dissolved in normal saline to a concentration of IOmgImL. The mixture was administered through the tail vein at an injection rate of 1 mL/minute. A Sage Instrument synnge purnp (mode1

351) was used for administration. The infusion was stopped when tonic seizures were produced. The time to the onset of tonic seizures was recorded.

2.10 Procedure for Tests of Sedation and Ataxia

In al1 experiments, subjects were challenged with the righting-reflex test and

Loscher's ataxia rating scale immediately pnor to testing. The righting reflex test involves rolling the subject ont0 its back and observing whether it is able to rotate 180 degrees ont0 its limbs. Loscher's test of ataxia involves simple observation, and uses the following 6- stage catagones of behavioral impairments:

1. Slight ataxia tottering of hind limbs

2. More pronounced ataxia dragging of hind limbs

3. Further increase of ataxia more pronounced dragging of hind limbs

4. Marked ataxia animal occasionally loses balance during forward

locomotion

5. Very marked ataxia animal fiequently loses balance during forward

locomotion

6. Extreme ataxia animal, despite attempts, is unable to move fonvard

2.1 1 Statistical Analysis

Time-response and dose-respome cuves were plotted for both the clonic and tonic seinues. Quantal curves were constructed by ploning the percentage of animais at each time or dose that were protected against the clonic or tonic seizures. Whenever possible, dose-response curves were fitted with sigrnoidal curves and using linear regression, and ED& were calculated fiom the fitted curves. Interval-ratio data (e.g., latencies) were expressed as mean standard error of the mean (SEM). When two groups were king compared and the data were normally distributed, a student t-test or a paired t-test was used. When the data were not normally distributed, the Mann-Whitney Rank Sum test was used. When more than two groups were being compared and the data were normally distributed, analysis of variance (one- way ANOVA) was used. Provided that the overall analysis of variance was significant, post-hoc cornparison used Duncan's Multiple Range test. When the data were not normally distributed, the Kniskal-Wallis one-way analysis on ranks was done. A cntical significance level of P < 0.05 was used for al1 tests. Chapter 3

Anticonvulsant Effects of Progesterone

in Amygdala-kindled Rats 3 Anticonvulsant Effects of Progesterone in Amygdala-kindled Rats

3.1 Rationale

In small, open clinical triais, progesterone has proven to be an anticonvulsant.

Women with catamenial seizures, who are refiactory to other treatments, consistently respond to progestemne (see table 2a). These patients generally suffer fiom intractable complex partial seizures. Complex partial epilepsy is the most common adult fom of epilepsy, and it is often intractable (see Introduction).

The kindling preparation provides an animal mode1 in which to study the effects of progesterone on complex partial seizures (Introduction). Past studies on the anticonvulsant efTects of progestemne in the kindling mode], however, have produced conflicting results. Edwards et al. (1 999) found that 5 mgkg of progesterone abolished amygdala-kindled seinires in 60% of adult female Wistar rats. Mohammad et al. (1998), however, fond that in male Wistar amygdala-kindled rats, doses bel ow 75 mgkg were ineffective at blocking generalized kindled seizures. Edwards did not perform dose- response studies, whereas Mohammad did. These data seem to suggest that progesterone may have low-dose anticonvulsant effects in femaies but not males.

Expenment 1 was designed to clarify the effects of progesterone on kindled seizures. Dose-response studies with progesterone were perfonned in amygdala-kindled female and male Wistar rats. ED,'s were determined, and the influences of endogenous and exogenous hormones on the response were explored. It was originally hypothesized that progesterone would be anticonvulsant at low doses (5 mgkg), at least in fernales.

In an attempt to replicate the procedure of Edwards et al., adult Wistar rats were used, and progesterone was injected 2 hours before seizures were triggered. It was assurned that the responses would be genomic and rnediated by the classic progesterone receptor. Genomic responses take some time to develop.

When initial trials failed to show low-dose progesterone effects (data not shown), a number of subsequent trials were performed, varying different expenmental parameters

(The same ovariectomized subjects were used in Experiments la and 1b).

Experiment la

In Experiment la, parameters related to administration were manipulated. To assess the possible importance of first-pass effects, the route of delivery was varied. The drug was administered IP (first pas effects) or SC (no first pass cffkct). To assess the possible effect of the vehicle, the drug vehicle was varied. The drug was administered either in beta-hydroxy cyclodextrin or corn oil (Edwards et al. had used oil as their drug vehicle). Experiment 1b

In Experiment 1b, the possible role of estrogen was investigated. The effect of

"estrogen priming" on after-discharge threshold (ADT) was determineci. Subjects were then "estrogen-primed" before tests of the anticonvulsant effects of progesterone.

Experiment le

In Experiment 1c, to determine effect of endogenous hormones (including estrogen) a new group of intact (not ovariectomized) subjects were used. This allowed the evaluation of anticonvulsant effects of progesterone within a normal hormonal milieu

(Edwards et al. used intact subjects). (Ail other kindling experiments used castrated subjects to prevent the confounding effect of endogenous hormones when testing progesterone.)

Experiment Id

In Experiment 1 d, castrated male subjects, with and without hormone replacement, were used to provide a cornparison with the fernale subjects in Experirnents la, lband fc.

3.2 Methods Subjects

The same subjects were used in Experiment la and 1b. They were 33 right amygdala-implanted, ovariectomized female Wistar rats. In Experiment 1c, the new subjects were 24 right amygdala-implanted, intact (not ovariectomized) female Wistar rats. In Expenment Id, the subjects were 36 right amygdala-implanted, gonadectomized and hormone-nplaced, male Wistar rats. Hormone replacement consisted of SC implantation of Silastic@capsules containhg cholesterol, estrogen or testosterone.

Subjects were housed as described in the General Methods.

Kindling and Threshold Testing

The subjects (in al1 experiments) were kindled daily to a critenon of 30 stage 5 seizures (procedure as in General Methods). Their thresholds were measured using the ascending series rnethod as described in the General Methods. Thresholds were checked for stability as described in the Generai Methods.

Dose-Response Testing

Four days fottowing the last kindling stimulus (stability testing), progesterone dose-response studies were initiated. In al1 studies, the following doses of progesterone were used: 4,8, 16,3 2, and 64 mgkg. For all expenments (except l a, when corn oil was used as vehicle), progesterone was dissolved in beta-hydroxy cyclodextrin to a concentration of 32mgfmL for the highest dose, followed by seriai dilution for lower doses. It was adrninistered in a volume of 2mL/kg. In part of Experiment 1b, the drug (progesterone) was prepared in a similar mawras above, except that corn oit was used as the vehicle.

In the first part of Experiment la, the subjects were randomized and injected SC or IP with different doses (4,8, 16,32,64 mgkg) of progesterone. Two hours after drug injection, the anticonvulsant action of progesterone was tested by administering a kindling stimulus equivalent to ADT+40%. The same subjects were tested repeatedly with different doses. A four-day interval was allowed between tests. It permits complete clearance of progesterone.

In the second part of Experiment la, the same subjects (ovariectomized fernales) were randomized and injected SC with different doses of progesterone dissolved in corn oil or beta-hydroxy cyclodextrin. Seizure testing was the same as described above.

In the first part of Experiment lb, the same subjects (ovariectomized females) were "estrogen primed" with a SC injection of I Ougkg estradiol benzoate 48 hours prior to ADT testing. Estradiot was dissolved in corn oit, and injected in a volume of ZmL/kg.

ADTs were detemined (ascending series method as described in General Methods) before and after "estrogen priming".

In the second part of Experiment lb, a progesterone dose-response study was then perfonned in "estrogen primeà" subjects (ovariectomized females). Forty-six hours after estrogen priming, they were randomized and injected SC with different doses of progesterone. Seizure testing was identical to Experiment la, except that a newly found

ADT + 40% was used. It was based on the ADT measured in the presence of estradiol.

In Experiment le, new subjects (intact females) were randomized and injected

SC with different doses of progesterone. Seimtesting was identical to Experiment la.

It should be noted that the intact subjects were not cycling and exhibited persistent

estrous. The kindling stimulus cause female rats to cycling anest and the subjects

perpetually appear to be in the proestrous/estrous phase. (Edwards et al., 1999).

In Experiment Id, adult mate subjects were (gonadectomized and hormone-

replaced males) with cholesterol (control), estrogen or testosterone. They were then

randomized and injected SC with different doses of progesterone. Seinire testing was

identical to Experiment la.

Seizure Scoring

Two compomnts of the kindled seizure were scored: the generalized convulsion

and the focal component. Each component was marked as "present" or "absent".

Generalized seizures were defined as stage 4 and 5 seizures, as catagorized by Racine's classification (General Methods). The generalized seiwe was scored as absent if seinire stages 4 and 5 were suppressed. The focal seizure was defined as 3 or more seconds of

"spiking" in the EEG record. This component was scored as absent, if a normal EEG followed the stimulation. These definitions are consistent with Albright's scoring technique (Albright and Bumharn, 1980).

Data Analysis

Dose-response curves were plotted for progesterone's effects on both the generalized and focal components of the seizures. Quantal curves were constmcted by plotting the percentage of animals that were protected at each dose. ED,o's and Rmax's were calculated. A paired Student t-test was used to compare ADT's in Experiment 1b.

3.3 Results

Sedation

Sedation was generally not apparent for doses of 32rngkg or less. A dose of

64mgkg in the SC, cyclodextrin groups fiom experiments 1a and 1 b (data are combined), however. caused significant sedation, including loss of the righting reflex in 36 % of the subjects (N=14). Ataxia was measured on the Loscher scale (General Methods) and found to be 4.9 +/- 0.4 in these groups. This was significantly different from control subjects (P

Experiment la: Progesterone Dose-Response: Route of Administeration and

Vehicle Figure 3 shows the progesterone (injected IP or SC) dose-response curve for the suppression of generalized and focal seizures in well-kindled, ovariectomized, female

Wistar rats. For generalized seizures in both the IP and the SC treated groups, the ED, was 33 mgkg and the maximal response (100%) occurred at 64mgkg. The route of administration did not matter. SC administration was therefore used in al1 subsequent studies. Dose mglkg

4 8 16 32 64 Dose mgkg

Fiaure 3: The Anti-convulsant Effects of Proaesterone in the Kindlina Model Proclesterone Was Administered IP or SC The anticonvulsant effects of progesterone against generalized kindled convulsions (closed circles) or focal seizures (open circles) Each point represents data from at least 7 subjects. Subjects were adult, ovariectomized female Wistar rats. Figure 3a: progesterone was administered IP. Figure 3b: progesteorne was administered SC. No substantial suppression of focal seizure activity was seen at any of the doses

tested in either group. No low-dose anticonvulsant effects were seen.

Figure 4 shows the progesterone (dissolved in cyclodextrin or corn oil) dose-

response curve for the suppression of generalized and focal seinires in well-kindled,

ovariectomized, fernale Wistar rats. For generalized seinires in the beta-hydroxy

cyclodextrin treated group, the ED,, was 33 mgkg and the maximal response (100%)

occuned at 64mgkg. There was no substantial suppression of generalized seizures in the

corn oil treated group. Cyclodextrin was, therefore, used in al1 subsequent studies.

No substantial suppression of focal seizure activity was seen at any of the doses tested in either group. No low dose progesterone effects were seen.

Experiment 1b: Progesterone Dose-Response: Estrogen Priming

Figure 5a shows the effect of "estrogen-priming" on the ADT as measured by the axending series method (General Methods). The same subjects were used as in

Experiment la. The ADT was significantly reduced by 23% (P < 0.00 1, using t-test).

Thus, estrogen priming lowea seinire thresholds at the seimre focus. In the subsequent 4 8 16 32 64 Dose mgkg

l Progesterone , corn oil

Dose mgkg

Fiaure 4: The Anticonvulsant Effects of Frocresterone in the Kindlina Model Prociesterone Was Administered in Cvclodextrin or Corn Oil The anticonvulsant effects of progesterone against generalized kindled convulsions (closed circles) or focal seizures (open circles) Each point represents data from at least 7 subjects. Subjects were adult, ovariectomized female Wistar rats. Figure 4a: progesterone was administered in cyclodextrin. Figure 3b: progesteorne was administered in corn oil. The Effect of Estrogen on Afterdischarge Threshold (ADT)

control "estrogen-prirned"

Fiaure 5a: The Effect of Estroaen on After-discharae Threshold The proconvulsant effects of estrogen on after-discharge threshold. Data are represented as mean +!- standard error. Each group contains 33 subjects; C) indicates a signifiant difference from the control group using a paireâ t-test (Pc0.001).

Progesterone, "estrogen-primed"

4 8 16 32 64 Dose mgkg

Ficlure Sb: The Anticonvukant- Effects of Proaesterone in the Kindlina Model Subiects Were Estroaen-Primed The anticonvulsant effects of progesteione against generalized kindled convulsions (closed circles) or focal seizures (open circles). Each point represents data from at least 5 subjects. Subjects were adult, ovariectomized (estrogen-primed) female Wistar rats. dose-response study (Figure Sb), estrogen primed subjects were tested at their newiy measured ADTYs+40%.

Figure Sb shows the progesterone dose-response curve for the suppression of generalized and focal seizures in well-kindled, ovariectomized, "estrogen-primed" female

Wistar rats. For generalized seizures in the "estrogen primed" subjects. the ED,, was

40mg/kg and the maximal response (80%) occurred at 64mgkg. No low-dose anticonvulsant effects were seen. Therefore, estrogen does not synergize progesterone's anticonvulsant effect in the kindling model.

No substantial suppression of focal seinire activity was seen at any of the doses tested.

As a positive control for the testing method, a dose-response of phenobarbital was prefonned (data not shown). The ED,. against generalized seims was 12.5 mg/kg.

Experiment le: Progesterone Dose-Respoose: Intact subjects

Figure 6 shows the progesterone dose-response curve for the suppression of generalized and focal seizures in well-kindled, intact, female Wistar rats (These subjects were not the sarne as those used in previous studies). For generalized seizures in the intact subjects, the ED, was 30mglkg and the maximal response (100%) occurred at Progesterone, "intact"

4 8 16 32 64 Dose mglkg

Fiaure 6: Anticonvulsant-- Effects. of Proaesterone in the Kindlina Model Intact Subiects Were Used The anticonvulsant effects of progesterone against generalized kindled convulsions (closed circles) or focal seizures (open circles) Each point represents data from at least 5 subjects. Subjects were adult, intact (not ovariedomized) fernale Wstar rats. 64mgkg. No low-dose anticonvulsant effects were seen. Therefore, normal levels of endogenous hormones do not alter progesterone's anticonvulsant effect in the kindling model.

No suppression of focal seinire activity was seen at any of the doses tested.

Experiment Id: Progesterone dose-response, male rats, hormone

replacement

Figure 7 shows the progesterone dose-response curves for the suppression of generalized and focal seizures in gonadectomized, cholesterol-replaced, estrogen- replaced and testosterone-replaced, well-kindled, male Wistar rats. For generalized seizures, the ED, was 35mg/kg in the cholesterol-replace subjects, 37mgkg in the estrogen-replaced subjects, and IO 1mgkg in the testosterone-replaced subjects. The maximal response (90%) occurred at 64mglkg in the cholesterol-replace subjects, and abin the estrogen-replaced subjects. The maximal response (60%) occurred at 98mg/kg in the testosterone-replaced subjects. No low-dose anticonvulsant effects were seen.

Hence, the anticonvulsant effects of progesterone in gonadectomized and cholesterol- replaced or estrogen-replaced rats are similar to the response in femaie rats. Testosterone in male rats, however, makes them less responsive to the anticonvulsant effects of progesterone. I /a o. -v

4 8 16 32 64 Dose mghg

100. Progesterone, estrogen-replaœd

cn 80. 2 a .Y u 60- Y> C 3 O r.r 40 3 * 20.

0

1 I b 1 b 4 8 16 32 64

Oose rngiûg

Progesterone, testosterone-replaced

Figure 7: The Anticonvulsant Effect of Progesterone in the Kindlina Model Subjects Were Gonadectomized and Hormone-Replaced Males The anticonvulsant effects of progesterone against generalized kindled convulsions (closed circles) or focal seizures (open circles) Each point represents data from at least 7 subjects. Subjects were adult , ovariectornized female Wistar rats. Figure 7a: progesterone was administered in cholesterol-replace subjects. Figure 7b: progesteorne was administered in estrogen-replaced su bjects. Figure 7c: progesteorne was administered in testosterone-replaced subjects. 75 No substantial suppression of focal seinire activity was seen at any of the doses tested in any of the groups.

3.4 Summary

Progesterone was an effective anticonvulsant in female rats, but at doses that were higher than expected. ED,s were usualiy in the range of 30-40mglkg (summarized in

Table 3). Manipulations of the route of drug delivery, the dnig vehicle, and of exogenous

(estrogen priming) and endogenous (intact subjects) hormonal States did not irnprove progesterone's anticonvulsant response. Estrogen prirning, however, lowered seizure thresholds, thus supporting the theory that estrogen is an important cause of catarnenial seizures of the complex partial vmiety.

Castrated males (cholesterol or estrogen replaced) had ED,s similar to the fernales. Testosterone-replaced castrated males, however, had high EDs,,s, suggesting that testosterone blocks the anticonvuf sant effects of progesterone.

The original hypothesis that progesterone would have anticonvulsant effects at low doses was not supported. Tabfe 3

The anticonvulsant EDSO's of progesterone for the suppression of Amygdala-generalized convulsions

-

Route of Vehicle Sex Hormonal Parameters Amygdala Administeration EDSO (mg/kg) - -

Cyclodextnn Female Ovariectomized

Cyclodextrin Female Ovariectomized

Corn oil Female Ovariectomized

Cyclodextrin Female Ovariectomized / Estrogen-replaced

Cyclodextrin Female Intact

Cyclodextrin Male Gonadectomized / Cholesterol-replaced

Cyclodextrin Male Gonadectomized / Estrogen-replaced

Cyclodextrin Male Gonadectomized / Testosterone-replaced Chapter 4:

Anticonvulsant Effects of Chronic Low-Dose

Progesterone and MPA in Six Seizure Models 4 Anticonvulsant Effects of Chronic Low-Dose Progutvone and MPAim Six

Seizure Models

4.1 Rationale

The results of Experiment 1 indicated that progesterone is not anticonvulsant at

low doses in amygdala-kindled rats. These findings are consistent with sirnilar studies by

Mohammad's group who also found suppression of kindled convulsion ai high doses

(Moharnrnad et al., 1998b).

The fact that only high doses - which caused sedation - were effective suggests a

GABAergic mechanism of action for progesterone's anticonvulsant effects. These

findings. however, are not consistent with the clinical data, which indicate that low doses

of both progesterone and MPA (which is not metabolized into ailopregnanolone) are effective anticonvulsants.

One important difference between Experiment 1 study and the clinical studies is

that we administered progesterone acutely, whereas epileptic patients use progesterone or

MPA chronically. It seemed possible that low doses of progesterone or MPA might be anticonvulsant in rats if they were administered chronically (for 1 week).

The effects of chronicaily administered progesterone and MPA were therefore assessed in several different seizure models. In each model, seinire latency or threshold - rather than seinire absence or presence - was measured, so that even srnall changes would be seen. The tests included amygdala kindling, ECS-threshold, MES-threshold,

MET, MMT, and pentylenetetrazol-infusion. It was hypothesized that low doses of progestins, chronically administered, would have anticonvulsant effects in these tests.

4.2 Methods

Subjects

In al1 experiments. intact fernale Wistar rats were used as subjects. Subjects were

60 days old upon arriva1 from the breeding fms. They were allowed one week for acclimatization and daily handling before any treatments or tests were performed.

The subjects' stage in the estrous cycle was monitored by daily vaginal smears, as described in the General Methods.

Dmg Administration

In al1 experiments, subjects were randomized into three groups. The first group

(the control subjects) was injected with normal saline. The second group was injected with MPA (1 Smgkg, 0.2mL, IM,vehicle: saline). This dose of MPA reliably stopped the fertility cycle in subjects, indicating that it was effective. The third group was treated with X inch Silastic@capsules containing progesterone. The Silastic0 capsules were implanted subcutaneously in the mff of the neck under light halothane-induced anesthesia. This dose of progesterone has produced luteal phase levels of progesterone in

ovariectomized rats (Edwards et al., 1999).

