PDF hosted at the Radboud Repository of the Radboud University Nijmegen

The following full text is a publisher's version.

For additional information about this publication click this link. http://hdl.handle.net/2066/71267

Please be advised that this information was generated on 2021-09-26 and may be subject to change. Frank van Broekhoven Effects of and on stress, attention, cognition and mood | Frank van Broekhoven

ISBN 978-90-9023655-1

Copyright ©2008 Frank van Broekhoven. The copyright of articles that have already been published has been transferred to the respective journals. No part of this book may be reproduced, in any form, without prior written permission from the author. Niets uit deze uitgave mag worden verveelvoudigd en/of openbaar gemaakt in welke vorm dan ook, zonder voorafgaande schriftelijke toestemming van de auteur.

Coverdesign and layout by: Communicatie Kant, Dinxperlo, The Netherlands Printed by: Up2data, Bocholt, Germany

The financial support for the printing of this thesis by Eli Lilly Nederland BV, Janssen-Cilag BV, the Department of Psychiatry from the Radboud University Nijmegen Medical Centre, and Karakter, Child and Adolescent Psy- chiatry University Centre, Nijmegen, is gratefully acknowledged. Effects of progesterone and allopregnanolone on stress, attention, cognition and mood

Een wetenschappelijke proeve op het gebied van de Medische Wetenschappen

Proefschrift

ter verkrijging van de graad van doctor aan de Radboud Universiteit Nijmegen op gezag van de rector magnificus prof. mr. S.C.J.J. Kortmann, volgens besluit van het College van Decanen

in het openbaar te verdedigen op maandag 24 november 2008 om 15.30 uur precies

door

Frank van Broekhoven geboren op 8 december 1969 te Groenlo Promotores: prof. dr. J.K. Buitelaar prof. dr. T. Bäckström (Umeå Universiteit)

Copromotor: dr. R.J. Verkes

Manuscriptcommissie: Prof. dr. Y.A. Hekster Prof. dr. J.A. den Boer (Rijksuniversiteit Groningen) Prof. dr. J.A.M. Kremer Effects of progesterone and allopregnanolone on stress, attention, cognition and mood

An academic assay in Medical Sciences

Doctoral thesis

to obtain the degree of doctor from Radboud University Nijmegen on the authority of the Rector Magnificus prof. dr. S.C.J.J. Kortmann, according to the decision of the Council of Deans to be defended in public on Monday, 24 November 2008 at 15.30 hours

by

Frank van Broekhoven born in Groenlo on 8 december 1969 Doctoral supervisors: prof. dr. J.K. Buitelaar prof. dr. T. Bäckström (Umeå University)

Co-supervisor: dr. R.J. Verkes

Doctoral Thesis Committee: Prof. dr. Y.A. Hekster Prof. dr. J.A. den Boer (University of Groningen) Prof. dr. J.A.M. Kremer Effects of progesterone and allopregnanolone on stress, attention, cognition and mood | Frank van Broekhoven

Contents

page Chapter 1 Introduction 9

Chapter 2 Pharmacological characteristics of progesterone and allopregnanolone, nomenclature, and development of neuroactive to drugs 21

Chapter 3 Allopregnanolone and serotonin 33

Chapter 4 in depression: a review 39

Chapter 5 Effects of PhD examination stress on allopregnanolone and plasma levels and peripheral receptor density 59

Chapter 6 Effects of allopregnanolone on sedation in men, and in women on oral contraceptives 67

Chapter 7 Oral progesterone decreases saccadic eye velocity and increases sedation in women 85

Chapter 8 Increased sensitivity to oral progesterone in the luteal phase in healthy women 103

Chapter 9 Progesterone selectively increases amygdala reactivity in women 117

Chapter 10 Discussion 133

Chapter 11 Summary 147

Chapter 12 References 153

Samenvatting 181

Publications 187

Dankwoord 193

Curriculum vitae 197 Effects of progesterone and allopregnanolone on stress, attention, cognition and mood | Frank van Broekhoven

Chapter 1 | INTRODUCTION Chapter 1 | INTRODUCTION

Chapter 1 | INTRODUCTION

Neuroactive steroids are important for psychiatry. Being metabolites of sex and stress hormones, they are involved in the pathophysiology of menstrual cycle and stress linked psychiatric disorders like the Premenstrual Dysphoric Disorder (PMDD), depression and dementia. Neuroactive steroids influence the excitability of neurons by acting on membrane-bound neurotransmitter receptors. Changes in concentrations of neuroactive steroids or in sensitivity of receptors can cause psychiatric symptoms like negative mood, anxiety and cognitive disor- ders. Most data comes from animal research while human data is mostly limited to the assessment of levels of neuroactive steroids in plasma from psychiatric patients and healthy controls and from postmenopausal women during hormone replacement therapy. The latter model has the advantage that dose response studies have been performed in humans.

Definitions

Steroid hormones are synthesized in the gonads, adrenal glands and placenta. Some steroid hormones are also synthesised in the brain and are called neurosteroids (Baulieu, 1998). The synthesis of steroids within the brain was first discovered in 1981 when -sulphate (DHEAS) was detected in the brain of rats in concentrations much more higher than in plasma and not being influenced by adrenalectomy or adreno- corticotropic hormone (ACTH) administration (i.e. by manipulation peripherally) (Corpechot et al., 1981). When steroid hormones or neurosteroids have effect on the central nervous system (CNS) they are called neuroactive steroids (Paul and Purdy, 1992).

Steroid hormones are known to act through intracellular receptors that act as ligand-activated transcrip- tion factors in the regulation of gene expression (genomic way). Neuroactive steroids have effects through membrane-bound receptors such as ligand-gated ion channels (non-genomic way) (McEwen, 1991; Rupprecht, 2003). This a-genomic mechanism of action accounts for rapid steroid effects that occur within milliseconds to seconds. Although neuroactive steroids can act on many different membrane-bound receptors, neuroactive steroids that act on the GABA(A) receptor are among the best studied (Baulieu, 1998). These so called GABA- steroids are defined as 3alpha-hydroxy-5alpha/beta metabolites of endogenous steroid hormones (Rupprecht and Holsboer, 1999). The main endogenous steroid hormones that generate GABA-steroids are stress hormones (cortisol and desoxycorticosterone (DOC)), and sex hormones (progesterone and ). Many of the CNS effects of stress and sex hormones are mediated via the 3alpha-hydroxy-5alpha/beta metabolites while these GABA-steroids themselves do not activate the classical hormonal nuclear receptors of their parent compounds (Backstrom et al., 2003).

11 Chapter 1 | INTRODUCTION

GABA, GABA(A) receptor, and GABA-steroids

Gamma-amino butyric acid (GABA) is the major inhibitory neurotransmitter in the CNS and activates the GABA(A) receptor. The transmembrane heteropentameric GABA(A) receptor is composed of five subunits that belong to different subunit classes (Rudolph et al., 2001). These subunits form an ion channel through which chloride ions can enter the cell. When chloride ions flow into the neuron the possibility to generate an action potential is inhibited. Opening and closing of the channel is regulated by substances that bind to the GABA(A) receptor such as GABA itself, , and GABA-steroids. GABA-steroids enhance the GABA effect on the GABA(A) receptor in a similar way as , benzodiazepines and alcohol and are therefore potent positive allosteric modulators of the GABA(A) receptor (Majewska et al., 1986). The concentrations of GABA-steroids in the brain change in parallel with the ovarian and adrenal production. Conditions in which high GABA-steroid production occurs are stress, the luteal phase of the menstrual cycle, preg- nancy, progestagen treatment (oral contraceptive use, hormonal replacement treatment), and steroid treatment. GABA-steroids and GABA(A) receptors interact with each other. The sensitivity of GABA(A) receptors to GABA- steroids depends on the subunit composition of the receptor, with receptors containing the delta-subunit being the most sensitive (Spigelman et al., 2003), and on the phosphorylation state of the receptor (Lambert et al., 2001). Subunit composition differs between brain regions (Gee et al., 1988; Lambert et al., 2001). Vice versa, GABA-steroids influence the subunit composition of the GABA(A) receptor (Follesa et al., 2004); withdrawal of GABA-steroids increase alpha 4 subunit resulting in decreased sensitivity to GABA-steroids (Smith et al., 1998) while exposure to GABA-steroids increase alpha 1 and 2 subunits (Birzniece et al., 2006; Guerra-Araiza et al., 2008). Thus, the interaction between GABA-steroids and GABA(A) receptors is complex. GABA-steroids can induce CNS disorders (Backstrom et al., 2003). These disorders can be divided into menstrual cycle linked disorders (for example negative mood during hormone replacement therapy, PMDD, and catamenial epilepsy, which is a form of epilepsy characterised by increased frequency of epileptic seizures) and acute and chronic stress disorders (for example depression). In addition, GABA-steroids induce symptoms like cognitive disturbances (impaired memory and learning (Johansson et al., 2002)), increased appetite (Chen et al., 1996), and increased risk of substance abuse (Finn et al., 1997). They also induce negative mood symp- toms and aggressiveness (Backstrom et al., 2003; Miczek et al., 2003). Long-term treatment with precursors to GABA-steroids increases the risk of developing dementia (Shumaker et al., 2003). Similar to benzodiazepines, GABA-steroids can produce tiredness or sedation, and the development of tolerance leading to rebound effects at withdrawal. After a long exposure to GABA-steroids the GABA-system becomes dysfunctional by down regu- lating the GABA(A) receptors (Yu and Ticku, 1995) which is a state similar to long-term benzodiazepine use. The pathophysiological mechanisms that are responsible for the GABA-steroid induced effects are direct action of GABA-steroids on GABA(A) receptors, induction of tolerance for GABA-steroids, and withdrawal of GABA-steroids (Backstrom et al., 2003).

12 Chapter 1 | INTRODUCTION

Allopregnanolone

Allopregnanolone, which is one of the main circulating metabolites of progesterone, is one of the most well-documented examples of those neuroactive steroids that interact with the GABA(A) receptor complex (Lan and Gee, 1994). It is a positive allosteric modulator of the GABA(A) receptor (Majewska et al., 1986). Allopreg- nanolone potentiates the action of GABA by prolonging the mean life time of the chloride channel open state (Prince and Simmonds, 1993), increases the duration of opening of the chloride channel in a manner similar to barbiturates and at high doses increases the frequency of opening of the chloride channel in a manner similar to benzodiazepines (Twyman and Macdonald, 1992). The exact site on the GABA(A) receptor to which allopregna- nolone was thought to bind to was always believed to be distinct from that of barbiturates and benzodiazepines (Peters et al., 1988; Paul and Purdy, 1992; Lambert et al., 1995; Brot et al., 1997; Wang et al., 2000). The possibil- ity of multiple binding sites on a single GABA(A) receptor or distinct sites on different GABA(A) isoreceptors was also considered (Morrow et al., 1990; Prince and Simmonds, 1993). Later the exact binding site was discovered: two distinct sites on the GABA(A) receptor (Hosie et al., 2006). For a small proportion, allopregnanolone can be entirely synthesized de novo within the brain from (Paul and Purdy, 1992; Marx et al., 2000a). It can also arise within the central or peripheral nervous system from metabolization of peripherally (adrenal medulla and ovary) produced steroid hormones (Calogero et al., 1998; Baulieu and Schumacher, 2000; Marx et al., 2000b). The positive modulation of the GABA(A) receptor by allopregnanolone leads to sedative, hypnotic, anaesthetic, anxiolytic and anticonvulsant effects in animals (Arafat et al., 1988; Bitran et al., 1991; Frye and Duncan, 1994; Lancel et al., 1997; Reddy et al., 2004).

Allopregnanolone, stress, and the peripheral benzodiazepine receptor

Acute stress, as induced by swimming in ambient temperature water (Purdy et al., 1991), inhalation of CO2 (Barbaccia et al., 1994; Barbaccia et al., 1996), and handling in naive rats or mild foot shock in handling-habitu- ated rats (Barbaccia et al., 1997), increases plasma and brain concentrations of allopregnanolone in animals. Chronic stress, however, decreases these baseline concentrations (Serra et al., 2000). The increase in allopreg- nanolone concentration in response to acute stress, however, is maintained during chronic stress (Serra et al., 2007). No data exists about alterations in allopregnanolone concentration following stress in humans. Stress has also influence on the peripheral benzodiazepine receptor (PBR) density in humans. The PBR plays an important role in the synthesis of neuroactive steroids (Papadopoulos et al., 2006). Situated on the outer membrane of the mitochondrion, the PBR translocates steroid precursors from outside to inside this or- ganell. In the mitochondrion, the steroid precursor is then metabolised into neuroactive steroids. Several clinical studies have been conducted on PBR density, indicating that acute stress increases and chronic stress decreases PBR density. Examination stress, forced swimming and surgical stress in rats, are associated with increased PBR

13 Chapter 1 | INTRODUCTION

density (Karp et al., 1989; Dar et al., 1991; Weizman et al., 1993; Weizman et al., 1995; Gavish et al., 1996; Soreni et al., 1999; Leschiner et al., 2000; Ritsner et al., 2003). Long-term anxiety, parachute training course, Post- traumatic Stress Disorder (PTSD), adolescents with repeated suicide attempts, are associated with decreased PBR density (Weizman et al., 1987; Dar et al., 1991; Gavish et al., 1996; Soreni et al., 1999). Depression, PMDD, and obsessive-compulsive disorder do not show any difference in PBR density with controls (Weizman et al., 1993; Weizman et al., 1995; Daly et al., 2001). Because of its relation with neuractive steroids, the PBR is suggested as a potential target for new anxiolytic drugs (Kita et al., 2004). Acute or chronic stress often precedes psychiatric disorders and it has been shown that patients suffer- ing from affective disorders have a dysregulation of the hypothalamic-pituitary-adrenal (HPA) axis (Owens and Nemeroff, 1991). The HPA-axis is activated during stress leading to elevated plasma levels of cortisol. For an effective response to stress, activation of the HPA-axis has to be followed by deactivation. In some conditions, however, the subect may have increased and/or prolonged activation of the HPA-axis due to an absent switch- off of the HPA-axis to normal state. When the HPA-axis remains in a state of overactivity, it will lead to negative effects. In depressed patients, for example, the negative feedback of cortisol on the hypothalamus and pituitary gland is disrupted and concentrations of corticotropin-releasing hormone (CRH), ACTH, and cortisol remain el- evated in the absence of the preceding stress factor (Holsboer, 2001). Allopregnanolone attenuates CRH-release and CRH-gene expression (Patchev et al., 1994) and also also decreases ACTH concentrations (Patchev et al., 1996). Vice versa, administration of CRH increases allopregnanolone concentrations (Genazzani et al., 1998). Taken into account that allopregnanolone plasma levels increase following acute stress, it is thought that allo- is formed as a calming compensatory response to stress. The role of allopregnanolone in stress- linked psychiatric disorders like depression deserves therefore further attention.