Afier one week of dnig treatment seinire testing was initiated. Control subjects

were tested during proestrous. Drug-treated subjects who stopped cycling were tested

once every four days (in studies where repeated testing was required). Othenvise,

subjects were tested during proestrous.

Testing and Seùure Sconng

Experiment 2a: Kindling Before dmg administration, subjects were amygdala-kindled to 5 stage 5 seizures,

using procedures described in the General Methods. Progesterone or MPA were then

administered for 1 week. On the 8'h day, the threshold for amygdala-kindled generalized convulsions was determined by the half-split method, as described previously. The

lowest current that produced generalized seizures (stage 4 or 5) as defined by Racine's classification was considered the "thresh~ld'~.

Experiment 2b: ECS Threshold ECS threshold was measured by the half-split method, as described in the General

Methods section. The lowest current required to produce forelimb clonus was considered the threshold. Experiment 2c: MES Threshold MES threshold was measured by the half-split method, as described in the

General Methods section. The lowest current required to produce hind-limb tonus was

considered the "threshold".

Experiment 2d: MET The MET test was administered as described in the General Methods section.

Subjects were scored for presence or absence of clonic seizures within 30 minutes after

injection, and also for the delay to onset of the seizures (latency).

Experiment 2e: MMT The MMT test was administered as described in the General Methods section.

Subjects were scored for presence or absence of tonic (forelimb extension) seinws within 30 minutes afier injection, and also for the delay to onset of the seizures (latency).

Experiment 2E Pentylenetetrazol Infusion Pentylenetetrazol was administered via the tail vein, as described in the General

Methods section. The dose of pentylenetetrazol required to cause forelimb clonous was calculated and was considered to be "threshoid". Data Analysis

In al1 experiments, the data were andyzed using one-way analysis of variance

(ANOVA) or - when they were not norrnally distributed - the Kniskal-Wallis one-way analysis on ranks.

4.3 Rcsults

Effects on the Estrous Cyck

The estrous cycle was arrested in 84% of subjects that were treated with MPA

(N=50). Vaginal smears from subjects that had stopped cycling displayed characteristics of cells from diestrous or metestrous. With progestemne, only 28% of the subjects stopped cycling (N=43).

Experiment 2a: Kindling As indicated by Figure 8, the mean thresholds (+/- SE) for generalized amygdala-kindled convulsions (GCT) for saline, progesterone and MPA-treated subjects were 63 f 2 1,68 f 32, and 83 + 53 PA, respectively. Analysis of variance revealed no significant difference among these means (P=O.94 1). These data indicate that chronic administration of progestins does not significantly alter the threshold for amygdala- Amygdala-Kindled Convulsive Threshold

control progesterone MPA

Fiaure 8: The Effect of Chronic Proclesterone and MPA on the Generalized Convulsive Threshold in Amvadala-Kindled Rats.

The control, progesterone, and MPA groups contain 13, 8, and 6 subjects respectively. The data are ptesented as mean +1- standard error. No significant difference was found using Kruskal-Wallis one-way analysis on ranks (P=O.Q4l). kindled convulsions in rats. There was, however, a bbtrend"towards

anticonvulsant activity with both progesterone and, to a greater extent, the MPA

treatments.

Experiment 2b: ECS Tbreshold As indicated by Figure 9, the mean ECS thresholds (+/OSE) for saline,

progesterone and MPA-treated subjects were 25 f 3,25 f 2, and 27 f 3 rnA, respectively.

Analysis of variance revealed no significant difference among these means (P=0.140).

These data indicate that chronic administration of progestins does not significantly alter

the ECS threshold in rats. There was, however, a slight "trend" towards anticonvulsant

activity with the MPA treatrnent.

Experiment 2c: MES Threshold As indicated by Figure 10, the mean MES thresholds (+/- SE) for saline,

progesterone and MPA-treated subjects were 40 f 6,42 + 6, and 42 t 5 mA, respectively.

Analysis of variance revealed no significant difference among these means (P=0.7 18).

These data indicate that chronic administration of progestins does not significantly alter the MES threshold in rats. There was, however, a slight "trend" towards anticonvulsant activity with both the progesterone and the MPA treatrnents. ECS Convulsive Threshold

control progesterone MPA

Fiaure 9: The Effect of Chronic Progesterone and MPA on the Threshold for ECS Seizures.

Each group contains at least 13 subjects. The data are presented as mean +/- standard error. No significant difference was found using the Kruskal-Wallis one-way analysis on ranks (P=O. 140). MES Convulsive Threshold

control progesterone MPA

Figure 10: The Effect of Chronic Prooesterone and MPA on the Threshold for MES Seizures.

Each group contains at least 13 subjects. The data are presented as mean +Iostandard error. No significant difference was found using the Kruskal-Wallis one-way analysis on ranks (P=O.718). Experiment 2d: MET Al1 test subjects had a clonic seizure within 30 minutes of pentylenetetrazol injection.

As indicated by Figure 1 1, the mean latencies (+/- SE) to onset of forelimb clonus following penty lenetetrazol injection in saline, progesterone and MPA-treated subjects were 457 k 133,590 I423, and 87 1f 465 seconds, respectively. Kruskal-Wallis one- way analysis on ranks revealed no significant difference among these means (P=0.082), although the data were approaching significance. These data indicate that chronic administration of progestins does not significantly alter response to the MET stimulus in rats. There was a "trend", however, for progesterone and, to a greater extent, for MPA to increase latency. Using the Mann-Whitney rank sum test to compare only the saline and

MPA groups, the MPA group would have differed significantly fiom the saline group

(P=O.O 18).

Experiment 2e: MMT Eight out of 9 test subjects in each group had a tonic seizwe (forelimb extension) within 30 minutes of pentylenetetrazol administration. There was no noteworthy difference in seizure occurrence between two groups. Latency to MET Clonic Seizures

control progesterone MPA

Fiaure 11: The Effect of Chronic Prwesterone and MPA on tatencies ta the Onset of MET Clonic Seizures.

Each group contains at least 8 subjects. The data are presented as mean +/- standard error. No significant differenco was found using the Kruskal-Wallis one-way analysis on ranks (Pz0.082). Latency to MMT Tonic Seizures

control progesterone MPA

Ficiure 12: The Effect of Chronic Proclesterone and MPA on Latencies to the Onset of MMT Tonic (Forelirnb Extension) Seizu res .

Each group contains at least 8 subjects. The data are presented as mean +/- standard error. No significant difference was found using the Kruskal-Wallis one-way analysis on ranks (P-0.254). As indicated by Figure 12, the mean latencies (+/- SE) to onset of forelimb

extension following pentylenetetrazol injection in saline, progesterone and MPA-treated

subjects were 547 f 1 11,678 f 374, and 774 f 323 seconds, respectively. Kniskal-

Wallis one-way analysis on ranks revealed no significant difletence among these means

(P-0,254). These data indicate that chronic administration of progestins does not

significantly alter response to the MMT stimulus in rats. There was, however, a "trend"

for progesterone and, to a greater extent, MPA to increase latency.

Experiment 2f: Peatylenetetrazol Infusion As indicated by Figure 13, the mean doses (+/- SE) that produced clonic seizures in the pentylenetetrazol inhision test for saline, progesterone and MPA-treated subjects were 52 i 17,48 f 20, and 56 f 21 mgkg, respectively. Andysis of variance revealed no significant difference among these means (P=0.666). These data indicate that chronic administration of progestins does not significantly alter the convulsant pentylenetetrazol infusion dose in rats. There was, however, a "trend" towards anticonvulsant activity with

MPA treatment. Tonic Threshold: Pentylenetetrazol Infusion

control progesterone MPA

Fiaure 13: The Effect of Chronic Proaesterone and MPA on the Threshold Dose of Pentylenetetrazol Reauired to lnduce Clonic Seizures (FLC). The control. progesterone, and MPA groups contain 22, 12. and 13 subjects, respectively. The data are presented as mean +/- standard error. No signifiant difference was found using the Kruskal-Wallis one-way analysis on ranks (P=0.666) 4.4 Summary

Chronically administered progesterone and, to a greater degree, MPA displayed trends towards anticonvulsant activity in several tests. In no case, however, was the anticonvulsant activity of progesterone and MPA statistically significant. Our hypothesis that low doses of progestins, chronically administered, would have significant anticonvulsant effects was not supported. Chapter 5

The Anticonvulsant Effects of Progesterone in

Adult and 15-day-old Male and Female rats in the

MMT and the MES Seizure Models 5 The Anticonvubant Effects of Progesteroae in Adult and 15-day-old Male and

Female rats in the MMT and the MES Seizure Models

5.1 Rationale

Experiment 1, the initial dose-response study, was conducted with a 2-hour injectiodtest interval. This interval was the interval used by Edwds et al. (1999), and was based on the assurnption that pmgesterone's effects were genomic effects, mediated by the "classic" intracellular receptors. Genomic effects take some time to appear.

Recently, Edwards et ai. (submitted) have performed fiuther experiments, which showed anticonvulsant effects of a single dose of progesterone using a 15-minute injectiodtest interval. In Experiment 3, therefore, the time-response of progesterone was studied. These studies established that 15 minutes is an optimal injectiodtest interval.

Subsequently, dose-response studies were perfomed using a 15-minute injectiodtest interval.

The time-response studies were performed in the MMT model. The dose- response studies were performed in both the MMT and the MES rnodels. The MES model was used because it is a standard model of tonic-clonic seizures. The MMT model was used because it allows the evaluation of both tonic and the clonic responses simultaneously. In experiment 3, studies were done in both female and male adults rats, and in male and female 1S-day-old rat pups. The following experiments were performed:

Experiment 3a: the anticonvulsant efTects of progesterone in female rats

Experiment 3b: the anticonvulsant effects of progesterone in male rats

Experiment 3c: the anticonvulsant effects of progestemne in female rat pups

Experiment 3d: the anticonvulsant effects of progesterone in male rat pups

These experiments were designed to determine the phannacology of progesterone in tonic-clonic seinire models, and also to detemine whether progesterone is as effective in stopping seinires in males and infants as it is in females. It was hoped that these studies would provide a basis for using progesterone in clinical trials in a broader epileptic population (children and men as well as women). It was hypothesized that progesterone would be as effective in infants and adult males as it is in adult females.

5.2 Methods

Subjects

In experiments 3a and 3b, intact young adult female or male Wistar rats were used as subjects. Subjects were 60 days old upon arriva1 fiom the breeding fm. Subjects were allowed one week for acclimatization and daily handling before any treatments or tests were performed. In experiment 3c and 3d, 1S-day-old, Wistar male and femaie rats were used as subjects. Subjects were 8 days old upon arriva1 fiom the breeding farm. They arrived as

"made up" Mers of 12 with their dam.

Drug Preparation and Administration

In dl of the experiments, progesterone was prepareù with beta-hydroxy cyclodextnn and administered SC in the rough of the neck.

In experiments 3a and 3b, progesterone (various doses) was injected in a volume of 4mLkg. Pentylenetetrazol(8Smgkg) was prepand in normal saline and administered in a different part of the rough of the neck in a volume of 4.2mLIkg.

In experiments 3c and 3d, progesterone (various doses) was injected in a volume of 1OrnLkg. Pentylenetetrazol(1 SOmg/kg) was prepared in normal saline and administered in a different part of the rough of the neck in a volume of 10 xnL/kg.

Testing and Seizure Scoring

In the MMT test, pentylenetetrazol was administered at 85mg/kg in experiments

3a and 3b and at 1 50mgkg in experiment 3c and 3d as described above. Subjects were observed for 30 minutes, and scored as described in the General Methods. MES was administered according to the standard procedures that were described

in the General Methods.

In the time-response studies, a dose of 6Omgkg of progesterone was used, plus the following injectiodtest intervals: 5 minutes, 15 minutes, 1 hour, 2 hours, 6 hours and

24 hours.

Data Analysis

Time-response and dose-response curves were plotted for both the tonic and clonic elements of the seizures. Quantal curves were constructed by plotting the percentage of animals at each time or dose that were protected against the clonic or tonic seizures. Dose-response curves were fitted with linear regression, and ED,'s and maximal responses were calculated.

5.3 Results

5.3.1 Experimeat 3.: Adutt Female Subjects

Figure 14a shows the time-response curve in adult females for the suppression of tonic and clonic seimes in the MMT mode1 &er a single dose of progestemne

(6OmgIkg). Both tonic (forelimb extension) and clonic components of the seizure were considerably suppressed 15 minutes after treatment. Protection against the chic component began to disappear after 1 hour. Thereafter, protection against both Figure 14a,b,c: Anticonvulusant effects of Progesterone in Adult, Female Wistar Rats

The anticonvulsant effects of progesterone on suppression of forelimb clonus (m), forelimb extension (O), and, in figure 14c. hind-limb extension (w). Each point represents data fiom at least 5 subjects. Figure 14a: time-response study following a 60mgkg injection of progesterone in the MMT seinire test (8Smgkg of pentylenetetrazol, SC). Figure 14b: dose-response study in the MMT seimtest. Progesterone was administered 15 minutes prior to the seinw test. Figure 14c: dose-response study in the MES seinire test (1 XhAof current delivered via comeal electrodes). Progesterone was administered 15 minutes prior to the seizure test. time

Progesterone Dose-Response, MM2--

dose (mglkg)

1 I 1 I I I I 1 5 1O 20 40 80 160

Progesterone Dose-Response, MES

1 dose (mglkg)

Firiure 14a.b.c: adult fernales components declined. Curiously. there was major protection against both components of

the seizure only 5 minutes &er SC progesterone injection.

Based on the tirne-response study, dose-response tests were carried out using a

15-minute injectiodtest interval. Figure 14b shows the dose-response cwefor the

suppression of MMT-induced tonic and clonic seizures in adult females. The ED,s for

tonic (forelirnb extension) and clonic seinws were 14 and 19 mgkg, respectively.

Maximal responses (1 00% protection) were observed at 80 and 160 mgkg for tonic and

clonic seizures, respectively .

Figure 14c shows the dose-response curve for the suppression of MES-induced tonic and clonic seimes in adult fernales. The ED, for both tonic (hind-limb extension) and clonic seizuies was >320 mgkg. Maximal response for tonus (40% protection) was observed at 320rng/kg.

5.33 Experiment 3b: Adult Male Subjects

Figure 15a shows the time-response curve in adult males for the suppression of tonic and clonic seizures in the MMT mode1 after a single dose of progesterone

(6Omgkg). 60th tonic (forelimb extension) and clonic components of the seizures were drastically suppressed 15 minutes after treatment. Protection against both components began to disappear after 2 hou. At 5 minutes, there was complete protection against the tonic seizure, but only partial protection against the clonic seinire.

Based on the time-response study, and also for consistency with the experiments on female rats, dose-response tests were carried out using a 15-minute injectiodtest interval. Figure 1Sb shows the dose-response curve for the suppression of MMT-induced tonic and clonic seinires. The ED,s for tonic (forelimb extension) and clonic seizures were 23 and 35 mgkg respectively. Maximal responses (1 00% protection) were observed at 80 rnglkg for both tonic and clonic seizws.

Figure 1Sc shows the dose-response curve for the suppression of MES-induced tonic and clonic seizures. The ED,*s for tonic (hind-limb extension) and clonic seizures were 220 and >320mg/kg respectively. The maximal response for tonic hind-limb extension (1 00%) was reached at 320mgkg.

5.3.3 Experiment 3c: Immature Female Subjects

Figure 16a shows the tirne-response curve in immature fernales for the suppression of tonic and clonic seizures in the MMT mode1 after a single dose of progesterone (6Omgkg). Both the tonic (forelimb extension) and clonic components of the seizure were considerably suppressed 15 minutes &er treatment. Protection against Figure 1Sa,b,c: Anticonvulusant effects of Progesterone in Adult, Male Wistar Rats

The anticonvulsant effects of progesterow on suppression of forelimb clonus (O), forelimb extension (O), and, in figure 15c, hind-limb extension ( Each point represents data fiorn at least 5 subjects. Figure 1 Sa: time-response study following a 6Omg/kg injection of progesterone in the MMT seizure test (85mgkg of pentylenetetrazol, SC). Figure 15b: dose-response study in the MMT seinire test. Progesterone was administered 15 minutes prior to the seinire test. Figure 1 Sc: dose-response study in the MES seiwtest (150rnA of current delivered via comeal electrodes). Progesterone was administered 15 minutes prior to the seizure test. Progesterone Time-Response, MMT

time

1, r 1 I Sm 1Sm lhr 2hr 6hr

dose (mglkg) 1 1 I I 1 1 r 1 5 10 20 40 80 160

O n n n Q 1 m w O a Dose (mglkg)

Ficiure 1Sa.b.c: adult males both components began to disappear after 2 hours. Even at 5 minutes, there was

major protection against both the clonic and the tonic components of the seinires.

Based on the the-response study, dose-response tests were carried out using a

15-minute injectiodtest interval. Figure 16b shows the dose-mponse curve for the

suppression of MMT-induced tonic and clonic seizures. The EDs0sfor tonic (forelimb

extension) and clonic seizures were 15 and 28 mg/kg respectively. Maximal responses

(1 00% protection) were observed at 80 and 160 mgkg for tonic and clonic seizures,

respectively .

Figure 16c shows the dose-response cwefor the suppression of MES-induced

tonic and clonic seizures. The ED,,s for tonic (hind-limb extension) and clonic seizures were 23 and 16Omgkg respective1y. Maximal responses (1 00% protection) were observed at 320 mgkg for both tonic and clonic seizures.

53.4 Experiment 3d: Immature Male Subjects

Figure 17a shows the time-response curve in immature males for the suppression of tonic and clonic seizures in the MMT mode1 after a single dose of progesterone

(6Omglkg). Both tonic (forelimb extension) and chiccomponents of the seinires were considerably suppressed 15 minutes after treatment. Protection against the tonic and clonic components began to disappear aRer 2 hours and 1 hour, respectively. Even at 5 Figure 16a,b,c: Anticonvulusant effects of Progesterone in 15-Day-Old, Female Wistar Rats

The anticonwlsant effects of progesterone on suppression of forelimb clonus (e), forelimb extension (O), and, in figure 16c, hind-limb extension ( Each point represents data fiom at least 8 subjects. Figure 16a: time-response study following a 6Omg/kg injection of progesterone in the MMT seizure test (1 SOmgkg of pentylenetetrazol, SC). Figure 16b: dose-response study in the MMT seizure test. Progesterone was administered 15 minutes prior to the seinire test. Figure 16c: dose-response study in the MES seizure test (1 5OmA of current delivered via comeal electrodes). Progesterone was adrninistered 15 minutes prior to the seimtest. & i5m lhr P hr 6hr

Progesterone Dose-Response, MMT

dose (mglkg)

Progesterone Dose-Response, MES I

- Q a dose (rnglkg)

I 1 I 1 .

Fiqure 16aabac:15-day-old female rat pups Figure 1 7a,b,c: Anticonvulusant efTects of Propesterone in 15-Daysld, Male Wistar Rats

The anticonvulsant effects of progesterone on suppression of forelimb clonus (*), forelimb extension (O), and, in figure 17c, hind-limb extension (D. Each point represents data fiom at least 8 subjects. Figure 17a: time-response study following a 6Omgkg injection of progesterone in the MMT seizure test (1 50mgkg of pentylenetetnwl, SC). Figure 17b: dose-response sîudy in the MMT seizure test. Progesterone was administered IS minutes prior to the seizure test. Figure 17c: dose-response study in the MES seizure test (1 SOmA of current delivered via corneal electrodes). Progesterone was administered 15 minutes pior to the seiaire test. Progesterone Time-Response. MMT 100 - l Er iiLEFE] 80 -

60 -

40 - O

time

-20 1 , m 1 l 1 5m 15m 1hr 2hr 6hr 24hr

Progesterone Dose-Response, MMT 100 1 O 13

dose (mglkg) I I I r I I 1 5 1O 20 40 80 160

Prqlestemne Dose-Response, MES 700 1 v 7

1 Dose (mglkg)

Figure 17a.b.c: 15-day-old male rat pups minutes there was important protection against the tonic seizures, and partial protection against the clonic seizures.