Allopregnanolone and emotion regulation

It has been shown in several animal models of anxiety that allopregnanolone has an anxiolytic effect. The amygdala is important in this effect (Akwa et al., 1999). The amygdala plays a crucial role in emotional process- ing. It is important for the identification of the emotional significance of stimuli and the response to it. The amygdala is part of a larger emotion circuitry and connected to other important centres such as the anterior cingulate gyrus, the fusiform gyrus and the prefrontal cortex. Anxiety is a symptom that is frequently seen in psychiatric patients and there are some psychiatric disor- ders in which anxiety is the core symptom, the so called anxiety disorders. Although in mixed anxiety-depressive disorder (Bicikova et al., 2000), generalised anxiety disorder (Semeniuk et al., 2001), and generalised social pho- bia (Heydari and Le Melledo, 2002), allopregnanolone plasma concentrations were not different from healthy controls, patients with panic disorder have significantly higher plasma concentrations than healthy controls which decrease at induction of panic attacks (Strohle et al., 2003). Dysregulation of allopregnanolone levels are

14 Chapter 1 | INTRODUCTION

suggested to be a part of the pathophysiology of panic disorder as healthy controls did not show a decrease in plasma allopregnanolone concentrations during the induced panic attacks (Strohle et al., 2003). The knowledge about the role of allopregnanolone in emotion regulation seems thus far to be restricted to anxiety disorders. In PMDD, a disorder characterised by negative emotions during the premenstrual phase, however, patients show a reduced sensitivity to pregnanolone, the stereo- of allopregnanolone, due to a dysfunction of the GABA system. Dysfunction of the GABA(A) receptor is part of the pathophysiology that is shared with panic disorder; both patients with PMDD and patients with panic disorder show a shift in sensitivity to flumazenil toward inverse agonism leading to panic attacks in response to flumazenil (Nutt et al., 1990; Le Melledo et al., 2000). More research on the influence of allopregnanolone in emotional processing is recom- mended.

Allopregnanolone and depression

Protracted social isolation in mice induces chronic stress and is followed by a behavioural adaptation syndrome, which shares some features of depression. Cerebral cortex levels of allopregnanolone have been shown to be decreased in this animal model of depression (Serra et al., 2000), due to a decreased activity of the 5alpha-reductase enzyme (Dong et al., 2001). Olfactory bulbectomy, which results in several anatomical, physiological and behavioural abnormalities similar to those reported in depression in human patients, induces also decreased levels of allopregnanolone in brain (Uzunova et al., 2003). After three weeks of treatment with antidepressants of three different classes (desipramine, , sertraline, and venlafaxine), these levels normalized (Uzunova et al., 2004). It is suggested that the increase in allopregnanolone by fluoxetine is due to a change of the activity of the enzyme 3alpha-hydroxysteroiddehydrogenase (HSD) (or 3alpha-hydroxyster- oidoxidoreductase (HSOR)), to the reductive site (Griffin and Mellon, 1999). The decreased GABA(A) receptor sensitivity in patients with PMDD is normalized after citalopram treatment suggesting a different mechanism of action by the antidepressants (Sundstrom and Backstrom, 1998). In humans, plasma (Romeo et al., 1998) and cerebro spinal fluid (Uzunova et al., 1998) levels of allopreg- nanolone were found to be decreased in depressed patients. Pharmacological treatment with antidepressants increased allopregnanolone levels in both studies. To investigate whether this normalization of decreased levels of allopregnanolone is part of the improvement of depressive symptoms or merely due to the pharmacological effect, the Max Planck Institute in Munich, Germany, performed three studies in which depressed patients were treated non-pharmacologically: partial sleep deprivation (Schule et al., 2003), repetitive transcranial magnetic stimulation (Padberg et al., 2002), and electroconvulsive therapy (Baghai et al., 2005). The conclusion of all these three studies was that there were no alterations in allopregnanolone levels following treatment. The observed effects on allopregnanolone levels following pharmacological antidepressant treatment is due to spe- cific pharmacological properties of antidepressant drugs (Eser et al., 2006). Normalization of allopregnanolone

15 Chapter 1 | INTRODUCTION

levels is neither required nor sufficient for clinical response. This was later supported by the finding that follow- ing five weeks of treatment with mirtazapine allopregnanolone levels increased in depressed patients both in responders and non-responders (Schule et al., 2006). The increase in the allopregnanolone levels was abolished when mirtazapine treatment was combined with lithium or (Schule et al., 2007).

Allopregnanolone and PMDD

Even though premenstrual symptoms had been already described by Hippocrates, PMDD was only first mentioned as a special psychiatric diagnosis in the Diagnostic and Statistical Manual of Mental Disorders (DSM- IV) in 1994 (Association, 1994). In this edition only research criteria for PMDD are given with the aim that results of further research will validate the inclusion of PMDD as an official DSM diagnosis. The syndrome itself has been described with different names before. In DSM-III-R-Appendix A it was called late luteal phase dysphoric disorder (LLPDD) (American Psychiatric Association, 1987). Before this diagnosis was established as a DSM disorder, the term premenstrual syndrome (PMS) was used for patients with severe premenstrual mood disturbances and physical symptoms (Lenzinger et al., 1997). The term PMS is still generally used in medicine to indicate serious premenstrual complaints but that do not reach the extent of severity as in PMDD. Another classification system than the DSM, the International Statistical Classification of Diseases and Related Health Problems 10th Revision (ICD-10), contains the diagnosis premenstrual tension syndrome to indicate serious premenstrual symptoms (World Health Organization, 2004). A prevalence rate for PMDD of 3-5% is mentioned in the DSM-IV (Associa- tion, 1994), and is confirmed by others (Gehlert and Hartlage, 1997). During the normal menstrual cycle the plasma levels of progesterone fluctuate. In the follicular phase there is a low concentration. After the ovulation the luteal phase starts and a corpus luteum is formed that pro- duces progesterone, and via metabolisation allopregnanolone, in high amounts. A few days before menstruation the progesterone concentration starts to decrease. In PMDD it is remarkable that symptoms of negative mood and irritability start after ovulation, increase in parallel with progesterone and allopregnanolone concentrations and continue to grow even after levels of progesterone and allopregnanolone have started to decrease. This indicates that there is a direct effect of allopregnanolone on responsible receptors in the brain first, followed by development of tolerance to allopregnanolone, and finally withdrawal of allopregnanolone starts. In the past it was believed that differences in plasma concentrations of progesterone and allopregnanolone between women with PMDD and normal women were the cause of symptom development. However, contradict- ing information has been reported concerning neuroactive steroid plasma concentrations in PMDD patients and controls. Some studies in PMDD patients found significantly lower allopregnanolone levels in the follicular phase (Bicikova et al., 1998) or in the luteal phase compared to control subjects (Rapkin et al., 1997; Bicikova et al., 1998; Monteleone et al., 2000). In other studies, no statistically significant differences in allopregnanolone levels were found (Wang et al., 1996; Sundstrom and Backstrom, 1998b). One study found higher allopregnanolone

16 Chapter 1 | INTRODUCTION

levels in the luteal phase of patients with PMDD (Girdler et al., 2001). Thus, assessment of plasma concentrations seem to be of little help in explaining the cause of PMDD. Moreover, actual allopregnanolone concentrations in the brain may differ from the periphery and can only be measured in liquor by lumbar puncture (Uzunova et al., 1998). Research at the receptor level is more promising. There might be a difference in sensititvity to allo- pregnanolone due to, for example, changed subunit composition of the GABA(A) receptor (Grobin and Morrow, 2000). Nowadays, a dysfunctional GABA system, i.e. a changed sensitivity of the GABA(A) receptor, is suggested to be part of the pathophysiology which suggestion is supported by the observation of decreased sensitivity to GABAergic substances such as benzodiazepines, alcohol and pregnanolone (Sundstrom et al., 1997b; Sundstrom et al., 1998; Nyberg et al., 2004), and a shift in benzodiazepine sensitivity toward inverse agonism in women with PMDD (Le Melledo et al., 2000). A third observation that indicates a changed GABA(A) receptor function is that the negative mood and irritability in women with PMDD is related to the increase in allopregnanolone during the menstrual cycle. Normally, allopregnanolone induces anxiolysis and sedation but in these women the opposite occurs. It is suggested that allopregnanolone has a bimodal effect in women with PMDD, i.e. that these women respond to low concentrations of allopregnanolone with anxiety and irritability and to higher concentrations with positive feelings. Such a bimodal effect has been demonstrated in postmenopausal women during hormo- nal therapy with progesterone. Their negative mood aggravated with increasing allopregnanolone concentra- tions in the beginning but improved with further increase of these concentrations later (Andreen et al., 2005; Andreen et al., 2006b) Women with PMDD show increased amygdala response to negative stimuli (emotional words) in the luteal phase compared to the follicular phase (Protopopescu et al., 2008). Again, this enhanced negative emotional processing in the luteal phase could be due to a different sensitivity of the GABA(A) receptor to allopregnanolone in this phase of the cycle.