Based on the time-response study, dose-response tests were canied out using a

15-minute injectiodtest interval. Figure 17b shows the dose-response curve for the suppression of MMT-induced tonic and clonic seizures. The EDg for tonic and clonic seizures were 12 and 85 mgkg, respcctively. Maximal responses (1 00% protection) were observed at 80 and 160 mgkg for tonic and clonic seizures, respectively.

Figure 17c shows the dose-response curve for the suppression of MES-induced tonic and clonic seinires. The ED,,s for tonic (hind-limb extension) and clonic seizures were 8 and 160 mgkg, respectively. Maximal responses (100% protection) were observed at 20 and 320 mgkg for tonic and clonic seinws, respectively. Table 4 The anticonvulsant EDSO's of progesterone for the suppression of MMT / MES seizures

*Tonic seizures were defined as forelimb extension in the MMT model, and hind-limb extension in the MES model.

- -- Seizure Gender EDSO Test Owkg) (Tonic*)

MT Adult Female 14

MMT Adult Male 23

MMT PUP Female 15

MMT PUP Male 12

MES Adult Female >320

MES Adult Male 220

MES PUP Female 23

MES PUP Male 8 The results of Expriment 3 are smarized in Table 4. Progesterone is an anticonvulsant in both female and male rats. Progesterone is an anticonvulsant in both female and male rat pups. The ED,,s in the MMT model are fat lower than in the MES model.

The original hypothesis was supported that progesterone is anticonvulsant in males and in infants. These data are novel and warrant Merstudy in clinical trials.

In addition, the very rapid onsets of progesterone's anticonwlsant actions cal1 into question the previous assumptions about the genomic nature of progesterone's anticonvulsant effects. Preliminary studies of progesterone's mechanism of action were perfomed in Experiment 4. Chapter 6:

Mechanistic Studies of the Anticonvulsant E ffects

of Progesterone 6 Mechanistic Sîudies of the Anticonvulsant ERects of Progesterone

6.1 Rationale

It seems probable that the anticonvulsant effects of progesterone are mediated by one of three different mechanisms: 1) by its active metabolite, allopregnanolone; 2) by binding to its "classic" (intracellular) receptors; or 3) by binding to the newly discovered ceIl surface receptors: Expriment 4 was designed as an initial investigation of these mechanisms.

Most investigators believe that allopregnanolone, progesterone's GABA-A-acting metabolite, is responsible for its anticonvulsant effects (General Introduction). We could not test this hypothesis by directly antagonizing GABA-A receptors, because al1 known

GABA-A antagonists are proconvulsant. Therefore, we blocked the conversion of progesterone to allopregnanolone by using indomethacin. Indomethacin binds to, and blocks the action of, Sa-reôuctase, fi rst of the enzymes involvcd in the conversion of progesterone to allopregnanolone.

The initial hypothesis of this thesis had been that progesterone worked through the classic (intracellular) progesterone receptor. RU486 (Mefipristone), an antagonist of the classic progesterone receptor, was used to test this hypothesis. nie onset of progesterone's anticonvulsant effects in Experiment 3 seems too rapid to be explained either by conversion to allopregnanolone or by binding to the intracelldar (genomic) receptors. It was hypothesized. therefore, that progesterone's

"fast" anticonvulsant effects were mediated by binding to the ce11 surface receptor. The expectation was that neither indomethacin nor RU486 would block progesterone's anticonvulsant effects. The MMT mode1 was used in these tests.

6.2 Methoda

Subjects

Intact, young adult, female Wistar rats were used as subjects. Subjects were 60 days old upon arriva1 from the breeding fm. They were allowed one week for acclimatization and daily handling before any treatrnents or tests were performed. After one week, subjects were randomly sorted into the following eight treatment groups:

Experiment 4a (Indomethesin)

1) Vehicle + Vehicle + MMT

2) Vehicle + P4 + MMT

3) Indomethacin + Vehicle + MMT

4) Indomethacin + P4 + MMT

Expenment 4b (RU486)

5) Vehicle + Vehicle + MMT

6) Vehicle + P4 + MMT + Vehicle + MMT

+ P4 + MMT

Dmgs and Dmg Administration

Progesterone (1 OOmg/kg) was dissolved with beta-hydroxy cyclodextrin to a concentration of 20mgkg. It was administered SC in the rough of the neck in a volume of 4mLkg. This dose is above the ED,, (Experiment 3).

Afier 15 minutes, pentylenetetrazol(85mglkg) - prepared in normal saline at a concentration of 2 1 mg/mL and injected in a volume of 4mLkg - was administered in a different part of the rough of the neck in a volume of 4mLkg. This dose prodixes maximal seizures.

Indomethacin (1 50 mgkg) was pnpared as a suspension, using normal saline, at a concentration of 37mghL. It was administered IP in a volume of 4mLkg, fifieen minutes before injecting progesterone. This dose of indomethacin is above doses used previously to block 5-alpha reductase. RU486 was prepared as a suspension, using cydo-h ydroxy cyclodex- in a concentratration of SrnghL. It was administered IP, at a dose of 20 mg/Kg in a volume of 4mL/kg, 15 minutes before injecting progesterone. This dose of RU486 is well above doses used previously to block the progesterone receptor.

Testing and Seizun Scoring

Seven days after arrival fiom the breeding fam, testing wa initiated. Subjects were preinjected with either RU486 or indomethacin. Fifteen minutes later, progesterone was administered and, 15 minutes after that, pentylenetetrazol was administered.

Subjects were scored for the presence or absence of tonic seizures within 30 minutes after injection.

Data Analysis

In both experiments, the Chi Square test was used to analyze the data.

6.3 Results

6.3.1 Experiment 4.: Indomethacin Figure 18 shows the e ffect of pre-treatment with indomethach on the anticonvulant effects of a single high dose of progesterone. These data show that: 1) pentylenetetrazol was an effective convulsant agent (vehicle+vehicle group), 2) progesterone at 100mgkg was an effective anticonvulsant (vehicle~progesteronegroup),

3) indomethacin did not interfere with the convulsant action of pentylenetetrazol

(indomethacin+vehicle group), and 4) indomethacin did not block the anticonvulsant action of progesterone (indomethacin+progesterone group). These data are statistically significant using the Chi Square test (P 5 0.001).

6.3.2 Experiment 4b: RU486

Figure 19 shows the effect of pre-treatment with RU486 on the anticonvulant effects of a single high dose of progesterone. These data show that 1) pentylenetetrazol was an effective convulsant agent (vehicle+vehicle group), 2) progesterone was an effective anticonvulsant (vehicle+progesterone group), 3) RU486 did not interfere with the convulsant action of pentylenetetrazol (RU486+vehicle group), and 4) RU486 did not block the anticonwlsant action of progesterone (RU486+progesterone group). These data are statistically significant using the Chi Square test (P 5 0.001)

6.3 Summay These data suggest that the "rapid" anticonvulsant effects of progesterone are not mediated by the classic progesterone receptor or by the GABA-A-binding metabolite of progesterone, allopregnanolone. They suggest that progesterone's "rapid" anticonvulsant effects are mediated by the newly discovered ce11 surface receptors. Anticonvulsant effect of progesterone in the presence or absence of indomethacin

vehicle vehicle indomethacin indomethacin + + + + vehicle progesterone vehicle progesterone

Fiaure 18: The Influence of lndomethacin (150 mglkri. IP) on the Anticonvulsant Effects of Proaesterone (1 00 malka

Seizures were triggered by pentylenetetrazol. Subjects were young adul fernale rats. Each point represents data from 10 rats. (*) indicates a significant difference frorn the ''vehicle+vehicle"controfs. using the Chi Square test (P<0.001). Anticonvulsant effect of progesterone in the presenee or absence of RU486

vehicle vehicle RU486 RU486 + + + + vehicle progesterone vehicle progesterone

Figure 19: The Influence of RU486 (20 mdks. IP) on the Anticonvulsant Effects of Proaesterone 11 O0 malka

Seizures were triggered by pentylenetetrazol. Subjects were young adult female rats. Each point represents data from 10 rats. (*) indicates a significant difference from the "vehicle+vehicle" controls, using the Chi Square test (PeO.001) Chapter 7:

Discussion 1. Discussion

7.1 The Kindling Data: Experiment 1

Experiment 1 was designed to test the anticonvulsant effects of progesterone in dose-response studies involving the kindling (arnygdala) mode!. Since Experiment 1 attempted to replicate a single-dose study by Edwards et al. (1999), an injectiodtest interval of 2 hours was used. The hypothesis was that progesterone would be anticonvulsant at low doses. It was thought that the low-dose anticonwlsant effects would be "genomic" effects, and that they might be seen in female, but not male, rats.

Anticonvulsant effects were observed only at high doses in Experiment 1 (ED,,>

30mgIkg). These findings are in agreement with Mohammad et al.'s (1998) findings.

Moharnmad et al. (1 998), working in amygdala-kindled male rats, found seizure suppression with ED,,s of greater than 60mg/kg. The results of Experiment 1, however, do not agree with the findings of Edwards et al. (1 999), since low dose effects were not observed in either males or fernales. Edwards et al. (1999) found that a low dose of progesterone (5mgkg) suppressed seizures in 60% of female subjects. Since progesterone was administered two hours before testing, high concentrations of allopregnanolone, but not of the parent compound (progesterone) would have been present at the time of seizure testing. The data fiom Experiment 1 appear to support the hypothesis that the anticonvulsant effects of progesterone are mediated by allopregnanolone. They appear to refbte the original hypothesis that low doses of progesterone, acting through genornic mechanisms, have anticonvulsant effects.

Although the animal data - fiom Experiment 1 and Mohammad's (1998) - appear to show considerable seizure suppression only at high doses, the clinical data

(Introduction) indicate that progesterone can suppress seizures at low doses. Therefore, low-dose effects were further investigated. Some features of the technique, which might have "masked" the low-dose effect of progesterone, were considered, and appropriate experiments were performed: 1) Phararncokinetic effects were ruled out by altering the route of administration and the vehicle. Neither changing the route nor the vehicle revealed a low dose effect. 2) The role of estrogen in synergizing progesterone's anticonvulsant effect was explored by using estrogen-primed (in ovariectomized rats) and intact (as opposed to ovariectomized) subjects. Since subsets of progesterone receptors are "inducible" by estmgen (Introduction) and the initial preparation used ovariectomized rats, it was thought that these "inducible" receptors might mediate a low-dose anticonvulsant effect of progesterone. Neither estrogen "priming", however, nor the use of intact subjects revealed a low dose anticonvulsant response to progesterone. 3) As a positive control, dose-response studies were preformed with phenobarôital in ovariectomized and estrogen "primed" subjects. These studies replicated Albnght and Burnham's findings (1980) in male rats, and indicated that anticonvulsant effects could be found in the paradigm used in the present study.

7.2 The 'Chronic, Low-Dose" Data: Experiment 2

In a Mersearch for low-dose effects, Experiment 2 was designed to test the anticonvulsant action of chronically adrninistered, low-dose progesterone and MPA. This treatment regimen imitates clinical practice more closely than the acute treatments used in

Experiment 1. The hypothesis was that low-dose of progesterone and MPA adrninistered chronically would be anticonvulsant. It was assumed again that the anticonvulsant effects would be "genomic" effects.

Chronic (7 day), low-dose administration of progesterone (Silastic@capsules resulting in luteal phase levels) or MPA (1 Smgkg, IM), however, failed to show a significant anticonvulsant action in a variety of seinire models. These findings are in agreement with Holmes and Weber (1 984) who reported that npeated administration of low doses of progesterone (Smgkg) had no effect on kindling parameters. These findings are in disagreement with Edwards et al. (1 999) who reported that chronic low-dose progesterone (Silastic@capsules) retarded kindling rates and raised seizure thresholds. It appears then that chronic, treatment with progesterone does not produce noteworthy anticonvulsant effects at low doses. These data are in agreement with Stitt and Kinnard

(1968) who found that chronic low-doses of progesterone (5mg/kg) and MPA (1Omglkg) did not alter MES seizure thresholds. These data are also in agreement with Woolley et al. (1 962) who found that chronic progesterone (5mg/kg) only altered MES sehe

thresholds by 5%. These data refùte the original hypothesis. It should be noted,

however, that trends were seen in several models. These represented the first suggestion of low-dose effects that were seen.

7.3 The Influence of gender and Age: Studies at 15-Minute InjectiodTest

Intewal

Experiment 3 was undertaken &ter Edwards et al. (submitted) reported anticonvulsant effects of progesterone in 15-day-old rat pups using a 15-minute injectiodtest interval and a dose of 2Omgkg. These data suggested that Experiments 1 had failed to show low dose efiects because the injectiodtest interval had been too long, and Experiment 2 had failed to show low dose eEects because the dose had been too low.

Time-response studies were therefore performed in adults and pups. These were designed to rneasure the time of onset and offset of progesterone's anticonvulsant actions.

Anticonvulsant actions were seen at 5 minutes derSC injection. These peaked by 15 minutes, and declined after 2 hours.

Subsequent to the time-response studies, dose-response studies were performed in adult and 15-day-old female and male rats in the MMT model. It was hypothesized that progesterone would be an effective anticonwlsant in the female and male adults and in

15-day-old rat pups. This was found to be true in the MMT seinw test, where relatively low dose effects were found (for tonic foreiimb exteasion), but not in the MES seizure

test, where only high dose effects were found, at least in adults.

Comparing the effects of progesterone in the different MMT groups, few

differences are seen. The ED,,s for tonic forelimb extension are similar in adult females

and 15-day-old rat pups (1 2- 14 mg/kg). They are somewhat higher in adult males

(3Omgkg) but still relatively low. These data are consistent with the kindling data

(Expriment 1). where male testosterone-replaced subjects had higher EDs0sthan female

subjects. The MMT data are a new addition to the literature.

Comparing the effects of progesterone in the different MES groups, ED,s for tonic hind-lirnb extension are much higher in adults (>320mgkg in females. 240mg/kg

in males) than in 15-day-old subjects (24rnglkg in females, 7mg/kg in males). 11 is interesting that progesterone is not a7 effective anticonvulsant against adult MES seizures, but it is quite effective against infant MES seizures. This may reflect the difference in seizure thresholds between adult and 15-day-old subjects - immature subjects having much higher thresholds. The MES stimulus may be closer to threshold in infants than adults. Consequently, a lower anticonwlsant dose of progesterone is required to raise thresholds above the seizure threshold.

In the MES model. the data regarding the immature subjects are a new addition to the literature. The data regarding the adult subjects are consistent with Craig (1966) and

Kokate et al. (1998), who found high EDs0sfor progesterone in the MES model using adult mice (300mgkg and 2SOmg/kg, respectively). These data are also consistent with

Stitt and Kirnard (1968), who found that a 70 mgkg of progesterone had no effect on

MES seizure threshold in adult female rats.

lt is unclear why progesterone was a more effective anticonvulsant in the MMT

seizure test in adults than in the MES seizure test. It is possible, however, that in the

MMT test the epileptogenic stimulus is closed to threshold than in the MES test. The

MES stimulus is 5 times threshold (Experiment 2), whereas the MMT stimulus is

possibly just above threshold, since some rats full to seize. This would suggest that

progesterone raises seizw thresholds by a small amunt.

The time-course studies provided some surprising findings, which suggest a

possible novel mechanism of action for progesterone. Progesterone was effective as early

as five minutes after administration. It must be remembered that MMT seizures have a

latency of 5 to 10 minutes, so it would be fair to Say that effects were seen at 10 to 15

minutes afier injection. Still, this is a very rapid response. These findings suggested that

progesteront's anticonvutsant action was not mediated by the gene transcription through the classical PR. Genomic effects do not occur this fast. They Mersuggested that the

parent compound, and not the active metabolites, might mediate progesterone's anticonvulsant effect. The time-course data led to Experiment 4. l 7.4 Studies of Mechanism: Experiment 4 IT-

Experiment 4 was designed to test whether. at an early time after injection (1 5

minutes), progesterone's metabolites were involved in the anticonvulsant action of

progesterone. It was hypothesized that they were not. Indomethacin was used to block

the Sa-reductase-mediated conversion of progesterone to metabolites such as

al lopregnanolone. Lt was hypothesized that progesterone would be an effective

anticonvulsant in the presence of indomethacin. This was found to be true. niese

findings are in disagreement with Kokate et al. (1 998), who found that finasteride -

another inhibitor of 5-alpha reductase - blocked the anticonvulsant effects of

progesterone. This conflict will be discussed below.

Experiment 4 was also designed to test whether genomic effects via the "classic"

receptors participated in progesterone's anticonvulsant action. RU486 (20mglkg) was

used to block the classical progesterone receptor. It was hypothesized that RU486 would

not block progesterone's anticonvulsant action. This was found to be true. This finding

is consistent with the previous findings of Mohammad et al. (1998), who also found that

RU486 does not block progesterone's anticonvulsant effects. It appears, then, that the

progesterone's anticonvulsant action does not involve the classic progesterone receptor.

This is consistent with the initial hypothesis. 7.5 How Does Progesterone Stop Seizum? :The Role of AUopregnanolone

Herzog has argued that the clinical anticonvulant actions of progesterone are completely mediated by its GABA-modulating metabolite, allopregnanolone. A nurnber of animal researchers agree with this point of view (eg. Mohammad et ai., Kokate et al.,

Landgren et al., and Frye et al.) Two facts, however, have always suggested that this is not the case: 1) Matson's Clinical study (1985) found that MPA was effective in treating catamenial seizures. MPA is not metabolized to any major degree to a GABA-A receptor-binding compound like allopregnanolone. 2) Relatively low doses of progesterone are used clinically. Herzog's progesterone treatment achieves normal luteal phase levels of 5-25ng/mL (Herzog, 1986). This results in allopregnanolone concentrations lower than those necessary to produce its anticonwlsant effects(Kokate et al., 1994;Kokate et al., 1996).

Data fiom present experiments provide new and direct evidence that non-GABA- mediated anticonvulsant effects exist for progesterone. In Experiment 4, blockade of the conversion of progesterone to allopregnanolone had no effect on progesterone's anticonvulsant actions. This implies that progesterone has major anticonvulsant effects that are entirely independent of GABA.

These data are not in complete agreement with the past studies of Kokate et al.

(1 998). There are only iwo cornmercially available Sa-reductase inhibitors, finasteride (Proscd) and indomethacin (IndocidO). Finasteride has been used by other investigators in the epilepsy field to determine the mechanism of action of progesterone.

Kokate et al. (1 998) used high doses of finasteride in conjunction with progesterone, and found that it completely blocked progesterone's anticonvulsant actions. Hence, Kokate concluded that progesterone's anticonvulsant action is entirely due to its metabolite, allopregnanolone. Recent studies (Edwards et al., personal communication), however, have shown that finasteride is itself a convulsant, and that high doses of finasteride cause seizures in both mature and immature rats (unpublished data). Therefore, this excludes the use of finasteride for mechanistic studies in seizure models. The finding, by Kokate, that finasteride blocks the anticonvulsant effects of progesterone could be due to finasteride's proconvulsant effects, rather than its inhibitory effects on Sa-reductase. It seems unlikely, therefore, that allopregnanolone plays a major role either in the clinical, or in the "fast" effects seen in the present study.

7.6 How Does Progesterone Stop Seizures? : Non-GABA-Mediated Actions

What are the possible non-GABA-modulating mechanisms of progesterone's

"fast" anticonvulsant action? Four possible mechanisrns may be considered: 1) non- specific effects, 2) binding to the intracellular receptor, or 3) non-GABA-acting metabolites, and 4) binding to the membrane receptor. 1) Non-Specilc Effects

Non-specific effects refer to alterations in neuronal excitability due to changes in membrane fluidity, or membrane expansion. In other worâs, might progesterone be acting as an anesthetic? If this were an important mechanisrn of progesterone's anticonvulsant action, then it would be expected that cholesterol, which has a similar chemical structure, would be an equally good anticonvulsant. Expriments on cholesterol have not shown it to be an effective anticonvulsant at similar doses (Likhodi et al., unpublished data).

2) Binding to the Intracellular Receptor

The assumption that guided the original hypothesis of this thesis was that progesterone was acting through the "classic" intracellular receptors. The rapid onset of progesterone's anticonvulsant actions in Experiment 3 made this dubious, and the fact that RU486 in Experiment 4 failed to block progesterone's anticonvulsant action seems to rule it out. This indicates that the classic progesterone receptor is not involved in progesterone's anticonvulsant actions.