Aims of this thesis

Although it has repeatedly been shown that allopregnanolone is increased after stress in animals, the ef- fect of stress on plasma levels of allopregnanolone in humans has not been investigated. Further, although it has been shown that PBR density is increased after stress in humans, the relationship between plasma levels of allopregnanolone and PBR density is not known. It has been consistently shown that administration of allopregnanolone in animals leads to sedation, anxiolysis and anticonvulsion. Human in-vivo data is missing because until recently there was no pharmaceutical formula- tion available for human use. We administered a newly available pharmaceutical formulation for intravenous administration in humans. Until now, the effect of the phase of the menstrual cycle on the sensitivity to progesterone is not known. It has been shown that PMDD patients have a different sensitivity to GABAergic substances like , and pregna- nolone across the cycle compared to healthy controls (Sundstrom et al., 1997a; Sundstrom et al., 1997b; Sundstrom

17 Chapter 1 | INTRODUCTION

et al., 1998). The effect of neuroactive steroids on the functional activity in brain function is not known. Several find- ings suggest that allopregnanolone is involved in the modulation of the amygdala. Infusion of GABA agonists into the amygdala induces anxiolytic effects (Davis and Whalen, 2001b). In the brain of healthy women there are significant changes in functional magnetic resonance imaging (fMRI) responses during the menstrual cycle which are related to, among other hormones, progesterone variations (Fernandez et al., 2003). Progesterone and allopregnanolone are associated with a decrease in amygdala reactivity during the intentional encoding of neutral and happy faces into memory (van Wingen et al., 2007). As the amygdala is a part of the brain that is related to emotional experiences and allopregnanolone is involved in psychiatric disorders, it is important to study the effect of progesterone on the response of the amygdala to emotional stimulation. There are many different neuroactive steroids and they have many different effects via a number of dif- ferent neurotransmitter receptors. In this thesis we have focused on the neuroactive steroid allopregnanolone and its effect on the GABA(A) receptor. As the available data on allopregnanolone is mostly limited to animal research, we have decided to investigate the effects of allopregnanolone in humans. The overall aim of this the- sis is to address this gap in knowledge and to investigate the GABAergic effects of allopregnanolone in men and women. Secondary aims were:

− to investigate the influence of the menstrual cycle on these effects in women − to test the hypothesis that in acute stress plasma allopregnanolone concentrations increase in men and women together with an increase in PBR density To investigate the GABAergic effects in humans it was important to administer allopregnanolone in hu- mans. At the start of the PhD project, however, there was no allopregnanolone formulation for human use available. Therefore we started challenge studies with the precursor progesterone, which is a technique applied by other researchers earlier. In these studies, progesterone as well as allopregnanolone plasma concentrations were measured and related to effect parameters. Meanwhile an allopregnanolone formulation for human use had become available. We were able to administer it in only one study given the time constraints of this thesis project.

Outline of the study

The thesis starts with an outline about the pharmacological characteristics of progesterone and allopreg- nanolone and the relation between allopregnanolone and serotonin (chapters 2 and 3). It proceeds with a review in which the role of neuroactive steroids in depression, with emphasis on allopregnanolone, is discussed in detail (chapter 4). In chapter 5 the effect of stress on allopregnanolone plasma levels and PBR density is assessed. We assumed that PhD examination could serve as a model for stress. PBR density on platelets was used as a marker for central PBR density. In chapters 6, 7, and 9 the GABAergic effect of exogenous allopregnanolone (chapter

18 Chapter 1 | INTRODUCTION

6) and progesterone (chapters 7 and 9) in humans is assessed by measurement of saccadic eye velocity (SEV) (chapters 6 and 7), electroencephalography (EEG) (chapter 6), reaction time (chapter 7), self-reports of mood, and fMRI (only in chapter 9) in relation to allopregnanolone levels. A saccade is a rapid eye movement from one target to another. SEV is the speed by which the eye moves from one target to another. Once a saccade has started, it is generally regarded to be outside conscious control and not subjected to motivational influences (Gentles 1971), and therefore SEV is considered to provide an objective measure of sedation. To investigate GABA(A) receptor sensitivity, SEV in response to sedative drugs effective on the GABA(A) receptor can be meas- ured (Hommer 1986). Several studies have shown that SEV is reduced in a dose-dependent manner by benzodi- azepines (Hommer et al., 1986) and this effect is reversed by the benzodiazepine antagonist flumazenil (Ball et al., 1991). The variability in SEV between individuals is large, but the intra-individual variability, both between test days and within a testing session, is generally low (Gentles and Thomas, 1971; Mercer et al., 1990; Roy-Byrne et al., 1990; Glue, 1991; Sundström and Bäckström, 1998b). Furthermore, benzodiazepine- or pregnanolone-induced increase in self-ratings of sedation is highly correlated with reduction in SEV (Hommer 1986, Sundström 1998a). Benzodiazepines increase sigma and beta power on pharmaco-EEG recordings and the same has been shown for allopregnanolone in animals (Lancel et al., 1997b). The benzodiazepine has been shown to decrease amygdala activity dose-dependently during emotion processing (Paulus et al., 2005). Recordings of SEV, self- ratings of sedation, pharmaco-EEG, and fMRI of amygdala are used in this thesis to measure the GABAergic effects of neuroactive steroids. In chapter 8 the influence of the phase of the menstrual cycle on the effects of oral progesterone in women is examined. The effect of oral progesterone on the CNS in the follicular phase is, placebo-controlled, assessed by measurement of SEV, reaction time, and self-reports of mood, in relation to allopregnanolone levels and compared with that obtained in the luteal phase. Chapter 10 is dedicated to a discussion of the results and to the question if the aims as formulated are fulfilled, and outlines avenues for further research. Chapter 11 is a summary of the thesis.

19 Chapter 1 | INTRODUCTION

Chapter 2 | Pharmacological characteristics of Progesterone and allopregnanolone, steroid nomenclature, and development of neuroactive steroids to drugs

20 Chapter 2 | Pharmacological characteristics of Progesterone and allopregnanolone, steroid nomenclature, and development of neuroactive steroids to drugs

Chapter 2 | Pharmacological characteristics of progesterone and allopregnanolone, steroid nomenclature, and development of neuroactive steroids to drugs

Progesterone

Progesterone (C21H30O2, molecular weight: 314.5) is a female which is also pharmacologically used in, among others, cycle disorders and as postmenopausal substitute.

Pharmacokinetic characteristics of progesterone

60% of oral progesterone is absorbed by the intestines as progesterone and metabolites. The bioavailabil- ity is 6-10% due to a significant ‘first-pass’ effect. The maximum blood concentration is achieved after 1-4 hours. The main metabolites are active 20alpha-dihydroxyprogesterone and inactive (5alpha-- 3alpha,20alpha-diol). A significant proportion of the metabolites follow a enterohepatic cycle. Progesterone is mainly eliminated via the urine (Commissie Farmaceutische Hulp van het College voor Zorgverzekeringen, 2006). Progesterone has a short half-life, about 30 minutes (Organon, personal comunications). Metaboliza- tion occurs successively in the gut (intestine bacteria with 5beta-reductase activity), intestinal wall (which has 5alpha-reductase activity and initiates conjugation with glucuronic acid at carbon atom C3) and liver (5beta- reductase, 3alpha- and 20alpha-hydroxylase and also conjugation with glucuronic acid at C3). Only a fraction of the native steroid eludes these enzymes and circulates in the plasma as remaining progesterone. However, micronization of progesterone, by speeding up its absorption, significantly increases the plasma level of the remaining progesterone (Hargrove et al., 1989; Simon et al., 1993; McAuley et al., 1996). Most of the metabo- lites circulate as inactive pregnanediol (5beta-pregnane-3alpha,20alpha-diol) and are subsequently excreted in the urine as conjugated pregnanediol (Adlercreutz and Martin, 1980; Ganong, 1991; de Lignieres et al., 1995). However, also pregnanolone and allopregnanolone are formed (5beta-pregnane-3alpha-ol-20-one and 5alpha- pregnane-3alpha-ol-20-one, respectively). There is a significant interindividual variability in these metabolites formed after administration of progesterone (Arafat et al., 1988; Freeman et al., 1993). These metabolites are ex- creted as or as 3alpha,6alpha-dihydroxy-5alpha-pregnan-20-one glucuroniside. About 2% of the circulating progesterone is free, whereas 80% is bound to albumin and 18% is bound to corticoster- oid-binding globulin (Ganong, 1991).

23 Chapter 2 | Pharmacological characteristics of progesterone and allopregnanolone, steroid nomenclature, and development of neuroactive steroids to drugs

Summary of known and potential risks of progesterone

Side effects of acute progesterone administration are short-lasting sleepiness or dizziness with onset shortly after administration. This is probably due to certain metabolites of progesterone some of which have anaesthetic properties (e.g., allopregnanolone, pregnanolone, tetrahydrodeoxycorticosterone (THDOC), etc.). In one study, severe sedation that lasted for approximately 2 hours was induced in a female subject after having re- ceived 400 mg of micronized progesterone (Arafat et al., 1988). Blood pressure, pulse and respiration remained stable throughout this time. When the woman repeated the study with 100 mg of micronized progesterone, she experienced only some minor transient sleepiness (Arafat et al., 1988). It is known that blood concentrations of progesterone metabolites after progesterone administration show large inter and even intraindividual differenc- es. In this woman concentrations of GABA(A) receptor potentiating metabolites were probably far more higher than concentrations of metabolites that inhibit or decrease GABA(A) receptor function. Androgenic effects such as water retention, edema, icterus, nausea and headache, and virilisation may occur only after long-term treat- ment.

Antidote for progesterone intoxication

There is no antidote for progesterone intoxication. Although some metabolites of progesterone, for exam- ple allopregnanolone and pregnanolone, affect the GABA(A) receptor, the benzodiazepine antagonist flumazenil is not effective as it interacts with another site of the receptor complex than these metabolites (Brot et al., 1997).

Allopregnanolone

Different notations for indicating allopregnanolone (C21H34O2, molecular weight: 318.5) are being used: 3alpha-OH-5alpha-pregnan-20-one, 5alpha-pregnane-3alpha-ol-20-one, and 3alpha,5alpha-tetrahydroproges- terone (3alpha,5alpha-THP). Allopregnanolone is derived from progesterone, within the brain or peripherally. In both cases, progesterone is the direct precursor which is reduced in a two-step reaction. The first step is irreversible and is mediated by 5alpha-reductase that converts progesterone to 5alpha-hydroxyprogesterone.

There are two isoenzymes of 5α-reductase: 5α-R1 and 5α-R2. In humans 5α-R2 is found mostly in the prostate gland. 5α-R1 is abundant in the brain. 5α-R2 is also present in the brain but in much lesser amount (Celotti et al., 1997). 5α-DHP has some progestational potency. The second step is reversible and is mediated by 3alpha- hydroxysteroiddehydrogenase (HSD) (or –oxidoreductase: HSOR) that converts 5alpha-hydroxyprogesterone to allopregnanolone (Krieger and Scott, 1984; Karavolas and Hodges, 1990; Mellon and Griffin, 2002). This is a

24 Chapter 2 | Pharmacological characteristics of progesterone and allopregnanolone, steroid nomenclature, and development of neuroactive steroids to drugs

reversible step: by reoxidation by this same enzyme allopregnanolone can be reconverted to 5α-DHP. There exist four isoforms of the 3α-HSD (Matsuura et al., 1998). Although a lot of research has been done in animals with ad- ministration of allopregnanolone, human studies have been limited to assessment of plasma concentrations. It has not been administered to humans. Pharmacokinetic data is therefore limited. No information on allopregna- nolone is available about the extent of protein binding, distribution volume, half life, clearance, or bioavailability. Plasma concentrations of allopregnanolone have been reported (see further). To be able to have some kind of an idea of what dosage has to be given to gain an effect without serious adverse events, we summarised data from studies that had investigated the stereoisomer of allopregnanolone, pregnanolone, in humans. Pregnanolone had been been subject of research in the development of a new anaesthetic agent in humans, called eltanolone, and was later, after further development had been canceled due to adverse events caused by the solubilizing agent Intralipid, used in clinical scientific research.

Pharmacokinetic characteristics of (allo)pregnanolone

Protein binding of pregnanolone is 1.1 (0.7-1.7)% in women (Dale et al., 1999). Distribution volume of preg- nanolone ranges from 1 to 4 l/kg depending on compartment model used (total or central compartment) and time phase (steady state, elimination phase) (Carl et al., 1990; Gray et al., 1992; Carl et al., 1994; Parivar et al., 1996; Dale et al., 1999). Half-life ranges from 21 to 42 min (distribution phase) (Carl et al., 1990; Gray et al., 1992; Carl et al., 1994; Parivar et al., 1996; Dale et al., 1999; Sundstrom et al., 1999). Clearance was about 25 ml min−1 kg−1 (Carl et al., 1990; Gray et al., 1992; Carl et al., 1994; Parivar et al., 1996; Dale et al., 1999) and 60 ml min−1 kg−1 (Sundstrom et al., 1999). The clearly higher clearance rate in the latter study might be due to different methods of calculation. About metabolisation and excretion of allopregnanolone a little bit more is known. Al- lopregnanolone can (1) be metabolised back to 5alpha-dihydroprogesterone (5alpha-DHP) via oxidative reaction catalyzed by the 3alpha-hydroxysteroidoxidoreductase enzyme (3alpha-HSOR). Then, 5alpha-DHP is converted to 3beta-OH-5alpha-pregnane-20-one by 3beta-HSOR (Stromstedt et al., 1993). 3beta-OH-5alpha-pregnane-20- one is then further hydroxylated into more polar metabolites (Stromstedt et al., 1993). Circulating levels of the 3beta-OH-5alpha-pregnane-20-one seem to be approximately half the level of allopregnanolone in both the follicular and luteal phases (Havlikova et al., 2006). Allopregnanolone can also be metabolized by (2) 6alpha- hydroxylase enzyme in extrahepatic tissues to 3alpha,6alpha-dihydroxy-5alpha-pregnan-20-one which is subse- quently delivered to the liver and then excreted in the urine as a glucuronoside (Chantilis et al., 1996), or by (3) 20alpha-hydroxysteroid oxidoreductase to 5alpha-pregnan-3alpha,20alpha-diol (pregnanediol) and conjugated with glucuronic acid and subsequently excreted with the urine as pregnanediol glucuronide (sodium pregnan- ediol-20-glucuronide). In order to get an impression of the pregnanolone concentrations that arise after pregnanolone admin- istration, we collected data from the control subjects in the studies from Sundström who administered preg-

25 Chapter 2 | Pharmacological characteristics of progesterone and allopregnanolone, steroid nomenclature, and development of neuroactive steroids to drugs

nanolone to humans in clinical scientific research. Cumulative doses of pregnanolone of 0.18 mg/kg resulted in concentrations of 164-180 nmol/l (Sundstrom et al., 1998; Sundstrom and Backstrom, 1998a; 1999; Sundstrom et al., 1999). Because there was only limited data on humans on suitable doses for sedation, Sundström gave doses corresponding to only one fourth of the dose required for anaesthesia (Sundström - personal communications). The total dosage was distributed over three increasing dosages (0.03, 0.06 and 0.09 mg/kg) with intervals of 25 minutes. In this manner a dose-response curve could be established. No adverse events were reported. Administration of pregnanolone is not complicated by neuroendocrine disturbances, regardless of cycle phase (Sundstrom and Backstrom, 1999). Hereby, the allopregnanolone level that is reached just before parturition of