3) Other Metabolites

As well as being metabolized to allopregnanolone, progesterone is also metabolized to IO-alpha dihydroprogesterone and 1 1-deoxycorticosterone (Figure 1).

There is no evidence that 20-alpha dehydroprogesterone has any anticonvulsant effects.

1 1 -deoxycorticosterone, however, is Merbroken dom to GABA-A-binding metabolites and does have GABA-mediated anticonvulsant effects. Recently ,Edwards et al. (unpublished) have shown that 1 1-deoxycorticosterone has sigaif~cantanticonvulsaat properties even in the presence of finasteride. This contrasts with the anticonvulsant effects of progesterone, which can be blocked by finasteride (Kokate et al., 1998). This irnplies that progesterone's and 1 1-deoxycorticosterone's non-GABA-mediated effects are different. Therefore, it is highly unlikely that progesterone's anticonvulsarit effects are mediated via l 1-deoxycorticosterone.

4) Binding to the Membrane Receptor

The most likely candidate for the non-GABA-mediated "fast" anticonwlsant effects of progesterone is the membrane progesterone receptor. The RU486 (Experiment

4) data support this mechanism. RU486 blocks the intracellular progesterone receptor but has a very low affiinity for the membrane progesterone receptor.

The MPA data are also consistent with this hypothesis. MPA binds the intracellular, but not the cell surface receptor.

Assuming that propesterone's anticonvulsant action is mediated by its membrane receptor, what downstream mechanisms could be involved? This is not clear as yet. the pi-imary known effect of the membrane progesterone receptor is to increase calcium currents, which should be procovulsant (Introduction) in the brain. These findings, however, are from sperm data. There is no evidence as yet, that the progesterone membrane receptors enhance calcium currents in the brain. There is currently a lack of comprehensive information on the

electrophysiological effects of progesterone membrane receptors in the brain. Further

research will be required. It seems probable. however, that progesterone's "fast"

anticonvulsant effects are mediated by binding to ce11 surface progesterone receptors.

7.7 How Does Progesterone Stop Seizures? Tbresbold Rises

What is the effect of progesterone on the "systems" level? The fact that progesterone was effective against MMT seinws, but not against MES seizures (in adults), suggests that progesterone raises seim thresholds, but that the threshold rises are not large.

Previous studies have also shown that small changes in seime thresholds can have clinical significance. The ketogenic diet, for instance, raises seizure thresholds minutely in animal seizure models by only about 20% (Thavendiranathan et al., 2000), yet it is an effective treatment for intractable childhood seizures. Smail rises in threshold can be important if they occur in patients that resist the standard anticonvulsant drugs.

7.8 How Does Progesterone Stop Seizures: Clinical Actions The present work has shown "fast" anticonvulsant effects of progesterone,

unrelated to aliogregnanolone and possibly mediated by membrane receptors. Do these

play a role in progesterone's clinical actions? While they could play an important role

(below), it seems unlikely that they can account for the effects seen clinically by Herzog and others.

Clinical studies establish progesterone concentrations within the normal physiological range. Even Smg/kg would produce supra-physiological concentrations, and the doses used in Experiment 3 (EDso's were 10-1 5 mgfkg.) are certainly supra- physiological.

Thus, while we have discovered new anticonvulsant actions of progesterone, they are probably unrelated to the effects seen in clinical studies. Those effects probably relate to progesterone's contraceptive and anti-estrogen effects.

1) Contraceptive effectg

The contraceptive-dated anticonvulsant effects of progesterone relate to its ability to eliminate the seizure triggers in catamenial exacerbation - the seinire triggers being estrogen and progesterone withdrawal. Progesterone does this by stopping the production of endogenous hormones through negative feedback to the hypothalamus and the anterior pituitary (Introduction). The progesterone intracellular receptor may be involved in this process. This mechanism would require prolonged exposure to progesterone. 2) Antkstrogen Effects

Progesterone's anti-estrogen effects may also be nlated to its actions on the classic progesterone receptors (Introduction). This mechanism involves the ability of

PR-A to act as a transcriptional inhibitor of other hormone receptors, including the estrogen receptors. Together, these mechanisms may explain progesterone's low dose clinical effects. This hypothesis is tenned the bbcontraceptive"hypothesis.

The "contraceptive hypothesis" of the anticonvulsant actions of progesterone - outlined above - would deal with dl three patterns of catamenial epilepsy: Pattern 1) exacerbation at the time of ovulation, Pattern 2) exacerbation during the entire luteal phase, and Pattem 3) exacerbation during the preimenstnial period. Progestins (used as contraceptives) stop the production of sex hormones through negative feedback in the hypothalamus and the ovaries. Hence, a patient who uses progestins as contraceptives: 1) would not have an estrogen surge which is associated with ovulation, 2) would not have elevated levcls of estrogen which are associated with the luteai phase, and 3) would not have the progesterone withdtawal which is associated with the perimenseual period.

Bauer's clinical studies ( 1992) support the hypotheses that the anticonvulsive properties of progestins are linked tc their contraceptive mechanisms of action. He used a synthetic GnRH analogue to suppress the fertility cycle in intractable epileptic patients with catamenial exacerbation. Using this treatment, he was able to reduce the freguency or abolish seizures to an equal or greater degree as in the studies, which have used progestins.

In summary, progesterone may stop seinws in three ways. At tmly low doses, in the chic, it may have "contraceptive" effects. At very high doses, in animal studies, it may have "slow" anticonvulsant effects related to its sedative metabolite, allopregnanolone. In addition, as we have show in the present work, at modem doses, progesterone has "fast" anticonvulsant effects, possibly related to binding to the ce11 surface receptor.

The "fast effects" of progesterone constitute a novel finding. It may provide a new target, and a new rnechanism of action, for anticonvulsant drug development.

7.9 Clinical Relevance of the Present Study

To date, only a srna11 number of female patients suffering fiom catarnenial seinires have benefited From progesterone therapy. The present study found major moderate-dose anticonvulsant actions of progesterone in adult male as well as female subjects (Expriment 3). Similady, major anticonvulsant effects were seen in immature subjects (Experiment 3). These "fast", moderate-dose effects were not associated with sedation. These findings imply that progesterone might be used clinically in males, infants and non-catarnenial females. A series of expanded clinical trials are in order. 7.10 Future Studies

Future studies must investigate the mechanisms of action of progesterone, and

also the nature of catamenial and intractable seizures:

1) To investigate the "fast" effects of progesterone, specific agonist and antagonists against the membrane progesterone receptor should be tested. Currently, these compounds exist, although they are not available commercially. This work should be carried on in whole animals, as well as in slice preparations. An alternative to specific agonists is BSA-conjugated progesterone. This cornpound does not cross ceIl membranes and its activity is limited to membrane receptors. Since BSA-conjugated progesterone also does not cross the blood-brain-bamer, it should be used in slices or administered

ICV via an implanted cannula in whole animals.

2) To show more conclusively that the "fast* effects of progesterone are not mediated by its GABA-modulating metabolite, plasma levels of progesterone and allopregnanolone should be measured. Hormone levels should be measured in a time response study, using the same parameters as Experiment 3. Hormone levels should also be measured in the presence or absence of indomethacin. 3) In the kindling experiments, dose-response studies were performed two hours

afler the administration of progesterone. These studies should be repeated at a test

interval of fifteen minutes. This would show whether the "fmt" effects are also present in

the kindling model, a model of complex partial seinires.

4) One of the anticonvulsant mechanisms of progestemne may be antiestrogenic

effect. This cm be tested easily and directly using antiestrogens such as roloxifene.

These should be tested as anticonvulsants.

5) The intractable nature of seizures with catamenial exacerbation may relate to the

role that estrogen plays in complex partial seizures. In Experiment 1, Estrogen treatment,

alone, lowered seizure thresholds at the (amygadala-kindled) seinve focus. This finding that estrogen is proconvulsant is in agreement with al1 previously published data. The

fact that seizure thresholds drop at the seizure focus, however, explains why the catamenial exacerbation is often associated with complex partial seizures. Estrogen

lowers seizure thresholds focally (in limbic structures) and also globally by a small dcgm (less than 30°/0). While this wvould not be important for the treatment of generaiized seizures, it has radical consequences for complex partial seizum. Albright and Bumham (1980) showed that AED's raise seime thresholds globally, but not in limbic structures. Hence, an endogenous substance (estrogen) that lowers limbic seizure thresholds by as much as 30% could cause seinires despite the presence of therapeutic doses of AED's. In fact, toxic doses of AED's would be required to return limbic seizure thresholds back to normal. The role that estrogen plays in cornplex partial seizures, therefore, can be studied in the following series of experiments. First dose-response studies should be performed on the effect of estrogen on after-discharge threshold in amygdala and cortical implanted subjects (before and after kindling). Next. the effect of AED's and antiestrogens on raising focal and generalized seizure thresholds in estrogen-pretreated, amygdala and cortical kindled subjects should be studied. If there is a difEerence in response between cortical and amygdala kindled subjects (as hypothesized), then the mechanism of action should be studied in slice preparations, which compare the effect of estrogen (on current fluxes for instance) on tissue fiom the cortex and the amygdala.

6) The finding that estrogen lowers seimre thresholds at the seizure focus also has significance for the role that glia play in the development of cornplex-partial seizures.

There are no estrogen receptors in the neurons of the baselateral amygdala (Table 1) - the site of the seinire focus - in the normal rat. Garcia-Segura's group (1 999), however, found that in response to injury activated glia express aromatase (Appendix 2). This response is irrespective of the site of injury. Hence, it is IikeIy that at the site ofthe seimre focus, estrogen is stimulating activated glia rather than neurons to lower the seizure threshold.

This theory can be verified by studying the effect of activation and inactivation of astrocytes on kindling thresholds. Alpha-arninoadipate, a compound that selectively destroys astrocytes, can be used at the site of kindling focus to test this hypothesis. Appendix 1:

AEDs Appendix 1: AED's

Pharmacological intervention in epilepsy began with the use of potassium bromide by Locock in 1857(Friedlander, l986b).

In 19 12, Hauptman introduced phenobarbital as the first organic antiseizure compound (cited by Woodbur~and Fingl, 1975). Ticku and Olsen discovered that phenobarbital binds to the barbiturate site on the gamma-aminobutyric acid (GABA) receptor, and enhances GABA-mediated inhibitory pathways(Ticku and Olsen, 1978).

Phenobarbital is no longer used as a first line therapy due to its severe sedative side effects. It is still used however, in poly-pharmacy and, in some cases, in the treatment of tonic-clonic and partial seizures(Leman-Sagie and Leman, 1999; Yukawa, 2000).

Meria and Putnarn introduced phenytoin in 1938(Glazko, 1986). The discovery of phenytoin was the result of a conscious effort to find a phenobarbital-like dmg with less sedative side effects. Many variants of the phenobarbital molecule were generated and tested. Phenytoin was the compound discovered. Phenytoin also marked an important advancement in the development of anticonvulsant drugs, since it was the first marketed antiseinire drug to be developed using an animal mode1 of epilepsy(Fned1ander 1986a). Willow and Catterail later discovered that phenytoin exerts i ts efiect by blocking voltage gated Na' channels(Willow and Catterall, 1982). Phenytoin is used pnmarily in the treatment of tonic-clonic and partial seizures(Penicca, 1996). Phenobarbital, and phenytoin are not effective in the treatment of absence

seinues(Murphy and Delanty, 2000). In 1945, Lemox found that trimethadione - a

variant of phenobarbital - was selectively effective against absence seizures. It was

ineffective against tonic-clonic or partial seimes. In 1963, it was found that

ethosuimide - another variant of Phenobarbital - was effective in the treatment of

absence seizures (Chen et al. 1963). Ethoswimide replaced trimethadione due to leu

severe side effects. Coulter has show that ethoswimide (and trimethidione) act through

inhibition of T-type voltage gated calcium channels in the thalamus (Coulter et al.,

1989;Davies, 1995). This finding was possible because of the work of Rowgoski who

discovered how neuronal firing in the thalamus is afFected by T-type calcium channels

(Suniki and Rogawski, 1989).

Benzodiazepines were introduced in the 1970's. They include clonazepam,

clobazam, and diazepam. They are effective against a broad range of seizures including partial. tonic-clonic and absence seimes(Kral1 et al., 1978). Mohler and Okada

discovered the benozodiazepine receptor, which was later found to be a site on the

GABA-A-related chlode chan.net (Mohler and Okada, 1977). Benzodiazepines increase

GABAergic inhibition by binding to the benzodiazepine site of the GABA-A receptor(Doble, 1999). B y tradition, clonazepam and cloba~unare used in the chronic treatment of seizures, whereas diazepam is adrninistered i.v. to stop recurrent seizures

('status epilepticus'). Two important drugs introduced in the 1970's were carbarnazepine and viilproate.

Carbarnazepine, has a similar mechanism of action to phenytoin (Macdonald and Kelly,

1995), but it has fewer side effects than phenytoin. It has therefore, become more widely used in the treatrnent of tonicslonic and partial seizures(Perucca, 1996).

Valproic acid was discovered accidentally when it was used as a solvent for potential AED. It has a broad range of activity, including activity against partial, tonic- clonic, and absence seinws(Perucca, 1996). This is possibly due to valproate's multiple mechanisms of action. Valproic acid both increases GABAergic inhibition and inhibits

NMDA-mediated ce11 firing (Loscher, 1993).

A number of new AED's have entered clinical PR-Actice in the past decade.

These are called "second generation" AED's (Loscher, 1998). Vigabatrin and tiagabine, enhance GABA-ergic inhibition. They are effective for the treatment of partial seinws and tonic-clonic seinires. Vigabatrin increases the levels of the neurotransmitter GABA by covalently binding to and inhibiting GABA transaminase, GABA's catalytic enzyme

(Graves and Leppik, 1993). Tiagabine hydrochloride increases synaptic GABA levels by blocking the re-uptake mechanism of GABA (Bourgeois, 1998).

Gabapentin and topiramate are "second generation" AED's with multiple mechanisms of action. They are effective for partial, tonic-chic and absence seinues.

Gabapentin has also been used as an add-on therapy in cases of intractable cornplex- partial seizures (Marson et al., 2000). Gabapentin's mechanisms of action include raising GABA levels in the brain and decreasing glutamte levels (McLean. 1999). Lt is unck which of gabapentine's mechanisms of action is primarily responsible for its anticonvulsant effects. Topiramate blocks sodium channels (Shank et al., 2000).

Lamotrigine acts mainly through sodium charme1 blockade. It also inhibits glutamate release (Coulter, 1997). It is effective against partial, tonic-clonic, and absence seinws. It also protects against some instances of drop attacks, which are a part of the

Lemox-Gastaut syndrome (Perucca, 1999). Appendix 2:

Neurosteroids - Tbree Important

Meta bolic Enzymes Appendix 2: Neurosteroids - Tbree Important Metabolie Enzymes

A2.1 Aromatase

Aromatase converts testosterone into estrogen. Aromatase is normally found only

in the neurons of the limbic structures such as the amygdala, the hippocampus, and the

hypothalamus. After brain injury however, astrocytes begin to express aromatase

regardless of the location of the injury(Bre~eret al., ). This strongly supports the notion

that estrogen is involved in brain repair, and helps explain the positive effects of this

hormone in the treatment of neurodegenerative diseases such as Alzheimer's(0riowo et al., 1980).

Aromatase distribution in the rat brain changes in three distinct, ontological, phases(0riowo et al., 1980). Aromatase is expressed in moderate levels in the following structures between GD16 and PND2 only: anterior mediai preoptic nucleus (aMPN), the stria1 part of the preoptic area (SPA), and the rostoral portion of the medial preoptic nucleus (MPN). In a second set of structures arornatase appears at GD 16, and reaches maximum levels between GD 18 and PND2. Aromatase expression then declines gradually to adult levels. Expression in these second set of structures forms a continuum that extends from the caudal parts of the lateral periphery of the media1 preoptic nucleus

(clMPN) to the principal nucleus of the bed nucleus of the stria teminalis (PR-BNST) and on Merto the posterdorsal part of the medial amygdaloid nucleus @dMAmg). In a third group of neurons aromatase appears only after PND 14. These newons are in the

lateral septal nucleus (iLS), the oval nucleus of the bed nucleus of the stria terminalis

(ovBNST), and the central amygdaloid nucleus (CAmg). These findings by Shinoda

( 1994) were in complete agreement with Lauber and Lichtestiger who also characterized the brain distribution of aromatase simultaneously(Oriowo et al., 1980). Using a different technique, Maclusky's group found that in addition to the high levels of ammatase in the hypothalamus-preoptic area and the amygdala, there are lower levels in the hippocarnpus, midbrain, and the cingulated cortex(0riowo et al., 1980).

The spatial as well as the temporal distribution of aromatase in the brain suggests that it plays an in important role in the sexual differentiation of the brain. Ontologically, masculinization of the brain, or lack of it, is determined by sex hormones. Mammalian brains are 'female' by default. It is the presence of estrogen that "masculinizes" the brain. This is possible because of aromatase(McEwen et al., 1997): the enzyme that converts testosterone to estrogen. Aromatase distribution and expression however, is very similar betweeti male and fernales(0riowo et al., 1980). The only di fference seems to be that males express higher Ievels of the enzyme(lauber et al., 1997;Oriowo et al.,

1980). It is primmily the pre and postnatal surge in testosterone in males (and lack thereof in fernales) that is converted to estrogen via arornatase and thereby masculinizes the brain(Shinoda, 1994). Matemal estrogen does not effect the fetal brain because of the expression of high concentrations of fetolneonatal estrogen binding protein (FEBP)that keep fiee-estrogen levels very low. FEBP binds estrogens with high affinity, but it does not sequester androgens(Komeyev et al., 1993). Subcellular localization of aromatase by electron microscopy has revealed the

presence of this enzyme on the surface of vesicles in presynaptic buttons(Med1ock et al..

1991). This finding lends support to theones on the actions of estrogen at surface

receptors modulating membrane potentiais(Leyendecker et al., 1975).

Two studies have looked at ammatase expression and activity in brain samples

fiom human patients with epilepsy(0riowo et al.. 1980). Expression and activity of

aromatase in hman brains is consistent with data obtained fiom the rat. The authors of

these studies suggest that enhanced aromatase activity in epileptic patients may increase their risk for having seizures. The present studies however, fail to prove this point since

proper controls are lacking.

A2.2 5-alpha reductase

5-alpha reductase metabolizes a number of hormones, including progesterone. 5- alpha reductase is ubiquitously distributed in the brain in both neurons and glia(Leyendecker et al., 1975 j. The enzyme activity however is not unifom in the brain.

Melcangi's group has show two seemingly contradictory results with respect to 5-alpha reductase activity. niey showed that neurons have greater enzyme activity than glia(Reddy and Kulkarni, 1998). They also showed that 5-alpha reductase activity is greater in white matter structures than in the cerebral cortex(Kokate et al., 1999). They fail to explain this discrepancy in their results, however, the latter finding has ken

reproduced and investigated in subsequent experiments(Gee et al., 1988;Komeyev et al.,

1993;Medlock et al., 1992).

The finding that 5-alpha reductase activity is greater in white matter led

Melcangi's group to speculate that this enzyme is involved in the process of myelin

formation. They found support for this hypothesis by showing that 5-alpha reductase is

present in myelin sheaths, while axons are devoid of the enzyme(Med1ock et al., 1992).

They also showed that dihyroprogesterone, the 5-alpha reduced metabolite of

progesterone that binds to the progesterone receptor, induces gene expression of

peripheral myelin protein zero (PO)(Kokate et al., 1999). This metabolite however is only

part of the mechanism regulating myelination since other metabolites of progesterone also regulate myelin related proteins. Baulieu's group simultaneously confimed

Melcangi's findings (and came to the same conclusions) by showing that in addition to its effects on Schwann cells, progesterone also increases the expression of myelin-specific proteins in oiigodendrocytes(0nowo et al., 1980).

Evidence from ontological expression and activity of 5-alpha reducatase also supports its role myelin formation and maintenance. In an early experiment, Melcangi's group fond that 5-alpha reductase activity in pwified myelin from rats was highest in the third week of life(Me1cangi et al., 1988). Lauber and Lichtensteiger preformed more detailed studies on the ontogeny of 5-alpha reductase type 1 and they found three distinct patterns of its mRNA expression(Lauber and Lichtensteiger, 1996). 1) In the embryonic stage, on GD 12- 18, 5-alpha reductase is expressed in the germinal and ventncular zones.