50 ng/ml (which is about 150 nmol/l) reported in pregnant women (Luisi et al., 2000) and of ≥30 ng/ml reported in third trimester of pregnancy (which is about ≥100 nmol/l) (Paul and Purdy, 1992) is not exceeded. The studies on which Sundstrom based the dosages of pregnanolone were as follows. Carl administered a bolus dosis of 0.5-0.6 mg/kg intravenous (i.v.) (Carl et al., 1990), and a bolus dosis of 0.6 mg/kg i.v. in 45 s (Carl et al., 1994). Gray adminstered subsequently 0.5 mg/kg i.v. in 30 s, 0.75 mg/kg i.v. in 45 s, and 1.0 mg/kg i.v. in 60 s (Gray et al., 1992). Parivar administered eltanolone as a bolus of 0.75 mg/kg over 20 s i.v. and as a constant rate infusion at 2 mg/kg/hr and 3.5 mg/kg/hr during two hours (Parivar et al., 1996). Dale administered eltanolone i.v. as a bolus of 0.75 mg/kg over about 20 s (Dale et al., 1999). Normal values (reference values) for allopregna- nolone are not available from clinical laboratories. Therefore we gathered data from plasma samples from other studies (table 1). From this table it can be concluded that in ovulatory menstrual cycles, allopreganolone concentration varies with the different phases and that normal plasma values of allopreganolone range from 0-1 nmol/l in the follicular phase to 1-4 nmol/l in the luteal phase. This is in accordance with data that became available later and which showed that circulating levels of allopregnanolone are two to four times higher in the luteal phase than in the follicular phase (Wang et al., 2001; Havlikova et al., 2006). Allopregnanolone levels do not change during the first two years of life but increase during pubertal development, suggesting that this steroid may be involved in the adaptive neuroendocrine mechanisms related to puberty (Fadalti et al., 1999). Allopregnanolone levels increase in parallel with progesterone during gestation and parturition (Luisi et al., 2000), and are detectable postpartum (Nappi et al., 2001). The highest concentration of allopreganolone is reached during pregnancy (Paul and Purdy, 1992; Bixo et al., 1997). In the third trimester of human pregnancy, plasma levels of alloprega- nolone and pregnanolone are approximately 100 nM (Paul and Purdy, 1992). In both cycling and pregnant rats, al- lopreganolone levels in the brain are higher than in plasma (Corpechot et al., 1993) and it can be suggested that the same is in humans. Plasma concentrations of allopregnanolone are correlated with those of progesterone (Schmidt et al., 1994). And the concentration of allopreganolone in brain correlates well with the concentration of progesterone in plasma (Paul and Purdy, 1992; Wang et al., 1996).

26 Chapter 2 | Pharmacological characteristics of progesterone and allopregnanolone, steroid nomenclature, and development of neuroactive steroids to drugs

Table 1. Allopregnanolone blood concentration in healthy subjects

27 Chapter 2 | Pharmacological characteristics of progesterone and allopregnanolone, steroid nomenclature, and development of neuroactive steroids to drugs

Summary of known and potential risks of (allo)pregnanolone

In the pregnanolone/eltanolone reports mentioned before, the main adverse event was the one reported in the study of Carl in 1990. This concerned an elevated body temperature for more than one hour (38.3°C after 11 hours) in one subject (Carl et al., 1990). The known pyrogenic effect of steroids was given as possible expana- tion. In the study of Gray, mean arterial pressure decreased with a maximum reduction of 17.8% at 4 minutes from the start of induction in the 0.75 mg/kg group. The investigators suggested this to be a study-design effect. Subjects were more relaxed as they became familiar with the study procedures. Respiratory depression (apnoe) occurred in one subject, after a dose of 1.0 mg/kg, and ventilation had to be assisted during 90 s. It was con- cluded that the degree of cardiovascular and respiratory depression observed was comparable to that seen with other i.v. induction agents (Gray et al., 1992). The second study on eltanolone done by Carl showed only minor cardiorespiratory changes (Carl et al., 1994). Parivar didn’t mention any side effect (Parivar et al., 1996). At a dose varying between 0.71 and 0.76 mg/kg, erythematous rash, nausea and vomiting with frequencies of 24, 67 and 19% respectively, were reported by Dale, while hemodynamics remained stable (Dale et al., 1999). Among other things, due to the cutaneous reactions, eltanolone was eventually withdrawn. Administration of allopregnanolone in mice in a dose of 20 mg/kg intraperitoneal produced decreased motor ac- tivity in ca. 40 % of animals (rotorod equilibrium procedure). Lower doses (5 and 10 mg/kg) had no such effects. (W. Lason and B. Budziszewska - personal communications). These are far higher dosages than we intended to apply.

Dosage and method of administration

Three intravenous consecutive doses with an dosage scheme of 0.015, 0.03 and 0.06 mg/kg allopregna- nolone with intervals of 25 minutes, each dose infused during 30 s, a total dose of allopregnanolone that was half of the dose that Sundstrom applied, and which latter one was already only a quarter of the needed dosage to induce anaesthesia with pregnanolone, was considered by us as a save dosing regime with no serious adverse events.

Antidote for allopregnanolone intoxication

There is no known antidote for allopregnanolone intoxication but considering the low-dose that we in- tended to administer and the short half-life of pregnanolone, no antidote would be needed, we assumed (see further the paragraph before about the antidote for progesterone).

28 Chapter 2 | Pharmacological characteristics of progesterone and allopregnanolone, steroid nomenclature, and development of neuroactive steroids to drugs

Nomenclature of (neuroactive) steroids

Steroids are numbered and rings are lettered as follows (one speaks of the steroid skeleton):

The hydrogen atom at C5 is termed alpha if it lies below the plane of the paper and beta if it lies above the plane of the paper. This is an aspect of stereochemistry. Stereochemistry concerns the study of the spatial ar- rangement of atoms within the molecule. This is important because a different spatial orientation of the atoms can make a huge difference between the properties of molecules that share basically exactly the same atoms (called ). Thus, the structure-activity relation of a molecule is dominated by its stereochemical structure. In formulae, bonds to atoms lying below the plane of the paper (alpha) are shown as broken lines and bonds to atoms lying above the plane of the paper (beta) are shown as solid lines:

a 5alpha-steroid a 5beta-steroid

The different configuration of the hydrogen atom at C5 in both molecules, has effect on the spatial arrangement of the total steroid skeleton:

a 5alpha-steroid a 5beta-steroid

29 Chapter 2 | Pharmacological characteristics of progesterone and allopregnanolone, steroid nomenclature, and development of neuroactive steroids to drugs

This is the case with allopregnanolone (3alpha-OH-5alpha-pregnan-20-one) and preganolone (3alpha-OH- 5beta-pregnan-20-one). They are each others cis- and transisomers determined by the alpha or beta position of the hydrogen atom bound to the fifth carbon (Morris et al., 1999). There is not so much difference between the two substances although allopregnanolone is more potent than pregnanolone. There is, however, another endogenous neuroactive steroid which has also almost exactly the same configuration as allopregnanolone. In this neuroactive steroid the hydroxyl group at C3 is not in the alpha but in the beta position; 3beta-OH-5alpha- pregnan-20-one. There is a huge difference between this molecule and allopregnanolone although they both share exactly the same atoms; allopregnanolone is one of the most potent positive allosteric modulators of the GABA(A) receptor while the 3beta isomer has been shown to be an allopregnanolone antagonist (Wang et al., 2000; Lundgren et al., 2003). The learning impairment induced in rats by allopregnanolone is reduced by this isomer (Turkmen et al., 2004). As we know that certain women experience negative symptoms premenstrually or during hormone replacement therapy and this is due to altered sensitivity to allopreganolone, the 3beta iso- mer of allopregnanolone might be a new drug against these symptoms.

Neuroactive steroids developed for pharmacological drugs

As early as 1941, Selye demonstrated the rapidly sedative effects of intraperitoneally injected progester- one and some of its metabolites in rats (Selye, 1941a; b). Even before that time it had been shown that the precur- sor of progesterone, cholesterol, could induce rapidly anesthesia in cats (Cashin and Moravek, 1927). It was not known how this could have such rapid actions. Steroid hormones were known to bind to nuclear receptors within the cell and in this way could influence the synthesis of receptors, peptides and so on. This was suggested to take place in a time course of days. Metabolites of steroids were known to exist but were regarded as inactive and waist from the parent hormone, at that time. But the metabolites gained more interest due to the findings of Selye. Other investigators subsequently tried to develop these substances to anaesthetics for human use. Because of poor solubility, bioavailability and side effects, however, only a few of these drugs have been ad- ministered to humans and were often withdrawn finally (Sear, 1998). Tens of years later the mechanism behind the rapid sedative effects of the steroid metabolites was discovered. They alter the excitability of neurons in mil- liseconds to seconds by binding to membrane-bound receptors (Majewska et al., 1986). Besides sedative effects also anticonvulsive properties of certain progesterone metabolites were observed, leading to the development of a synthetic which is now tested in clinical trials (Monaghan et al., 1997). The role of neurosteroids is not only subject of interest in the fields of anaesthesiology and neurology, but in gynecology and psychiatry as well. Through the years, researchers have tried to develop neuroactive steroids to drugs, encouraged by new findings continuously emerging like the reported neuroleptic like activity in rodents of progesterone and its reduced metabolite allopregnanolone (Rupprecht et al., 1999). Without suggesting to be complete, we list neuroactive steroids that have been or are being developed as new drugs:

30 Chapter 2 | Pharmacological characteristics of progesterone and allopregnanolone, steroid nomenclature, and development of neuroactive steroids to drugs

Althesin

The synthetic steroidal anaesthetic agent, alphaxalone (3alpha-hydroxy-5alpha-pregnane-11,20-dione, the active compound) is combined with alphadolone acetate (21-acetoxy-3alpha-hydroxy-5alpha-pregnane-11,20-di- one) in the anaesthetic agent Althesin® (Glaxo) (in veterinary science it is called Saffan®), which became avail- able in 1971. The only difference in the molecular structure between alphaxalone and alphadolone (alphadolone acetate after de-acetylation) is the hydroxyl group at the 21 position (Nadeson and Goodchild, 2001). Both in- gredients are 5alpha-reduced progesterone derivatives. The sedative and anaesthetic properties of the mixture were due to the alphaxalone content and all of the antinociceptive properties were due to the alphadolone content (Goodchild et al., 2001). Onset occurs in about 30 seconds, and it has a duration of action of 5-10 min- utes. It produces little respiratory effects, and nausea is unusual. Because of serious side effects, anaphylactoid reactions frequently occurred, which were related to its formulation with Cremophor EL, Althesin is now only being used in veterinary medicine (Cornet and Popescu, 1977; Sear, 1998).

Eltanolone

Eltanolone (pregnanolone) has been originally investigated as a potential anaestheticum (Carl et al., 1990; Carl et al., 1994). Eltanolone, though it seemed to be an attractive agent, was eventually withdrawn because of cutaneous reactions probably due to the Intralipid in which it was solubilised. Due to unfavorable properties, among which Pk/Pd studies revealed a slow onset/offset, eltanolone turned out to be an unsuitable agent for i.v. maintenance of anaesthesia (Sear, 1996; 1998).

Ganaxolone

This 3a-hydroxy-3b-methyl-5a-pregnan-20-one is the 3b-methylated, synthetic analog of allopregnanolone and possesses no detectable hormonal activity because it is not converted to the hormonally active 3-keto form (Reddy and Rogawski, 2000). It is very recently being used as an antiepileptic drug in humans (Nohria and Giller, 2007).

GR 2/146

GR 2/146 (5beta-pregnane-3alpha-ol, 11, 20 dione 3 phosphate disodium) has been withdrawn because of delayed onset anaesthesia in animals and man, and paraesthesia in arm and neck after i.v. dosing (Sear, 1998).

31 Chapter 2 | Pharmacological characteristics of progesterone and allopregnanolone, steroid nomenclature, and development of neuroactive steroids to drugs

Viadril Chapter 3 | Allopregnanolone and serotonin

This 21-hydroxy derivative of pregnanedione is also called and was made water soluble by esterfication at the C21 position as the sodium hemisuccinate but showed side-effects (P’An et al., 1955).

Minaxolone

Minaxolone citrate (2beta, 3alpha, 5alpha, 11alpha-2-ethoxy-3-hydroxy-11,N,N-dimethyl-amino-pregnane- 20-one) has been withdrawn because of slow onset of action and delayed recovery, with a high incidence of excitatory movements and hypertonus, and possible oncogenic effect in rats (Sear, 1998).

32 Chapter 3 | Allopregnanolone and serotonin

Chapter 3 | Allopregnanolone and serotonin

Serotonin (5-hydroxytryptamine, 5-HT) is a neurotransmitter used by many neurons throughout the brain (Brodie et al., 1955). Serotonin alters the function in different neurons by altering the rate of signaling and also modulates the release of other neurotransmitters. In serotonin biosynthesis the precursor to serotonin, tryp- tophan, an essential amino acid (Wurtman, 1983), is converted to 5-hydroxytryptophan (5HTP) by tryptophan hydroxylase, found in most tissues, including the brain. 5HTP is almost immediately decarboxylated to serotonin and the enzyme responsible for this conversion is aromatic L-amino acid decarboxylase. Serotonin can be stored in the cytoplasm, transported to vesicles, or degradated by monoamine oxidase. Serotonergic neurons are lo- calized in the raphe nuclei in mesencephalon and medulla oblongata. Physiological functions or symptoms in which serotonin is involved are numerous and include aggression, impulse control, anxiety, sexual behavior, pain, sleep and appetite (Spigset, 1997). Dysfunction of serotonergic transmission has been regarded as an impor- tant mechanism in several psychiatric disorders, particularly major depression and anxiety disorders (Meltzer, 1989; Charney et al., 1990). For over 30 years, the leading theory to explain the biological basis of depression has been the “monoamine hypothesis of depression”. This theory proposes that the biological basis of depres- sion is a deficiency in one or more of the three key neurotransmitter systems, which are thought to mediate the therapeutic actions of virtually every known antidepressant agent. The important neurotransmitters are norepinephrine, dopamine and serotonin. The effect of serotonin in the synapse is terminated partly by the re- uptake mechanism of a specific protein localized presynaptically, also referred to as the serotonin transporter. The development and introduction of pharmacological agents able to block this mechanism (selective serotonin re-uptake inhibitors, SSRIs), including fluoxetine, sertraline, paroxetine, fluvoxamine, and citalopram, represent an important advance in the pharmacotherapy of psychiatric disorders. SSRIs are not only used in depression (Meltzer, 1989), but also in a wide range of psychiatric disorders, e.g., panic disorder (Rosenbaum et al., 1995), obsessive compulsive disorder (Piccinelli et al., 1995), eating disorders (Vaswani et al., 2003) and PMDD (Steiner 1995). SSRIs inhibit the reuptake of serotonin into the presynaptic nerve terminal, thus increasing the serotonin concentration in the synaptic cleft, prolonging its activity at postsynaptic receptor sites. There are a number of different serotonin receptors and subtypes having different functions and different brain localizations (Barnes and Sharp, 1999). Most of the serotonin receptors are metabotropic (except the 5HT3 receptor which is an ligand-gated ion channel), affecting G-protein-stimulated adenylyl cyclase or phospholipase activity, depending on receptor subtype. Of these, the 5HT1A, 2A and 2C receptors have been shown to be involved in regulation of mood and anxiety (Palvimaki et al., 1996; Blier et al., 1997; Spigset, 1997; Toth, 2003; Van Oekelen et al., 2003). The 5HT1A receptor is present as a somatodendritic autoreceptor and as a postsynaptic receptor, depending on the brain area (Kia et al., 1996). After two to three weeks of treatment with SSRIs, initially activated presynaptic 5HT1A autoreceptors by the increased serotonin synapse content, are desensitized. Since the initial activation

35 Chapter 3 | Allopregnanolone and serotonin

during treatment with SSRIs of the somatodendritic autoreceptors results in decreased firing activity along the serotonergic axon, decreased sensitivity of these receptors will finally result in enhancement of serotonin neu- rotransmission (Chaput et al., 1986; Elena Castro et al., 2003; Hensler, 2003).