2) During the late fetal and early postnatal development 5-alpha reductase levels gradually decrease in the ventricular zone. but are expressed in differentiating regions of the brain including the cortical plate, the thalamus, the cerebellum, and the pyramidal ce11 layer of the hippocarnpus with the subiculurn. 3) From PND 15 and into adulthood lower levels of the enzyme are detected primarily in white matter structure. This last finding is in agnement with Melcangi's findings and hypothesis. The distribution of 5-alpha reductase in pre and early postnatal rat however, suggests a role in proliferation and differentiation(Lauber and Lichtensteiger, 1996).

A second isoform 5-alpha reductase has been found(Russell and Wilson, 1994). It is called type 2. 5-alpha reductase type 2 has a much higher affinity for various substrates than the type 1 isoform, however it is active only in a very narrow PH range around 5. The type 2 isoform has a different ontological profile from the type 1 enzyme.

It is absent at GD14. It is detected at GD18. It reaches its peak levels at PNDZ, and then decreases to extremely low levels or becomes absent altogether(Po1etti et al., 1998). In adutt rats the type 2 enzyme is detected primarily in the hypothalarnas(Po1etti and

Martini, 1999). Unlike 5-alpha reductase type 1, the type 2 isoform gene is highly inducible by testosterone, but only in cells derived from the hypothalamus. This is intriguing because both isofoms have a glucocorticoid response element in their promoter regions. Differences in presence or absence of cofactors might explain differrntial responses to testosterone. The correlation between the levels of 5-alpha reductase type 2 during the 'critical period' of sexual dittèrentiation and it's pattern of inducability by testosterone has led Poletti to hypothesizes that the type 2 isofomi may be

involved in sexual differentiation. Their studies to date however, have been in male rats

only(Po1etti et al., 1998). None-the-less there is strong support for this idea fkom the

observation that a genetically defective type 2 enzymes in the male human results in

pseudohemaphroditism(Imperato-McGinley et al., 1 974;Katz et al.. 1 995).

Recently Kellogg and Frye showed that the activity of 5-alpha reductase is different in infants and adults(Kellogg and Frye, 1999). In the fetal rat brain (fiom the

last five days of gestation) the levels of progesterone metabolites are twenty fold higher than the levels of progesterone, whereas the levels of testosterone metabolites are ten fold

lower than the levels of testosterone. In adults however, the main substrates for 5-alpha reductase seem to change. The conversion of progesterone to its metabolites becomes more inefficient, but the levels of testosterone 5-alpha reduced metabolites are three to ten folds higher than the levels of the parent hormone.

A2.3 3-alpha hydroxysteroid dehydrogenase (3-alpha HSD)

3-alpha HSD metabolizes a number of hormones, including 5-alpha reduced progesterone. 3-alpha HSD is not found in neurons(Me1cangi et al., 1993). 3-alpha HSD is found primarily in type 1 astrocytes and to a lesser degree in oligodendrocytes(Melcangi et al., 1994). Al1 regions of the rat brain express 3-alpha

HSD. The highest levels of 3-alpha HSD however, are found in the olfactory bulb(Cheng et al., 1994a). It is curious that the olfactory bulb also contains the highest concentration of intemewons in the brain. The distribution of 3-alpha HSD is consistent with the role of 3-alpha modified hormones in modulating the activity and genomic expression of

GABA-a receptors(Bele1li et al., 1990;Srnit.h et al., 1998a).

The promoter region of 3-alpha HSD contains response elements for estrogen, glucocorticoids and progesterone(Penning et al., 1997).

3-alpha HSD has a substrate affinity in the low micro molar range for 5-alpha reduced homiones. It has specific activity that is a 100 fold higher than 5-alpha reductase(Pe~inget al., 1985). This indicates that in vivo, 5-alpha reductase is the rate limiting step in producing 5-alpha,3-alpha modified steroids.

3-alpha HSD is responsible for giving testosterone and progesterone metabolites

GABA-modulating activity. There are however, important interspecies differences in the activity of this and other enzymes in the brain. Costa's group established the levels of various metabolites of progesterone in the braîns of rats, mice and monkeys(Korneyev et al., 1993). In the rat the major metabolites are 5-alpha dihydroprogesterone, and 3- alpha,S-alpha tetrahydroprogesterone. In the mouse the major metabolites are 5-alpha dihydroprogesterone, and 20-alpha dihyrdroprogesterone. In the monkey the main metabolite is 20-alpha di hydroprogesterone. Appendix 3:

The Role of Sex Hormones in Some Important

Aspects of Neuronal Plasticity Appendix 3: The Role of Sex Hormones in Some Important Aspects of Neuronai

Plasticity

A3.1 Some Behavionl Changea Dunng the Fertillty Cycle

A number of behavioral changes that occur during the estrous cycle, in pregnancy, at puberty or at menopause have provided insights into the role of sex hormones on brain plasticity. Figure 2 dernonstrates the variation in levels of progesterone and estrogen in the estrous cycles of hurnan and rat respectively.

Premenstnial syndrome is correlated temporally to progesterone withdrawal.

Smith's group suggests that premenstrual imtability is due to the effect of progesterone withdrawal on the genomic expression of the alpha4 subunit of the GABA-a receptor(Smith et al., 1998a). The differential expression of various GABA-a subunits altea the electrophysiological properties of the GABA-a cornplex. GABA-a complexes containing alpha-4 subunits have decreased chloride currents. A brain rich in alpha-4 subunits is more susceptible to seimres(Smith et al., 1998a) and is insensitive to the potentiating effect of benzodiazepines(Reddy and Rogawski, 2000; Smith et al., 1998~).

Smith suggests that postpartum depression and postmenopausal dysphoria also share the same mechanism as premenstmai syndrome. Tiredness is linked to elevated levels of progesterone. This phenornenon rnay be apparent in the late luteal phase; however, it is much mon pronounced in pregnancy where progesterone levels increase to more than ten times luteal levels(Behrenz and

Monga, 1999). There is a close relationship between plasma levels of progesterone and allopregnanolone(Concaset al., 1998). Tiredness during pregnancy is due to the agonistic affects of allopregnanolone on the GABA-a receptor(Rupprecht and Holsboer,

1999). The generai sedative effects of progesterone can manifest itself in depression as experienced by some women during pregnancy or during hormonal therapy(Herzog,

1986a).

The behavioural change that is of particular interest to this thesis is the catarnenial pattern of seizures. If pathological changes in the brin are an extention of neuronal placicity (McEachem and Shaw, 1999), then catamenial exacerbation too is a fonn of neuronal plasticity in response to hormones. In Catamenial epilepsy seizure fiequency is increased during the estrous cycle at times of high estrogen levels or with progesterone withdrawal, and it seems to decrease at times of high progesterone levels(Herzog et al.,

1997b).

Rat behavior in response to hormones is often similar to human behavior. There is a notewonhy exception. In the human female the height of sexual receptivity generally occurs at ovulation(Singer and Singer, 1972;Stanislaw and Rice, 1988). This corresponds to a peak in estrogen levels. In rats, sexual receptivity also occurs at ovulation; however, it is not pnmarily mediated by estrogen. This event occurs on estrous at the height of the progesterone peak(1chikawa et al., 1972).- . Estrogen priming in proestous however, is a

necessary prelude for the lordosis behavior(Freeman et ai., 1976).

A3.2 Neuronal Plasticity and the Fertility Cycle: Cellular Changes

It is possible to think of epilepsy as a pathologicai form of neuronal plasticity. In

this paradigm plasticity is a continuum encompassing development of neuronal circuitry,

learning, memory, and pathological disorders of the CNS(McEachem and Shaw, 1999).

Hence a number of cellular and molecular markers of neuronal plasticity can be used to

assess the extent of changes in the brain in response to various stimuli. Sex hormones

induce a number of cellular and molecular changes that are associated with neuronal

plasticity.

LTP and LDP are standard measurements of synaptic plasticity. LTP is very simply a semi-permanent enhancement in synaptic strength while LDP describes the

inverse process. Estrogen enhances long-terni potentiation (LTP) and attenuates long- term depression (LTD)(Good et al., 1999;Warren et al., 1995). This effect is mediated by

NMDA receptors(Foy et al., 1999). The effect of progesterone on these parameters is not clear. Allopregnanolone however, decreases LTP(Dubrovsky et al., 1993). LTP and LDP are also correiated with seinue activity. Electrical and chernical kindling enhance

LTP(Ben Ari and Gho, 1988;Suhila and Steward, 1986). Likewise, LTP speeds the kindling process(Sutu1a and Steward, 1987). In rats increased neurogenesis is obsemed during proestrous (Taubal1 Md

Lindstrom, 1993). Neurogenesis is a naturai process that occurs at a basal rate in the dentate gyrus and around the third ventricle(Kuhn and Svendsen, 1999). Many of the new neurons that are produced die soon after through apoptosis. Estrogen is an anti- apoptotic agent(Garcia-Segura et al., 2001), and it increases the number of new neurons by increasing cell sdval and also by increasing the rate of production(Tanapat et al.,

1999). Progesterone' s e ffect on neurogenesis is unclear.

New dentate granule neurons have a greater capacity for long-terni potentiation(Wang et al., 2000). They also have less GABA-ergic inhibition than older dentate granule neurons(Wang et al., 2000). Therefore, enhanced neurogenesis potentially enhances hippocampal excitability. Gould has proposed that neurogenesis is a potential mechanism through which hormones affect spatial learning and memory(Gou1d et al., 1999). Recently, abnormal neurogenesis has been linked to ôoth the onset and recovery fiom depression(Jacobs et al., 2000). Also, various methods of inducing seizures have been shown to enhance neurogensis(Parent et al., 1998;Scott et al., 2000).

Sex hormones affect spine density of the pyramidal neurons of the hippocampus(Woo1ley and McEwen, 1993). Spines are outgrowths of dendrites ont0 which afferent axons synapse. Woolley's group found that estrogen increases spine density by activating NMDA receptors(Wool1ey and McEwen, 1994). Hence increased spine density is associated with increased neuronal excitability. In the rat estrous cycle, spine density is greatest in proestrous. corresponding to an estrogen peak, and drop sharply in estrous corresponding to a progesterone peak(Wooi1ey and McEwen, 1994).

Hormonal regulation of mossy fiber sprouting is similar to the regulation of pyramidal ce11 spine densities. Mossy tibers are axons of dentate granule cells that synapse onto the dendrites of pyramidal cells in CA3 region of the hippocampus(Sutu1a et al., 1992). Estrogen enhances sprouting while the effects of progesterone have not been examined so far(Teter et al., 1999). Also, various foms of inducing seinires have been show to enhance mossy fiber sprouting(Sutu1a et al., 1992).

Recently Garcia-Segura proposed that estrogen increases synaptic excitability by altering synaptic architecture(Garcia-Seguni et al., 1999). This process involves the interaction between glia and neurons(M0r et al., 1999). On the afkemoon of proestrous estrogen activates astroglia by increasing the synthesis of GFAP. Activated astroglia ensheathe the soma of glutametergic neurons in their vicinity. In doing so they tem poraril y force the disinhibit ion of axonosomatic synapses b y intemeurons.

A3.3 Neuronal Plasticity and the Fertility cycle: Some Molecular Changes

A3.3.1 GABA GABA is the primary inhibitory amino acidin the &tain. GABA plays an important role in neuronal plasticity by regulating the expression its receptors or receptor subunits. Two alpha, two beta, and one gamma subunits make up the GABA receptor complex and each subunit is present in a multitude of isoforms(Mohler et al., 1996).

Abnonnalities in GABA-ergic response underlie the GABA hypothesis of epilepsy.

Loscher's group did not see any long-tenn changes in the number of GABA-ergic neurons in the hippocampus after kindling(Lehmann et al., 1996). Reductions however, were seen in the ipsilateral pirifom cortex(Lehrnann et al., 1998). A~so,regardless of the number of GABA-ergic neurons, kindling causes a long-lasting decrease in inhibitory effect of GABA in the hippocampus(Kamphuis et al., 1991). GABA receptor subunits are altered in various seizure models including kindling(Fol1esa et al., 1999;Kokaia et al.,

1994;Poulter et al., 1999).

Estrogen's effects on the GABA complex are still clouded in the literature.

Canonco has argued that estrogen decreases GABA receptor levels in the brain(Canonaco et al., 1993). Herbison on the other hand has argued that in areas of the bmin containing estrogen receptors, estrogen significantly increases the levels of some GABA receptor subunits(Herbison and Fenelon, 1995). Herbison's work is in agreement with Perez's findings that estmgen changes the function of GABA-a receptors by increasing the binding of GABA agonists in specific regions of the brain(Perez et al., 1988). This enhancement of GABA-ergic activity however, does not increase protection against convulsants(Perez et al., 1988). Progesterone's effects on G AB A receptors are ptunanly indirect and involve progesterone's tetra-hydroxy metabolite, also called allopregnanolone(Concas et al.,

1999). Allopreganaolone binds io the alpha subunit of the GABA-a receptor with a hundred fold greater afEnity than barbiturates(Gee et al., 1988). Allopregiuurolone enhances the effects of GABA and can alter subunit expression as well. Progesterone withdrawal increases the formation of the alpha-4 subunit of the GABA-a receptor(8mith et al., 1998b). This makes the complex less responsive to agonists of the alpha subunit.

A3.3.2 Glutamate

Glutamate is the primary excitatory amino acid in the brain. It plays an important role in neuronal plasticity via the AMPA, kainite, and NMDA receptors. Abnonnalities in glutametergic response underlie the glutamate hypothesis of epilepsy.

Wooley and McEwen have show that estrogen enhances certain excitatory glutamergic pathways. Estrogen increases NMDA agonist binding directly by regulating the agoinst binding siteWeiland et ai., 1997). Estrogen also increases the nurnber of synapses containing glutamatergic receptors(Wool1ey and McEwen, 1994). These axonodendritic synapses are made on denciritic spines. The estrogen-induced spine formation in the hippocampus during proestrous contain elevated levels of NMDA-1 receptors, but they lack AMPA receptors(Woolley, 1998). Progesterone does not alter the expression or bind directly to glutametergic

receptors. It may however, inhibit NMDA receptors indirectly through it's action on the

sigma receptor(M0nnet et al., 1995). Sigma receptor upon agonist binding potentiates

neuronal response to NMDA(Bergeron et al., 1999). Progesterone is a Sigma receptor

antagonist. Progesterone's effect on NMDA receptors is probably more apparent through

cross-talk with estmgen receptors. Woolley showed that initially, progesterone acts

synergistically with estrogen to enhance CA 1 dendritic spines containing NMDA receptors(Wool1ey and McEwen, 1993). Eighteen hours after progesterone treatment however, spine density is significantly decreased.

A3.3.3 Connexin

Gap junctions mediate a non-synaptic mechanism of cell-to-ce11 communication by pemitting direct electrical coupling between cells(Rozenta1 et al.,

2000b). Gap junctions are comprised of a large family of connexins. Six connexins form a hemichannel complex called a connexon at the cell surface(Dermietzel, 1998).

Co~exonsaggregate together into plaques. DisuIfide bonds adjoin connexons fiom plaques on adjacent cells(Foote et al., 1998;Manjunath et al., 1987). Gap junctions are involved in neuronal placticity and they are regulated by sex hormones(Perez et al.,

1990). Estrogen, alone or in combination with progesterone, increases connexin expression(Lye et al., 1993;Shinohara et al., 2000). Progesterone withdrawal also increases co~exinexpression(Micevych et al.. 1996). The combination of mechanical or electrical stimulation and estrogen synergistiacally increase gap junction mRNA expression. While electrical stimulation alone, or pmgesterone without estrogen however, do not increase connexin expression (Edwards and Maclusky, unpublished data).

Co~exin43 is the first isoform to be expressed in the brain(Chang and Balice-

Gordon, 2000). In the rat it is expressed at GD 12, peaks at PND2 1 and levels remain high into adulthood. It is primarily a glial gap junction although it is also found in some neurons such as the CA 1 pyramidal neurons(Rozenta1 et al., 2000a). Neuronal comexins such as connexin 32 and 36 are expressed most prominently in late gestation and drops to adult levets in the first few weeks of life(Bel1iveau and Naus, 1995;Rozental et al.,

2000~).An exception is the neuronal connexin 26, which is primarily expressed in the developing brain.

Outside the brain, gap junctions have a wide range of physiological functions.

Co~exin43 is responsible for the synchronous contractions of the heart. Gap junctions also synchronize labor contractions. In the brain, they mediate high frequency oscillations (100-200 Hz) in the hippcarnpus@raguhn et al., 1998;Draguhn et al., 2000).

These high fiequency oscillations seem to involve pyramidal neurons interconnected by axoaxonal gap junctions(Traub et al., 1999). Gap junctions also mediate slow calcium waves arnong astrocytes(Giaurne and Venance, 1998;Rottingen and Iversen, 2000).

Temporal expression of gap junctions suggests that they rnay also play a role in neuronal migration(Naus and Bani-Yaghoub, 1998). Hence, gap junctions may provide a general rnechanism through which excitatory or inhibitory pathways operate in a synchmnued fashion.

There is some evidence that gap junctions may be responsible for synchrony of epileptic seizures. Carlen's group showed that in absence of synaptic transmission, gap junctions mediate epileptiform activity in hippocampal slices(Carlen et al., 2000;Perez

Velazquez and Carlen, 2000). Ironically, in animal seinire models, connexin expression is decreased or remains unchanged(E1isevich et al., 1997a;Elisevich et al., 1998;Khurgel and Ivy, 1996). Likewise, hippocampal tissue From human epileptic patients does not show an increase in co~exinexpression(Elisevich et al., 1997b). Chang however, showed that following axonal injury there is increased gap junction coupling without an increase in expression(Chang et al ., 2000).

A3.3.4 c-fos

Sex hormones alter the expression of number of early response genes, such as c- fos. Both estrogen and progesterone enhance c-fos expression(Auger and Blaustein,

1997). c-fos expression is one of the hallmarks of neuronal plasticity and it is commonly used as a molecular market of seizure activity(Andre et al., 1998;Applegate et al.,

199S;Dragunow et al., 1988;Samonski et al., 1998;Simler et al., 1999). There is some evidence that c-fos is a convulsant agent and is involved in the spread of seizures in the kainic acid seizure model(Panegyres and Hughes, 1997). On the other hand, c-fos seems to be anticonvulsive in the kindling model(Rocha and Kaufman, 1998). A3.3.5 BDNF

Another family of early response genes is the neurotrophic factors. Neurotrophic

factors are an important part of the maintenance and repair mechanism of newons and gl ia(Mackay -Sima and C huahb, 2000). Estrogen alters the expression of nurnber of neurotrophic factors, such as brain-derived neurotrophic factor (BDNF). The response appears as an increase or decrease in expression depending on brain location or time interval afier application of estrogen(Gibbs, 1999;Murphy and Segal, 2000).

Progesterone can act in synergy with estrogen to enhance BDNF expression(Gibbs,

1999). Murphy's group has shown that BDNF prevents estrogen-induced dendritic spine formation in hippocampal neurons(Murphy et al., 1998). Furthemore, progesterone prevents spine formation through a non-BDNF mechanism(Murphy and Segal, 2000).

Consistent with Murphy's findings, BDNF decreases seinire susceptibility in the kindling model(0sehobo et al., 1999;Reibel et al., 2000). Ironically, BDNF enhances LTP(Lu and

Chow, 1999). Reference List

Aiken,S.P., Brown, W.M., 2000. Treatment of epilepsy: existing therapies and future developments. Front Biosci 5, E124-E152.

Albright,P.S., Bumharn,W.M., 1980. Development of a new phamiacological seizure model: effects of anticonvulsants on cortical- and arnygdala-kindled seinires in the rat. Epilepsia 2 1, 68 1-689.

Albright,P.S., Bumham,W.M., 1983. Effects of phenytoin, carbarnazepine, and clonazepam on coriex- and arnygdala-evoked potentials. Exp Neurol 8 1,308-3 19.

Andre,V., Pineau,N ., Motte, J.E., Marescaux,C., Nehlig, A., 1998. Mapping of neuronal nctworks underlying generalized seinires induced by increasing doses of pentylenetetrazol in the immature and adult rat: a c-Fos immunohistochemical study. Eur J Neurosci 10,2094-2 106.

Apgar,B.S., Greenbrg,G., 2000. Using progestins in clinical practice. Am Fam Physician 62, 1839-1850.

Applegate,C.D., Pretel,S., Piekut,D.T., 1995. The substantia nigra pars reticulata, seinires and Fos expression. Epilcpsy Res 20, 3 1-39.

Auger,A.P., Blaustein,J.D., 1997. Progesterone treatment incrcascs Fos-imrnunoreactivity within some progestin receptor-containing neurons in localized regions of female rat forebrain. Brain Res 746, 164- 170.

Avanzini,G., 1995. Anima1 models relevant to human epilepsies. Ital J Neurol Sci 16, 5-8.

Backstrom,T., 1976. Epileptic seinires in wornen reiated to plasma estrogen and progesterone during the menstrual cycle. Acta Neurol Scand 54,32 1-347.

Backstrom,T., Zetterlund,B., Blom,S., Rornano,M ., 1984. Effects of intravenous progesterone in fiisions on the epileptic discharge fkquency in women with partial epilepsy. Acta Neurol Scand 69,240-248.

Baird,D.T., McNeilly,A.S., 198 1. Gonadotrophic control of follicular development and hinction during the oestrous cycle of the ewe. J Reprod Fertil Suppl30, 119-133.

Baldi,E., Luconi,M., Muratori,M., Forti,G., 2000. A novel functional estrogen receptor on human spenn membrane interferes with progesterone effects. Mol Cell Endocrinol 16 1,3 1-35,

Behrenz,K.M., Monga,M., 1999. Fatigue in pregnancy: a comparative study. Am J Perinatol 16, 185-188.

Belelli,D., Lan,N.C., Gee,K.W., 1990. Anticonvulsant steroids and the GABA/benzodiazepine receptor- chloride ionophore cornplex. Neurosci Biobehav Rev 14,3 15-322.

Belliveau,D.J., Naus,C.C., 1995. Cellular localization of gap junction mRNAs in devetoping rat brain. Dev Neurosci 17,81-96.

Ben AR,Y ., Gho,M., 1988. Long-lasting modification of the synaptic properties of rat CA3 hippocampal neurones induced by kainic acid. iPhysiol (Lond) 404,365-384.

Berg,A.T., Testa,F.M., Levy,S.R., Shinnar,S., 1996. The epidemiology of epilepsy. Past, present, and future. Neurol Clin 14, 383-398. Bergeron,R., de Montigny,C., DebonneI,G., 1999. Pregnancy reduces brain sigma receptor function. Br J E - Phannacoi 127,1769-1776.

Berruti,G., 1998. Signaling events during male germ cell diflerentiation: bases and perspectives. Front Biosci 3, D1097-D1108.

Blankenstein,M.A., Koehorsf S.G., van der Kallen,C.J., Jacobs,H.M., van Spriel,A.B., Donker,G.H., van? Verlaat, J. W., Biaauw,G., Thijssen,J.H., 1995. Oestrogen receptor independent expression of progestin receptors in hurnan rneningioma--a review. J Steroid Biochern Mol Biol53,361365.

Bonuccclli,U., Melis,G.B., Paoletti,A.M., Fioretti,P., Murri,L., Muratori0.A.. 1989. Unbalanced progesterone and estradiol secretion in catarnenial epilepsy. Epilepsy Res 3, 100- 106.

Bourgeois,B.F., 1998. New antiepileptic dmgs [see comments]. Arch Ncurol55, 1 181- 1 183.

Brann,D.W., Rao,I.M., Mahesh,V.B., 1988. Antagonism of estrogen-induced prolactin release by progesterone. Biol Reprod 39, 1067- 1073.

Brenner,P.F., Guerrero,R., Cekan,Z., Diczfalusy,E., A radioimrnunoassay method for six steroids in plasma. pp. 25 1-64. In: Crosignani PG ,James VH ,ed Recent progress in reproductive endocrinology London, Academic Press, 1974 W3 SE4779W no 4 1973.

Britton,K.T., Koob,G.F., 1998. Neurophannacology. Prernenstrual steroids? [news; comment]. Nature 392, 869-870.

Brown,T.J., Yu,J., Gagnon,M., Sharma,M., MacLusky,N.J., 1996. Sex differences in estrogen receptor and progestin receptor induction in the guinea pig hypothalamus and preoptic area. Brain Res 725,374.

Buterbaugh,G.G., 1987. Acquisition of amygdala-kindled seizures in female rats: relationship behveen the effect of estradiol and intra-amygdaloid electrode location. Pharmacol Biochem Behav 28,291-297.

Calogero,A.E., Bumllo,N., Barone,N., Palenno,l., Grasso,U., DIAgata,R., 2000. Effects of progesterone on spenn function: mechanisms of action. Hum Reprod 15 Suppl 1,2845.

Carnacho-ArroyoJ., Guerra-Araiza,C., Cerbon,M.A., 1998b. Progesterone receptor isofonns are differentially regulated by sex steroids in the rat forebrain. Neworepon 9,3993-3996.

Camacho-Arroyo,l., Guerra-Araiza,C., Cerbon,M.A., 1998a. Progesterone receptor isofonns are differcntiaiiy regulated by sasteroids in the rat forebtain. Neuroreport 9,39933996,

Canonaco,M., Tavolaro,R., Maggi,A., 1993. Steroid hormones and receptors of the GABAA supramolecular complex. II. Progesterone and estrogen inhibitory effects on the chloride ion channel receptor in different forebrain areas of the female rat. Neuroendocrinology 57,974-984.

Carlen,P.L., Skinner,F., Zhang,L., Naus,C., Kushnir,M., Perez Velazquez,J.L., 2000. The role of gap Junctions in seizures. Brain Res Brain Res Rev 32,235-24 1.

Carter,R.B., Wood,P.L., Wieland,S., Hawkinson,J.E., Belelli,D., Larnbert,J.J., White,H.S., Wulf,H.H., Mirsadeghi,S., Tahir,S.H., Bolger,M.B., Lan,N.C., Gee,K.W., 1997. Characterization of the anticonvulsant properties of ganaxolone (CCD 1042; 3alpha-hydroxy-3beta-methyl-Salpha-pregnan-2O-one),a selective, high-affinity, steroid rnodulator of the gamma-aminobutyric acid( A) receptor. J Phmacol Exp Ther 2 80, 1284- 1295. Chang,Q., Balice-Gordon,R.J., 2000. Gap junctional communication among developing and injured motor neurons. min Res Brain Res Rev 32,242-249.

Chang,Q., Pereda,A., Pinter,M.J., Balice-Gordon,R.J., 2000. Ncrve injury induces gap junctional coupling among axotomized adult motor neurons. J Neurosci 20.674484.

Cheng,K.C., Lee,J., KhannqM., Qin,K.N., 1994b. Distribution and ontogeny of 3 alpha-hydroxysteroid dehydrogenase in the rat brain. J Steroid Biochem Mol Biol50,85-89.

Cheng,K.C., Lee,J., Khanna,M., Qin,K.N., 1994a. Distribution and ontogeny of 3 alpha-hydroxysteroid dehydrogenase in the rat brain. J Steroid Biochem Mol Biol 50,85089.

Compagnone,N.A., Mellon,S.H., 2000. Neurosteroids: biosynthesis and fiinction of these novel neuromodulators. Front Neuroendocrinol2 1, t -56.

Concas,A., Follesa,P., Barbaccia,M.L., Purdy,R.H., Biggio,G., 1999. Physiological modulation of GABA(A) receptor plasticity by progesterone metabolites. Eur J Phannacol 375,225-235.

Concas,A., Mostallino,M.C., Porcu,P., Follesa,P., Barbaccia,M.L., Trabucchi,M., Purdy,R.H., Grisenti,P., Biggio,G., 1998. Role of brain alloprcgnanolone in the plasticity of gamma-arninobutyric acid type A receptor in rat brain during pregnancy and aftcr delivery. Proc Natl Acad Sci U S A 95, 13284-1 3289.

Cooke,l.D., 1988. The corpus luteum. Hum Reprod 3, 153-1 56.

Coulter,D.A., 1997. Antiepileptic drug cellular mechanisms of action: where does lamotrigine fit in? J Child Neurol 12 Suppl 1, S2-S9.

Coulter,D.A., Huguenard,J.R., Prince,D.A., 1989. Characterization of ethosuximide reduction of low- threshold calcium current in thalarnic neurons. Ann Neurol23,582-593.

Cummings,L.N., Giudicc,L., Morrell,M.J., 1995. Ovulatory function in epilepsy. Epilepsia 36, 355-359.

Darne,F.J., McGarrigle,H.H., Lachelin,G,C., 1989. Diumal variation of plasma and saliva oestrogen, progesterone, cortisol and plasma dehydroepiandrosterone sulphate in late pregnancy. Eur J Obstet Gynecol Reprod Biol32,57-66.

Davies,J.A., 1995. Mechanisms of action of antiepileptic drugs. Seizure 4,267-27 1.

Delaunay,F., Pettersson,K., Tujague,M., GustafssonJ.A., 2000- Functional Werences betwecn the amino- terminal domains of estrogen receptors alpha and beta. Mol Pharmacol 58,584-590.

Dermietzel,R., 1998. Gap junction wiring: a 'new' principle in cell-to-cell communication in the nervous system? Brain Res Brain Res Rev 26, 176- 183.

Desmedt,L.K., Niemegeers,C.J., Lewi,P.J., Janssen,P.A., 1976. Antagonism of maximal metrazol seinires in rats and its relevance to an experimental classification of antiepileptic drugs. Ameimittelforschung 26, 1 592- 1603.

Devinsky,O., Kelley,K., Porter,R.J., Theodore,W.H., 1988. Clinical and electroencephalographic features of simple partial seizures. Neurology 38, i 347- 1352.

Doble,A., 1999, New insights into the mechanism of action of hypnotics. J Psychopharmacol 13, S 1 1 -S20. Draguhn,A., Traub,R.D., BibbigA., Schmitz,D., 2000. Ripple (approximately 200-Hz) oscillations in tmiporat stntctures f hr Process Citation]. I Cfin Neuroptiysïot 17,36ï-376.

Draguhn,A., Traub,R.D., Schmitz,D., Jefferys,J.G., 1 998. Electrical coupling underlies high-frequency oscillations in the hippocampus in vitro [see comments]. Nature 394, 189-192.

Dragun0w.M.. Robertson,H.A., Robertson,G.S., 1988. Arnygdala kindling and c-fos protein(s). Exp Neurol 1 02.26 1 -263.

Dubrovsky,B., Gijsbers,K., Filipini,D., Binningharn,M.K., 1993. Effects of adrenocortical steroids on long-term potentiation in the limbic system: basic mechanisms and behavioral consequences. Cell Mol Neurobiol 13, 399-4 14.

Edwar&,H.E., Burnham,W.M., MacLusky,N.J., 1999a. Testosterone and its metabolites affect aficrdischarge thresholds and the development of amygdala kindled seizures. Brain Res 83 8, 1 5 1 - 1 57.

Edwards,H.E., Bumharn, W.M., Mendonca,A., Bowlby,D.A., MacLusky,N.J., 1999b. Steroid hormones affect limbic allerdischarge thresholds and kindling rates in adult femalc rats. Brain Res 838, 136-150.

Edwards,H.E., Bumham, W.M., Mendonca,A., Bowlby,D.A., MacLusky,N.J., 1999c. Steroid hormones affect limbic afterdischarge thresholds and kindling rates in adult female rats. Btain Res 838, 136-150.

Elisevich,K., Rempel,S.A., Smith,B., Allar,N., 1997a. Connexin 43 mRNA expression in two experimental models of epilepsy. Mol Chem Neuropathol32,75-88.

Eliscvich,K., Rempel,S.A., Smith,B., Hirst,K., 1998. Temporal profile of connexin 43 mRNA expression in a tetanus toxin- induced seizure disorder. Mol Chem Neuropathol35.23-37.

Eliscvich,K., Rernpel,S.A., Smith,B.J., Edvardsen,K., 1997b. Hippocampal connexin 43 expression in human complex partial seizure disorder. Exp Neurol 145, 154-164.

Engel,J., Jr., 1996. Excitation and inhibition in epilepsy. Can J Neurol Sci 23, 167- 174.

Engel,J., Jr., Starkman,S., 1994. Overview of seizures. Emerg Med Clin North Am 12,895-923.

Ferrendelli,J.A., Holland,K.D., McKeon,A.C., Covey,D.F., 1989. Comparison of the anticonvulsant activitics of ethosuximide, valproate, and a new anticonvulsant, thiobutyrolactone. Epilepsia 30,6 17-622.

Feuer,G., 1983. Drug control of steroid metabolism by the hepatic endoplasmic reticuluni. Dnig Metab &v 14, 11 19-1 144.

Finn,D.A., Gee,K. W., 1994. The estrus cycle, sensitivity to convulsants and the anticonvulsant effect of a neuroactive steroid. J Pharmacol Exp Ther 27 1, 164- 170.

Fisher,R.S., 1989. Animal models of the epilepsies. Brain Res Brain Res Rev 14,245-278.

Follesa,P., Tarantino,A., Floris,S., Mallei,A., Porta& Tuligi,G., Cagetti,E., Caddeo,M., Mura,A., Sem,M., Biggio,G., 1999. Changes in the gene expression of GABAA mceptor subunit mRNAs in the septum of rats subjected to pentylenetetrazol-induced kindling. Brain Res Mol Brain Res 70, 1-8.

Foote,C.I., Zhou,L., Zhu,X., Nicholson,B.J., 1998. The pattern of disuifide linkages in the extracellular loop regions of connexin 32 suggests a mode1 for the docking interface of gap junctions. J Cell Biol 140, 1 187-1 197. Foresta$., Rossato,M., Di Virgilio,F., 1993. Ion fluxes through the progesterone-activated channel of the spenir plasma membrane. Biochem J 294 CR Ij, 279-283.

Foy,M.R., Xu,J., Xie,X., Brinton,R.D., Thompson,R.F., Berger,T. W., 1999. t 7beta-estradiol enhances NMDA receptor-mediated EPSPs and long-term potentiation. J Neurophysiol8 1,925-929.

Frang W.B.,111, 1988. Basic review: endocrinology of the normal menstrual cycle. Prim Care 15,607-6 16.

Freeman,M.C., Dupke,K.C., Croteau,C.M., 1976. Extinction of the estrogen-induced daily signal for LH release in the rat: a role for the proestrous surge of progesterone. Endocrinology 99,2230229.

Friedlander,W.J., 1986a. Puhum, Merritt, and ihe discovery of Dilantin. Epilepsia 27 Suppl 3, S 1-20.

Friedlander,W.J., 1986b. Who was 'the father of bromide treatmcnt of epilepsy'? Arch Neurol43,SOS-507.

Frucht,M.M., Quigg'M., Schwaner,C., Fountain,N.B., 2000. Distribution of seizure precipitants among epilepsy syndromes [In Process Citation]. Epilepsia 4 1, 1334-1 539.

Frye,C.A., Bayon,L.E., 1999. Cyclic withdrawal from cndogcnous and exogenous progesterone increases kainic acid and petforant pathway induced seizures. Phannacol Biochcm Behav 62,3 15-32 1.

Garcia-Segura,L.M., AzcoitiaJ., DonCarlos,L.L., 2001. Neuroprotection by estradiol [In Process Citation]. Prog Neurobiol63,29-60.

Garcia-Segura,L.M., Naf?olin,F., Hutchison,J.B., Azcoitia,l., Chowen,J.A., 1999. Role of astroglia in estrogen regulation of synaptic plasticity and brain repair. J Neurobiol40,574-584.

Gee,K. W., Bolger,M.B., Brinton,R.E., Coirini,H., McEwcn,B.S., 1988. Steroid modulation of the chloride ionophore in rat brain: structure- activity requircments, regional dependencc and mechanism of action. J Pharmacol Exp Ther 246,803-8 12.

George,F. W ., OjedqS.R., 1982. Changes in aromatase activity in the rat brain during embryonic, neonatal, and infantile development. Endocrinology 1 I 1, 522-529.

Giannakopoulou,M., Bozas,E., Philippidis,H., Stylianopoulou,F., 2001. Pmtooncogene c-fos Involvement in the Molecular Mechanism of Rat Brain Sexual Differentiation. Neuroendocrinology 73,387-396.

Giaume,C., Venance,L., 1998. Intercellular calcium signaling and gap junctional communication in astrocytes. Glia 24, 50-64.

Gibbs,R. B., 1 999. Treatment w ith estrogen and progesterone affects relative levek of brain-derived neurotrophic factor mRNA and protein in different regions of the adult rat brain. Brain Res 844,20-27.

Gilbert,M.E., Cain,D.P., 198 1. A developmental study of kindling in the rat. Brain Res 254,321-328.

Glazko,A.J., 1986. Discovery of phenytoin. Ther Dmg Monit 8,490-497.

Goddard,G.V., McIntyre,D.C., Leech,C.K., 1969. A permanent change in brain function resulting from daily electrical stimulation. Exp Neurol25,295-330.

Good,M., Day,M., Muir,J.L., 1999. Cyclical changes in endogenous levels of oestrogen modu lace the induction of LTD and LTP in the hippocampal CA1 region. Eur J Neurosci I 1,447614480. Gould,E., Bey1in.A.. TanapaSP., Reeves,A., Shors,T.J., 1999. Learningenhances adult neurogenesis in the hbpocampal formation [see comments]. Nat Neurosci 2,260-265.

Graves,N.M., Leppik,l.E., 1993. Advances in pharmacotherapy: recent developments in the treatment of epilepsy. J Clin Pham Ther 18,227-242.

Gu,Q., Korach,K.S., Moss,R.L., 1999. Rapid action of 17beta-estradiol on kainate-induced currents in hippocampal neurons lacking intracellular estrogen receptors. EndocrinoIogy 140,660-666.

Guerra-Amiza,C., Cerbon,M.A., Morimoto,S., Camacho-Arroyo,l., 2000. Progesterone receptor isofoms expression pattern in the rat brain during the estrous cycle. Life Sci 66, 1 743- 1752.

Hagihara,K., Hirata,S., Osada,T., Hirai,M., Kato,J., 1992. Distribution of cells containing progesterone receptor mRNA in the female rat di- and telencephalon: an in situ hybridization study. Brain Res Mol Brain Res 14,239-249.

Hall,S.M., 1977. Treatment of mensmal epilepsy with a progesterone-only oral contraceptive. Epilepsia 18,235-236.

Harden,C.L., Pulver,M.C., Ravdin,L., Jacobs,A.R., 1999. The effcct of menopause and perimenopause on the course of epilepsy. Epilepsia 40, 1402- 1407.

Hauser, W.A., Annegers,J.F., Kurland,L.T., 199 1. Prevalence of epilepsy in Rochester, Minnesota: 1940- 1980. Epilepsia 32,429-445.

Herbison,A.E., Fenelon,V.S., 1995. Estrogen regulation of GABAA receptor subunit mRNA expression in preoptic area and bed nucleus of the stria terminalis of female rat brain. J Neurosci 15,2328-2337.

Herzog,A.G., 1986b. Intermittent progesterone therapy and fiequency of complex partial seizures in women with menstruat disorders. Neurology 36, 1607-16 10.

Herzog,A.G., 1986a. Intermittent progesterone therapy and frequcncy of complex partial seinires in women with mensnual disorders. Neurology 36, 1607- 16 10.

Herzog,A.G., 199 1. Reproductive endocrine considerations and hormonal therapy for women with epilepsy. Epilepsia 32 Suppl6, S27433.

Herzog,A.G., 1995a. Progesterone therapy in women with complex partial and secondary generalized seiwres. Neurology 45, 1660-1662.

Henog,A.G., 1995b. Progesterone therapy in women with complex partial and secondary generalized seizures. Neurology 45, 1660- 1662.

Herzog,A.G., 1999a. Progesterone therapy in women with epilepsy: a 3-year follow-up. Neurology 52, 1917-1918.

Herzog,A.G., 1999b. Psychoneuroendocrine aspects of temporolimbic epilepsy. Part Il: Epilepsy and reproductive steroids. Psychosomatics 40, 102- 108.

Herzog,A.G., Klein,P., Ransil,B.J., 1997a. Three patterns of catarnenial epilepsy. Epilepsia 38, 1082- 1088.

Herzog,A.G., K1ein.P.. Ransi1,B.J.. 1997b. Three patterns of catarnenial epilepsy. Epilepsia 38, 1082- 1088. Holmes,G.L., Weber,D.A., 1984. The effect of progesterone on kindling: a developmental study. Brain Res ?tg* 45-53.

Howard,G., Peng,L., Hyde,J.F., 1997. An estrogen receptor binding site within the human galanin gene. Endocrinology 138,46494656.

Ichikawa.S., Morioka,H., Sawada,T., 1972. Acute effect of gonadotrophins on the secretion of progestins by the rat ovary. Endocrinology 90, 1356- 1362.

Imperato-McGinley,J., Guerrero,L., Gautier,T., Peterson,R.E., 1974. Steroid Salpha-reductase deficiency in man: an inherited form of male pseudohermaphroditism. Science 1 86, 12 13- 1 2 15.

Jacobs,B.L., Praag,H., Gage,F.H., 2000. Adult brain neurogcnesis and psychiatry: a novel theory of depression. Mol Psychiatry S. 262-269.

Joels,M., 1997. Steroid hormones and excitability in the mammalian brain. Front Neuroendocrinol 18-2- 48.

Kamphuis,W., Gorter,.l.A., da Silva,F.L., 199 1. A long-lasting decrease in the inhibitory effect of GABA on glutamate responses of hippocampal pyramidal neurons induced by kindling epileptogenesis. Neuroscience 4 1,425431.

Katz,M.D., Cai,L.Q., Zhu,Y.S., Herrera,C., DefilIo-Ricart,M., Shackleton,C.H., Imperato-McGinley,J., 1995. The biochernical and phenotypic characterization of females homozygous for 5 alpha-teductase-2 deficiency. J Clin Endocrinol Metab 80,3160-3 167.

Kellogg,C.K., Frye,C.A., 1999. Endogenous levels of 5 alpha-reduced progestins and androgens in fetal vs. adult rat brains. Brain Res Dev Brain Res 1 15, 17-24.

KepaJK., Jacobsen,B.M., Boen,E.A., Prendergast,P., Edwards,D.P., Takimoto,G., Wieman,M.E., 1996. Direct binding of progesterone receptor to nonconsensus DNA sequences represses rat GnRH. Mol Cell Endocrinol 1 17,27-39.

Kerrigan,J,F., Shields, W.D., Nelson,T.Y., Bluestone,D.L., Dodson, W.E., Bourgeois,B.F., Pellock,J.M., Morton,L.D., Monaghan,E.P., 2000. Ganaxotone for treating intractable infantile spasms: a rnulticenter, open-label, add-on trial [In Process Citation]. Epilepsy Res 42, 133- 139.

Khisti,RT., Mandhane,S.N., Chopde,C.T., 1998. The neurosteroid 3 alpha-hydroxy-5 alpha-pregnan-20- one induces catalepsy in mice- Neurosci Leû 25L,85-88.

Khurgel,M., Ivy,G.O., 1996. Astrocytes in kindling: relevance to epileptogenesis. Epilepsy Res 26, 163- 175.

Kokaia,M., Pratt,G.D., Elrner,E., Bengzon,J., Fritschy,J.M., Kokaia,Z., tindvall,O., Mohler,H., 1994. Biphasic differential changes of GABAA receptor subunit mRNA ievels in dentate gyrus granule cells following recurrent kindling-induced seizures. Brain Res Mol Brain Res 23,323-332.

Kokate,T.G., Cohen,A.L., Karp,E., Rogawski,M.A., 1996. Neuroactive steroids protect against pilocarpine- and kainic acid- induced limbic seizures and status epilepticus in mice. Neuropharmacology 35, 1049- 1056.

Kokate,T.G., Juhng,K.N., Kirkby,R.D., Llamas,J., Yamaguchi,S., Rogawski,M.A., 1999. Convulsant actions of the neurosteroid pregnenolone sulfate in mice. ûrain Res 83 1, 1 19- 124. Kokate,T.G., Svensson,B.E., Rogawski,M.A., 1994. Anticonvulsant activity of neurosteroids: correlation with gamma- aminobutyric acid-evoked chioride curent potenfration. 1PFiarmacoI Exp nier 270, 1223- 1229.

Kokka,N., Sapp,D.W., Witte,U., Olsen,R.W., 1992. Sex differences in sensitivity to pentylcnetetrazol but not in GABAA receptor binding. Pharmacol Biochem Behav 43,441-447.

Komeyev,A., Costa,E., 1996. Allopregnanolone (THP) mediates anesthetic cffécts of progesterone in rat brain. Hom Behav 30,37-43.

Korneyev,A., Guidotti,A., Costa,E., 1993. Regional and interspecies difletences in brain progesterone metabolism. J Neurochem 6 1,204 1-2047.

Krall,R.L., Penry,J.K., Kupferberg,H.J., Swinyard,E.A., 1978. Antiepileptic drug development: 1. History and a program for progress. Epilepsia 19,393408.

Kuhn,H.G., Svendsen,C.N., 1999. Origins, functions, and potential of adult neural stem cells. Bioessays 2 1, 625-630.

Landgren,S., AasiyJ., Backstrom,T., Dubrovsky,B., Danielsson,E., 1987. The eflect of progesterone and its metabolites on the interictal epileptiform discharge in the cat's cerebral cortex. Acta Physiol Scand 13 1, 33-42.

Lauber,A.H., Romano,G.J., Pfaff,D.W., 1991. Sex difference in estradiol regulation of progestin receptor mRNA in rat mediobasal hypothalamus as demonstrated by in situ hybridization. Neuroendocrinology 53, 608-6 1 3.

Lauber,M.E., Lichtensteiger,W ., 1994. Pre- and postnatal ontogeny of aromatase cytochrome P450 messenger ribonucleic acid expression in the mate rat brain studied by in situ hybridization. Endocrinology 135, 1661-1668.

Lauber,M.E., LichtensteigeqW., 1996. Ontogeny of 5 alpha-reductase (type 1) messenger ribonucleic acid expression in rat brain: early presence in germinal zones. Endocrinolagy 137,27 18-2730.

Lauber,M.E., Sarasin,A., Lichtensteiger,W., 1997. Sex di fferences and androgen-dependent regulation of aromatase (CYP19) mRNA expression in the developing and adult rat brain. J Steroid Biochem Mol Bi01 61,359-364.

Laxer,K., Blum,D., Abou-Khalil,B. W., Morretl,M.J., Lee,D.A., Data.J.L., MonaghaaEP, 2000. Assessment of ganaxolone's anticonvulsant activity using a randomized, double-blind, presurgical trial design. Ganaxolone Presurgical Study Group. Epilepsia 4 1, 1 187- 1 194.

Lehmann,H., Ebert,U., Loscher, W., 1996. Immunocytochemical localization of GABA immunoreactivity in dentate granule cells of normal and kindled rats. Neurosci Lett 212,4144,

Lehmann,H., Ebert,U., Loschet, W., 1998. Amygdala-kindling induces a lasting reduction of GABA- immunoreactive neurons in a discrete area of the ipsilatetal pirifonn cortex. Synapse 29,299-309.

Lephart,E.D., Sirnpson,E.R., McPhaul,M.J., Kilgore,M.W., Wilson,J.D., OjedqS .R., 1992. Brain aromatase cytochrorne P-450 messenger RNA levels and enzyme activity during prenatal and perinatal development in the rat. Brain Res Mol Brain Res 16, 187-192.

Leman-Sagie,T., Lerman,P., 1999. Phenobarbital still has a role in epilepsy treatrnent. J Child Neurol 14, 820-82 1. Leyendecker,G., Geppert,G., Nocke, W., Ufer,J., 1975. [Estradial- l7beta, estrone, LH and FSH in serum after administratron ofestradior-17beta, estradiolbenzoate, estradiol-valeriate and estradiol- undecylate in the female (author's transi)]. Geburtshilfe Frauenheilkd 35,370-374.

London,E.D., Buterbaugh,G.G., 1978. Modification of electroshock convulsive responses and thresholds in neonatal rais afier brain monoamine reduction. J Phannacol Exp Ther 206,8 1-90.

Loscher-W., 1993. Effects of the antiepileptic dnig valproate on metabolism and function of inhibitory and excitatory amino acids in the brain. Neurochem Res 18.485-502.

Loscher,W., 1998. New visions in the pharmacology of anticonvulsion. Eur J Pharmacol 342, 1-1 3.

Lu,B., Chow,A., 1999. Neurotrophins and hippocampal synaptic transmission and plasticity. J Neurosci Res 58, 76-87.

Lye,S.J., Nichotson,B.J., Mascarenhas,M., MacKenzie,L., Petrocelli,T., 1993. lncreased expression of conncxin-43 in the rat myometrium during labor is associated with an increase in the plasma estr0gen:progesterone ratio. Endocrinology 132,2380-2386.

Macdonald,R.L., Kelly,K.M., 1995. Antiepileptic drug mechanisms of action. Epilepsia 36 Suppl2, S2-12.

Macdonald,R.L., McLean,M.J., 1986. Anticonvulsant drugs: mechanisms of action. Adv Neurol44,7 13- 736.

Mackay-Sima,A., Chuahb,M.I., 2000. Neurotrophic factors in the primary olfactory pathway [In Process Citation]. Prog Neurobiol62,527-559.

MacLusky,N.J., Philip,A., Hurlburt,C., Naftolin,F., 1985a. Estrogen formation in the developing rat brain: sex differences in aromatase activity during eariy pst-natal lifc. Psychoneuroendoctinology 10,355-361.

MacLusky,N.J., Philip,A., Hurlburt,C., Naftolin,F., 1985b. Estrogen formation in the developing rat brain: sex differences in aromatase activity during early post-natal life. Psychoneurocndocrinology 10,355-36 1.

MacLusky,N.J., Walters,M.J., Clark,A.S., Toran-Allerand,C.D., 1994. Aromatase in the cerebral cortex, hippocampus, and mid-brain: ontogeny and devclopmental implications. Mol Cell Neurosci 5.69 1-698.

Maione,S., Berrino,L., Vitagliano,S., Leyva,J., Rossi,F., 1992. Pregncnolone sulfate increases the convulsant potency of N-methyl-D- aspartate in mice. Eur J Pharmacol 2 19,477-479.

Maitra,R., Reynolds,J.N., 1998. Modulation of GABA(A) receptor function by neuroactive steroids: cvidence for heterogeneity of steroid sensitivity of recombinant GABA(A) receptor isoforms. Can J Physiol Pharmacol 76,909-920.

Majewska,M.D., 1992. Neurosteroids: endogenous bimodal modulators of the GABAA receptor. Mechanism of action and physiological significance. Prog Neurobiol38, 379-395.

Majewska,M.D., Mienville,J.M., Vicini,S., 1988. Neurosteroid pregnenolone sulfate antagonizes electrophysiotogical responses to GABA in neurons. Neurosci Lett 90,279-284.

Majewska,M.D., Schwartz,R.D., 1987. Pregncnolone-sulfate: an endogenous antagonist of the gamma- aminobutyric acid receptor complex in brain? Brain Res 404,355-360.

Manjunath,C.K., Goings,G.E., Page,E., 1987. Human cardiac gap junctions: isolation, ultrastructure, and protein composition. J Mol Cell Cardiul 19, 13 1- 134. Marson,A.G.. Kadir,Z.A.. Hutton,J.L., Chadwick,D.W.. 2000. Gabapentin addsn for dmg-resistant partial epilepsy (Cochrane review) pn Process Citation]. Cochrane Database Syst Rev CD00 14 15.

Mattson,R.H., Cramer,J.A., Caldwell,B.V., Siconolfi,B.C., 1984b. Treatment of seizures with medroxyprogesterone acetate: preliminary report. Neurology 34, 1255- 1258.

Mattson,R.H., Cramer,l.A., Caldwell,B.V., Siconolfi,B.C., l984a. Treatment of seizures with medroxyprogesterone acetate: pwliminary report. Neurology 34, 1255- 1258.

McEachern,J.C., Shaw,C.A., 1999. The plasticity-pathology continuum: defining a role for the LTP phenornenon. J Neurosci Res 58.42-6 1.

McEwen,B.S., 1992. What makes a steroid a neurosteroid? Neurosteroids and Brain Function, sponsored by the Fidia Research Foundation, New Orleans, LA, USA, November 8-9, 199 1. New Biol4.2 12-2 16.

McEwen,B.S., 1999. Stress and hippocampal plasticity. Annu Rev Neurosci 22. 105- 1 22.

McEwen,B.S., Alves,S.E., Bulloch,K., Weiland,N.G., 1997. Ovarian steroids and the brain: implications For cognition and aging. Neurology 48, S8-15.

McLcan,M.J., 1999. Gabapentin in the management of convulsive disorders. Epilepsia 40 Supp16, S39- S50.

McNatty,K.P., 198 1. Hormonal correlates of follicular development in the human ovary. Aust J Biol Sci 34,249-268.

McNeilly,A.S., 1988. The control of FSH secretion. Acta Endocrinol Suppl (Copenh) 288,3 1-40,

McNicol,D., Jr., Crews,D., 1979. Estrogedprogesterone synergy in the control of female sexual receptivity in the lizard, Anolis carolinensis. Gen Comp Endocrinol38,68-74.

Medlock,K,L., Branharn, W.S., Sheehan,D.M., 1992. Long-term effects of postnatal exposure to diethylstilbestrol on uterine estrogen receptor and growth. J Steroid Siochem Mol Biol42,23-28.

Medlock,K.L., Forrester,T.M., Sheehan,D.M., 1991. Short-tenn effects of physiological and pharrnacological doses of estradiol on estrogen reccptor and uterine growth. J Recept Res I 1,743-756.

Melcangi,R.C., Celotti,F., Ballabio,M., Castanosp.,Po1etti.A.. Milan& Martini,L., 1988. Ontogenetic development of the 5 alpha-reductasc in the rat brain: cerebal cortex, hypothalamus, purifid myelin and isolated oligodendrocytes. Brain Res Dev Brain Res 44, 18 1 - 1 88.

Melcangi,R.C., Celotti,F., Castano,P., Martini,L., 1993. Differcntial localization of the 5 alpha-reductase and the 3 alpha- hydroxysteroid dehydrogenase in neuronal and glial cultures. Endocrinology 132, 1252- 1259.

Melcangi,R.C., Celotti,F., Martini,L., 1994. Ptogesterone 5-alpha-reduction in neuronal and in different types of glial ce11 cultures: type 1 and 2 astrocytes and oligodendrocytes. Brain Res 639,202-206.

Mermelstein,P.G., Becker,J.B., Surmeier,D.J., 1996. Estradiol reduces calcium currents in rat neostriatal neurons via a membrane receptor. J Neurosci 16,595-604.

Micevych,P.E., Popper,P., Hanon,G.l., 1996. Connexin 32 mRNA levels in the tat supraoptic nucleus: up- regulation prior to parturition and during lactation. Neuroendocrinology 63,3945. Mienville,J.M., Vicini,S., 1989. Pregnenolone sulfate antagonizes GABAA receptor-mediated currents via a reduction of chan& opening frquency . &sin Res 489, t WB&.

Mirsky,A.F., Duncan,C.C., Mystobodsky,M.S., 1986. Petit mal epilepsy: a review and iniegration of recent information. J Clin Neurophysiol3, 179-208.

Mohammad,S., Abolhassan,A., Pourgholami,M.H., 1998a. Evaluation of the anticonvulsant profile of progesterone in male amygdala-kindled rats. Epilepsy Res 30, 195-202.

Mohammad,S., Abolhassan,A., Pourgholami,M.H., 1998b. Evaluation of the anticonvulsant profile of progesterone in male amygdala-kindled rats. Epilepsy Res 30, 195-202.

Mohler,H., Fritschy,J.M., Luscher,B., Rudolph,U., Benson,J., Benke,D., 1996. The GABAA receptors. From subunits to diverse functions. Ion Channels 4,8901 13.

Mohier,H., Okada,T., 1977. Benzodiazepine receptor: demonstration in the central nervous systcm. Science 198, 849-85 1.

Monnet,F.P., Mahe,V., Robel,P., Saulieu,E.E., 1995. Neurosteroids, via sigma receptors, modulate the [3H]norepinephrine release evoked by N-methyl-D-aspartate in the rat hippocampus. Proc Natl Acsd Sci U S A 92,3774-3778.

Mor,G., Nilsen,J., H0rvath.T.. Bechmann,l., Brown,S., Garcia-Segura,L.M., Naftolin,F., 1999. Estrogen and microglia: A regulatoiy system that affects the brain. J Neurobiol40,484-496.

Moran,M.H., Smith,S.S., 1998. Progesterone withdrawal 1: pro-convulsant effects. Brain Res 807,84-90.

Morozova,T.M ., Levashova,Z.B ., NagibnevaJ .N., Rau,V.A ., Sidorkina,O.M., 1990. [The participation of transrnembrane messenger systems in the action of steroid hormones on target cells]. Fiziol Zh SSSR Im 1 M Sechenova 76, 1 1 79- 1 186.

Morrison,T., Waggoner,L., Whitworth-Langley,L., Stith,B.J., 2000. Nongenomic action of progesterone: activation of Xenopus oocyte phospholipase C through a plasma membrane-associatcd tyrosine kinase. Endocrinology 14 1.2 145-2 1 52.

Murphy,D.D., Cole,N.B., Sega1.M.. 1998. Brain-derived neurotrophic factor mediates estradiol-induced dendritic spine formation in hippocampal neurons. Proc Natl Acad Sci U S A 95, 1 14 12- 1 14 17.

Murphy,D.D., Segal,M., 2000. Progesterone prevents estradiol-induced dendritic spine formation in cultured hippocampaI neurons [In Process Citation], Neuroendocrinology 72, 13 3- 143.

Murphy,K., Delanty,N., 2000. Primary Generalized Epilepsies. Curr Treat Options Neurol2, 527-542.

Murri,L., Galli,R,, 1997. Catamenial epilepsy, progesterone and its metabolites. Cephalalgia 17 Suppl20, 46-47.

Narbone,M.C., Ruello,C., Oliva,A., Baviera,G., D8Amico,D.,Bramanti,P., Di Perri,R., 1990. Hormonal disregulation and catamenial epilepsy. Funct Neurol5,49-53.

Naus,C.C., Bani-Yagh0ub.M.. 1998. GAP JUNCTIONAL COMMUNICATION IN THE DEVELOPING CENTRAL NERVOUS SYSTEM. Cell Biol tnt 22,75 1-763. Nicoletti,F., Speciale,C., Sortino,M.A., Summa,G., Caruso& Patti,F., Canonico,P.L., 1985. Comparative effects of muadiot bene,the mthmgmctamipttme citrate, aircttk pragcstin medroxyprogesterone acetate on kainic acid- induced seizures in male and female rats. Epilepsia 26,252-257.

NissinenJ., Halonen,T.. Koivisto,E., Pitkanen,A., 2000. A new model of chronic temporal lobe epilepsy induced by electrical stimulation of the amygdala in rat. Epilepsy Rcs 38, 177-205.

O'Conner,J.L., Wade,M.F., PrendergashP., Edwards,D.P., Boonyaratanakomkit,V., Mahesh,V.B., 1997. A 36 1 base pair region of the rat FSH-beta promoter contains multiple ptogesterone receptor-binding sequences and confers progesterone responsivencss. Mol Cell Endocrinol 136.67-78.

Orikasa,C., McEwen,B.S., Hayashi,H., Sakuma,Y., Hayashi,S., 2000. Estrogcn receptor alpha, but not bcta, is expressed in the intemeurons of the hippocampus in prepubertal rats: an in situ hybridization study. Brain Res Dev Brain Res 120,245-254.

0riow0,M.A.~Landgren,B.M., Stenstrom,B., Diczfalusy,E., 1980. A cornparison of the pharmacokinetic properties of three estradiol esters. Contraception 2 1,4 15-424.

Osehobo, P., Adams, B., Sazgar,M., Xu,Y ., Racine,R.J., Fahnestock,M., 1999. Brain-derived neurotrophic factor infusion delays amygdala and perforant path kindling without affecting paired-pulse measures of neuronal inhibition in adult rats. Neuroscience 92, 1367-1375.

Osterlund,M.K., KellefiE., Hurd,Y.L., 2000. The human forebrain has discrete estrogen receptor alpha messcnger RNA expression: high levels in the arnygdaloid cornplex. Neuroscience 95,333-342.

Panegyres,P.K., Hughes,J., 1997. The anticonvulsant properties of antisense c-fos oligodeoxynucleotides in kainic acid-induced seizures. J Neurol Sci 153, 12-19.

Parent,J.M., Janumpalli,S., McNamara,J.O., Lowenstein,D.H., 1998. lncreased dentate granule cell neurogenesis following amygdala kindling in the adult rat. Neurosci Lett 247,9-12.

Pelletier,G., Luu-The,V., Labrie,F., 1994. lmmunocytochemical localization of 5 alpha-reductase in rat brain. Mol Cell Neurosci 5, 394-399.

Pellmar,T.C., Wilson,W.A., 1977. Synaptic mechanism of pentylenetetrazole: selectivity for chloride conductance. Science 197,9 12-9 14.

Penning,T.M., Pawlowski,J.E., Schlegel,B.P., Jez,J.M., Lin,H.K., Hoog,S.S., Bennett,M.J., Lewis,M., 1997. "Mammalian 3 alpha-hydroxysteroid dehydrogenases". Steroids 62,455-456.

Penning,T.M., Sharp,R.B., Krieger,N.R., 1985. Purification and properties of 3 alpha-hydroxysteroid dehydrùgenase from rat brain cytosol. Inhibition by nonsteroidal anti-inflamrnatory drugs and progestins. J Biol Chem 260, 15266- 15272.

Perez Ve1azquezJ.L.. Carlen,P.L., 2000. Gap junctions, synchrony and seizures. Trends Neurosci 23,68- 74.

PerezJ., Tranque,P.A., Naftolin,F., Garcia-Segura,L.M., 1990. Gap junctions in the hypothalarnic arcuate neurons ~Fovariectomizedand estradiol-treated rats. Neurosci Lett 108, 1 7-2 1,

Perez,J., Zucchi,l., Maggi,A., 1988. Estrogen modulation of the gamma-arninobutyric acid receptor com plex in the central nervous systern of rat. J Phamiacol Exp Ther 244, 1005- 10 10.

Perucca,E., 1996. Established antiepileptic drugs. Baillieres Clin Neurol 5,693-722. Perucca,E., 1999. The spectrum of the new antiepileptic drugs. Acta Neurol Belg 99, 23 1-238.

Petraglia,F., de Micheroux,A.A., Florio,P., Salvatori,M., Gallinelli,A., Cela,V., Palumbo,M.A., Genazzani,A.R., 1995. Steroid-protein interaction in hurnan placenta. J Steroid Biochem Mol Biol53,227- 23 1.

Pine1,J.P.. Skelton,R., Mucha,R.F., 1976. Kindling-related changes in afterdischarge "thresholds". Epilepsia 17, 197-206.

Poletti,A., Martini,L., 1999. Androgen-activating enzymes in the central nervous system. J Steroid Biochem Mol Bi01 69, 1 17- 122.

Poletti,A., Negri-Cesi,P., Rabuffetti,M.. Colciago,A., Celotti,F., Martini,L., 1998. Transient expression of the Salpha-reductase type 2 isozyme in the rat brain in late fetal and early postnatal life. Endocrinology 139,2171-2178.

Poulter,M.O., Brown,L.A., Tynan,S., Willick,G., Williarn,R., Mclntyre,D.C., 1999. Differential expression of alphal, alpha2, alpha3, and alpha5 GABAA receptor subunits in seinire-prone and seinire-resistant rat models of temporal lobe epilepsy. J Neurosci 19,4654-466 1.

Racine,R.J., 1972. Modification of seimre activity by electrical stimulation. 1. After- discharge threshold. Electroencephalogr Clin Neurophysiol32,269-279.

Ramirez,V.D., Zheng,J., Siddique,K.M., 1996. Membrane teceptors for estrogen, progestcrone, and testosterone in the rat brain: fantasy or reality. Cell Mol Neurobiol 16, 175-198.

Reddy,D.S., Kulkarni,S.K., 1998. Proconvulsant effects of neurosteroids pregnenolone sulfate and dehydroepiandrosterone sulfate in mice. Eur J Pharmacol 345,5559.

Reddy,D.S., Rogawski,M.A., 2000. Enhanced anticonvulsant activity of ganaxolone afier neurosteroid withdrawal in a rat model of catamenial epilepsy. J Pharmacol Exp Ther 294,909-91 5.

Reibel,S., LarmeGy., Le,B., Carnahan,J., Marescaux,C., Depaulis,A., 2000. Brain-derived neurotrophic factor delays hippocarnpal kindling in the rat [In Process Citation]. Neuroscience 100, 777-788.

Reindollar,R.H., Moretuzzo,R. W., Ellegood,J.O., Mahesh,V.B., McDonough,P.G., 1985. The effects of single-dose luteinizing hormone-releasing hormone on ovulatory menstrual function: support for a single luteinizing hormone- and follicle-stimulating hormone-releasing factor. Fertil Steril43, 185- 195.

Richard,S., Zingg,H.H., 1990. The human oxytocin gene promoter is regulated by estrogens. J Bi01 Chem 265,6098-6 103.

Rocha,L., Kaufinan,D.L., 1998. In vivo administration of c-Fos antisense oligonucleotides accelerates amygdala kindling. Neurosci Lett 24 1, 1 1 1-1 14.

Rogvi-Hansen,B., Gram,L., 1995. Adverse effects of established and new antiepileptic drugs: an atternpted cornparison. Pharmacol Ther 68,425434.

Rosse,R.B., Mastropaolo,J., Novitzki,M.R., Deutsch,S.I., 1990. Depot testosterone attenuates the anticonvulsant effect of flurazepam in mice. Psychoneuroendocrinology 15,83-85.

Rottingen,J., Iversen,J.G., 2000. Ruled by waves? lntracellular and intercellular calcium signalling. Acta Physiol Scand 169,203-2 19. Rozental,R., Giaume,C., Spray,D.C., 2000a. Gap junctions in the nervous system. Brain Res Brain Res Rev 32, 1145.

Rozental,R., Giaume,C., Spray,D.C., 2000b. Gap junctions in the nervous system. Brain Res Brain Res Rev 32, 11-15.

Rozental,R., Srinivas,M., Gokhan,S., Urban,M., Dennietzel,R., Kessler,J.A., Spray,D.C., Mehler,M.F., 2000c. Temporal expression of neuronal connexins during hippocampal ontogeny. Brain Res Brain Res Rev 32, 57-7 1.

Rupprecht,R., Holsbocr,F., 1999. Neuropsychopharmacological propeflies of neuroactive steroids. Steroids 64,83-9 1.

Russell,D. W., WiIson,J.D., 1994. Steroid 5 alpha-reductase: hvo genesbo enzymes. Annu Rev Biochem 63,25-61.

Sackeim,H., Decinqp., Prohovnik,I., Malitz,S., 1987. Seizure threshold in electroconvulsive therapy. Effects of sex, age, electrode placement, and number of treatments. Arch Gen Psychiatry 44,355-360.

Samoriski,G.M., Piekut,D.T., Applegate,C.D., 1998. Regional analysis of the spatial patterns of Fos induction in brain following flurothyl kindting. Neutoscience 84, 1209-1222.

Sanghera,M.K., Simpson,E.R., McPhaul,M.J., Kozlowski,G., Conley,A.J., Lephm,E.D., 1991 a. lmmunocytochemical distribution of aromatase cytochrome P450 in the rat train using peptide-generated polyclonal antibodies. Endocrinology 129,2834-2844.

Sanghera,M.K., Simpson,E.R., McPhaul,M.J., Kozlowski,G., Con!ey,A.J., Lephart,E.D., 199 1b. lmmunocytochemical distribution of aromatase cytochrome P450 in the rat brain using peptide-generated polyclonal antibodies. Endocrinology 129,2834-2844.

Scher,M.S., 1997. Seinires in the newbom infant. Diagnosis, treatment, and outcorne. Clin Perinatol24, 735-772.

Scon,B.W., Wojtowicz,J.M., Burnham,W.M., 2000. Neurogenesis in the dentate gyrus of the rat following electroconvuisive shock seizures [In Process Citation]. Exp Neurol l65,23 1-236.

Selye,H., 1970. Resistance to picrotoxin poisoning induced by catatoxic steroids. Agents Actions 1, 133- 135.

Shank,R.P., Gardocki,J.F., Streeter,A.J., Maryanoff,B.E., 2000. An overview of the preclinical aspects of topiramate: pharmacology, pharrnacokinetics, and mechanism of action. Epilepsia 4 1 Suppl 1, S3-S9.

Shinoda,K., 1994. Brain aromatization and its associated structures. Endocr J 4 1, 1 15- 13 8.

Shinohara,K., Funabashi,T., Mitushirna,D., Kimura,F., 2000. Effects of estrogen on the expression of connexin32 and connexin43 mRNAs in the suprachiasrnatic nucleus of female rats. Neurosci Lett 286, 107-1 IO.

Shorvon,S.D., 1996. The epidem iology and treatment of chronic and refractory epilepsy . €pi lepsia 37 SUPPI2, S 1-S3.

Shughrue,P.J., Lane,M.V., Merchenthaler.1.. 1997. Comparative distribution of estrogen receptor-alpha and -beta mRNA in the rat central nervous system. J Comp Neurol388,507-525. Shughrue,P.J., Lane,M.V., Scrimo,P.J., Merchenthaler,l., 1998a. Comparative distribution of estrogen recepbr-alpha (ER-alpha) mdbettr f ER-beta)mRNEt in-the rat ph-, ganad, and nptoda&e tract. Steroids 63,498-504.

Shughrue,P.J., Scrimo,P.J., Merchenthaler.1.. 1998b. Evidence for the colocaiization of estrogen receptor- beta mRNA and estrogen receptor-alpha immunoreactivity in neurons of the rat forebrain. Endocrinology 139,5267-5270.

Simler,S., Vergnes,M., Marescaux,C., 1999. Spatial and temporal relationships between C-Fos expression and kindling of audiogenic seimres in Wistat rats. Exp Neurol 1 57, 106- 1 19.

Singer,[., SingerJ., 1972. Periodicity of sexual desire in relation to time of ovulation in women. J Biosoc Sci 4,471-48 1.

Sm ith,S.S., 1991. Progesterone administration attenuates excitatory amino acid responses of cerebellar Purkinje cells. Neuroscience 42,309-320.

Smith,S.S., Gong,Q.H., Hsu,F.C., Markowitz,R.S., f'fiench-Mullen,J.M., Li,X., 1998b. GABA(A) receptor alpha4 subunit suppression prevents withdrawal properties of an endogenous steroid [see comments). Nature 392,926-930.

Smith,S.S., Gong,Q.H., Hsu,F.C., Markowitz,R.S., ffrench-Mullen,J.M., Li,X., 1998a. GABA(A) receptor alpha4 subunit suppression prevents withdrawal properties of an endogenous steroid [see comments]. Nature 392,926-930.

Smith,S.S., Gong,Q.H., Li,X., Moran,M.H., Bitran,D., Frye,C.A., Hsu,F.C., 1998c. Withhwal fiom 3alpha-OH-Salpha-pregnan-20-Oneusing a pseudopregnancy mode1 alters the kinetics of hippocampal GABAA-gated current and increases the GABAA receptor alpha4 subunit in association with increased anxiety. J Neurosci 18, 5275-5284.

Snead,O.C., 111, 1998. Ganaxolone, a selective, high-affinity steroid modulator of the gamma- aminobutyric acid-A receptor, exacerbates seimres in animal models of absence. Ann Neurol44,688-69 1.

So,E.L., 1995. Classifications and epidemiologic considerations of epileptic seizures and epilepsy. Neuroimaging Clin N Am 5, 5 13-526.

Spector,S., Cull,C., Goldstein,L.H., 2000. Seizure precipitants and perceived self-control of seizures in adults with poorly-controlled epilepsy. Epilepsy Res 38,207-216,

Stanislaw,H., Rice,F.J., 1988. Correlation between sexual desire and menstrual cycle characteristics. Arch Sex Behav 17,499-508.

Stefan0,G.B.. Cadet,P., Breton,C., Goumon,Y., Prevot,V., Dessaint,J.P., Beauvillain,J.C., Roumier,A.S., Welters,l., Salzet,M., 2000. Estradiol-stimulated nitnc oxide rclease in human granuiocytes is dependent on intracellular calcium transients: evidence of a cell surtace estrogen receptor. Blood 95,395 1-3958.

Steiner,M., 1992. Female-specific mood disorders. Clin Obstet Gynecol35,599-6 1 1.

Stitt,S.L., Kinnard,W.J., 1968a. The effect of certain progestins and estrogens on the threshold of electrically induced seizure patterns. Neurology 18,2 13-2 16.

Stitt,S.L., Kinnard,W.J., 1968b. The effect of certain progestins and estrogens on the threshold of electrically induced seizure patterns. Neurology 18,2 13-2 16. Stunn,G., Haberlein,H., BauerJ., Plaum,T., Stalker,D.J., 199 1. Mass spectrometric and high-performance liquid chromatogragbic sndies of medcoxyprogesterone matemetabotites in hmart plasma. i Chromatogr 562,35 1-362.

Sutula,T., Steward,O., 1986. Quantitative analysis of synaptic potentiation during kindling of the perforant path. J Neurophysiol56.732-746.

Sutula,T., StewardBo.,1987. Facilitation of kindling by prior induction of long-term potentiation in the perforant path. Brain Res 420, 109- 1 17.

Sutula,T.P., Golarai,G., Cavazos,J., 1992. Assessing the tirnctional significance of mossy fiber sprouting. Epilepsy Res Suppl7,251-259.

Suzuki,S., Rogawski,M.A., 1989. T-type calcium channels mediate the transition between tonic and phasic firing in thalarnic neurons. Proc Natl Acad Sci U S A 86,722807232.

Swoboda,K.J., Drislane,F. W.. 1994. Seizure disorders: syndromes, diagnosis, and management. Compr Ther 20.67-73.

Szabo,M., Kilen,S.M., Nho,S.J,, Schwartz,N.B., 2000. Progesterone receptor A and B messenger ribonucleic acid levels in the anterior pituitary of rats are regulated by estrogen. Biol Reprod 62,9S- 102.

Tanapat,P., Hastings,N.B., Reeves,A.J., Gould,E., 1999. Estrogen stimulates a transient increase in the number of new neurons in the dentate gynis of the itdult female rat. J Neurosci 19,5792-5801.

Tauboll,E., Lindstrom,S., 1993. The effect of progesterone and its metabolite 5 alpha-pregnan-3 alpha- ol- 20-one on focal epileptic seizures in the cat's visual cortex in vivo. Epilepsy Res 14, 17-30.

Teter,B., Harris-White,M.E., Frautschy,S.A., Cole,G.M., 1999. Role of apolipoprotein E and estrogen in mossy fiber sprouting in hippocampal slice cultures. Neuroscience 9 1, 1009- IO 16.

Thavendiranathan,P., Mendonca,A., Dell,C., Likhodii,S.S., Musa,K., Iracleous,C., Cunnane,S.C., Bumham,W.M., 2000. The MCT ketogenic diet: effects on animal seizure models. Exp Neurol 161,696- 703.

Ticku,M.K., Olsen,R.W., 1978. Interaction of barbiturates with dihydropicrotoxinin binding sites related to the GABA receptor-ionophore system. Life Sci 22, 1643- 1651.

Traub,R.D., SchmitqD., Jefferys,J.G., Draguhn,A., 1999. High-frequency population oscillations are predicted to occur in hippocampat pyramidal neuronal networks interconnected by axoaxonal gap junctions. Neuroscience 92,407-426.

Velisek,L., Kubova,H., Pohl,M., Stankova,L., Mares,P., Schickerova,R., 1992. Penty lenetetrazol-induced seizures in rats: an ontogenetic study. Naunyn Schmiedebergs Arch Pharmacol 346,588-591.

Wada,J.A., The clinical relevance of kindling: species, brain sites and seizure susceptibility. pp. 369-88. In: Livingston KE , Hornykiewicz O, ed Limbic mechanisms New York , Plenum Press, 1978 WL 3 14 L733L 1976.

Wada,J.A., Mizoguchi,T., Osawa,T., 1978. Secondarily generalized convulsive seizures induced by daily amygdaloid stimulation in rhesus monkeys. Neurology 28, 1026- 1036.

Wahnschaffe,U., Loscher, W., 1992. Lack of changes in seizure susceptibility during the estrous cycle in kindled rats. Epilepsy Res 13, 199-204. Wang& Scott,B.W., Wojtowicz,J.M., 2000. Heterogenous properties of dentate granule neurons in the rtduk rat. k Ncurobiot 42,248-257.

Warren.S.G., Humphreys,A.G., Juraska,J.M., Greenough,W.T., 1995. LTP varies across the estrous cycle: enhanced synaptic plasticity in proestrus rats. Brain Res 703.26-30.

Weiland,N.G., Orikasa,C., Hayashi,S., McEwen,B.S., 1997. Distribution and hormone regulation of estrogen receptor immunoreactive cefls in the hippocampus of male and female rats. J Comp Neurol388, 603-6 12.

Williams,R.F., Barber,D.L., Cowan,B.D., Lynch,A., Manit, EL., Hodgen,G.D., 198 1. Hyperprolactinemia in monkeys: induction by an estrogen-progesterone synergy. Stetoids 38,32 1-33 1.

Willow,M., Canerall,W.A., 1982. lnhibition of binding of [3H]batrachotoxinin A 20-alpha-benzoate to sodium channels by the anticonvulsant drugs diphenylhydantoin and carbarnazepine. Mol Pharmacol 22, 627-635.

Wong,M,, Thompson,T.L., Moss,R.L., 1996. Nongenomic actions of estrogen in the brain: physiological significance and cellular mechanisms. Crit Rev Neurobiol 10, 189-203.

Woolley,C.S., 1998. Estrogen-mediated structural and hnctional synaptic plasticity in the female rat hippocampus. Hom Behav 34, 140- 148.

Woolley,C.S., McEwen,B.S., 1993. Roles of estradiol and progesterone in regulation of hippocampal dendritic spine density during the estrous cycle in the rat. J Comp Neurol336,293-306.

Woolley,C.S., McEwen,B.S., 1994. Estradiol regulates hippocampal dendritic spine density via an N- methyl- D-aspartate receptor-dependent mechanism. 1 Neurosci 14,7680-7687.

Woolley,C.S., Schwartzkroin,P.A., 1998. Hormonal effects on the brain. Epilepsia 39 Suppl 8, 52-58.

Wu,F.S., Gibbs,T.T., Farb,D.H., 1991. Pregnenolone sulfate: a positive allosteric modulator at the N- methyl-D- aspartate receptor. Mol Pharmacol 40,333-336.

Xiong,F., Liu,D., Le Drean,Y., Elsholtz,H.P., Hew,C.L., 1994. Differential recruitment of steroid hormone response elements may dictate the expression of the pituitary gonadotropin II beta subunit gene during salmon maturation. Mol Endocrinol8,782-793.

Y ukawa,E., 2000. lnvestigation of phenobarbital-carbarnazepine-valproic acid interactions using population pharrnacokinetic analysis for optimisation of antiepi leptic drug therapy :an overview. Drug Metabol Drug lnteract 16, 86-98.

Zhou,Y., Watters,J.J., Dorsa,D.M., 1996. Estrogen rapidly induces the phosphorylation of the CAMP response element binding protein in rat brain. Endocrinoiogy 137.2 163-2 166.

Zimmerrnan,A. W., Holden,K.R., Reiter,E.O., Dekaban,A.S., 1973. Medroxyprogesterone acetate in the treatment of seizures associated with menstruation. The Journal of Pediatrics 83,959-963.

Zwain,l.H., Yen,S.S., 1999. Neurosteroidogenesis in astrocytes, oligodendrocytes, and neurons of cerebral cortex of rat brain. Endocrinology 140, 3843-3552.