Neuroactive steroids and serotonin system

The serotonin system is involved in progesterone related negative mood symptoms but the mechanism behind their interaction is unknown. Hormonal fluctuations in women are often accompanied by depressive symptoms, e.g during the premenstrual phase or postpartum. General conclusions about the interaction be- tween ovarian steroids and serotonergic function are difficult to draw since a wide variety of results have been obtained from animal and human studies on this subject. Allopregnanolone has been shown to have antidepressive effects in the mouse forced swim test model of depression (Khisti and Chopde, 2000; Khisti et al., 2000). It is suggested that interaction between allopregna- nolone and the serotonin system is, at least partly, responsible for the antidepressive-like effects of allopregna- nolone. Although allopregnanolone is a weak allosterical antagonist of the 5HT3 receptor, it increases the firing rate of dorsal raphe nucleus serotonergic neurons in animals (Robichaud and Debonnel, 2004; 2006). Firing rate is also increased at the end of pregnancy when progesterone and allopregnanolone levels are very high (Klink et al., 2002). Data on human subjects regarding ovarian steroid interactions with the serotonergic system are scarce, often variable, and sometimes contradictory. Of interest are two studies in which the hormonal influence on the serotonergic system, assessed by measurement of platelet serotonin transporter density and 5-HT2A receptor binding by means of radioactive paroxetine and LSD, respectively, was investigated. Subjects consisted of post- menopausal women before and during sequential hormone replacement therapy and reproductive women dur- ing six different stages of the menstrual cycle. In the postmenopausal women no significant effect of hormone replacement therapy was found on serotonergic function (Wihlback et al., 2001), but in the study on reproductive women, however, it was clear that a significant variation in platelet serotonin transporter and 5-HT2A receptor binding was evident during the menstrual cycle. A decrease in Bmax (indicating maximum number of binding sites, e.g. number of receptors) was observed for both the serotonin transporter and the 5-HT2A receptor during the luteal phase, a period of time during which women with PMDD experience negative mood changes (Wihlback et al., 2004). It has also been shown on gene level that ovarian hormones influence serotonin receptors. 5HT1A, 2A and 2C receptor mRNA expression is modulated in a hormone and brain area specific manner, involving brain areas important for mood and memory functions (Birzniece et al., 2001; Birzniece et al., 2002). Ovarian hormones have also effect on the biosynthesis of serotonin as progesterone increases the turno- ver rate of serotonin (Ladisich, 1977). Other evidence for the interaction between neuroactive steroids and the

36 Chapter 3 | Allopregnanolone and serotonin

serotonergic system comes from the observation that reduced pregnanolone sensitivity in PMDD patients is normalized during successful treatment with the SSRI citalopram (Sundstrom and Backstrom, 1998). The serotonergic system is thus intimately linked to, and influenced by, neuroactive steroids. This could explain part of the specific vulnerability that women have for the development of mood and anxiety disorders and for the deterioration of mood so frequently seen during the luteal phase. Women are approximately two times as likely as men to report a lifetime history of major depression or anxiety disorder and these disorders occur especially during the reproductive years. The sex difference begins in the early adolescence and persists through the mid 50s (Kessler et al., 1993; Kessler et al., 1994; Wittchen et al., 1994). Periods of hormonal vari- ability, that is, menarche (Angold et al., 1999), premenstrual periods (Soares et al., 2001), postpartum (Chaudron et al., 2001) and perimenopause (Freeman et al., 2004) have been suggested to increase the vulnerability to mood disorders in certain women. Therefore, it seems likely that sex steroid hormones can provide one possible explanation for the differences in mood disorders observed between the genders. Bearing this in mind, and considering the considerable effects of the fluctuating endogenous concentrations of steroid hormones during the menstrual cycle on the peripheral serotonergic measurements it seems obvious that neuroendocrine factors probably contribute to the overall increased risk of developing mood disorders in women.

Serotonin and GABA system

There is a complex interaction between the serotonin and GABA systems. The GABAergic system is in- volved in major depression, with decreased activity of enzymes needed for GABA synthesis and GABA levels in the brain (Brambilla et al., 2003b). Especially occipital cortex is affected and after SSRI treatment there is an increased GABA level in this brain region (Sanacora et al., 1999; Sanacora et al., 2002; Bhagwagar et al., 2004). Thus, the GABA system is highly involved in pathologies like major depression. There is also a direct interaction between the GABA and the serotonin system in the hippocampus, were serotonin neurons often end at inhibitory GABAergic interneurons (Gulyas et al., 1999).

Together with chapter 1 in which the relation between steroids and the GABA system is discussed, it can be concluded that steroids, GABA and serotonin influence each other.

37 Chapter 3 | Allopregnanolone and serotonin

Chapter 4 | Neurosteroids in depression: a review

38 Chapter 4 | Neurosteroids in depression: a review

Van Broekhoven F, Verkes RJ. Psychopharmacology (2003) 165:97–110

Chapter 4 | Neurosteroidsepression: a review

Abstract

Rationale: A deregulation in concentrations of the neurosteroids (allo)pregnanolone and 3alpha,5alpha-tetrahy- drodeoxycorticosterone (3alpha,5alpha-TH DOC) has been found in depressed patients. These levels normalize following treatment with Selective Serotonin Reuptake Inhibitors (SSRIs). Furthermore, administration of the neurosteroid dehydroepiandrosterone (DHEA) to depressed patients is associated with an improvement in the symptoms of depression. Objective: The aim of the present review is to clarify the mechanisms whereby neurosteroids, particularly al- lopregnanolone and DHEA, are involved in depression and to discuss the effect of SSRIs on allopregnanolone concentration. Methods: Literature on preclinical and clinical research has been analyzed in relation to the pathophysiology of depression. Results: Decreased plasma and cerebrospinal fluid concentrations of allopregnanolone in depressed patients in- crease to normal levels following effective psychopharmacological treatment. This might either be a physiologi- cal aspect of improvement in the symptoms of depression or a pharmacologically induced alteration. Several findings support the hypothesis of an antidepressant effect of allopregnanolone. These include an antidepres- sant effect demonstrated in an animal model of depression and a suppressing effect on corticotropin-releasing hormone (CRH) and arginine-vasopressin (AVP) gene expression. SSRIs increase levels of allopregnanolone but this effect is not confined to this class of drugs alone. The beneficial effect of DHEA administration in depressed patients might result from its sigma 1 receptor-mediated enhancement of noradrenaline and serotonin neuro- transmission, antiglucocorticoid effects, and cognition enhancing effects. Conclusions: Indirect genomic (allopregnanolone) and non-genomic (allopregnanolone and DHEA) mechanisms are involved in the neurosteroidogenic pathophysiology of depression. Clinical studies in homogeneous groups of non-pharmacologically treated depressed patients are required to further elucidate this relationship.

41 Chapter 4 | Neurosteroids in depression: a review

Introduction

The clinical importance of neurosteroids was already recognized in the early nineteen-seventies (Holz- bauer, 1976), but new impetus in the study of the implications of these steroids in clinical psychiatry did not emerge until the last decade. Neurosteroids are defined as steroids, the accumulation of which in the central and peripheral nervous systems occurs, at least in part, independently of supply by the steroidogenic endocrine glands. These neurosteroids can be synthesized de novo in the nervous system from sterol precursors (Baulieu et al., 1999). They rapidly (milliseconds to minutes) influence the excitability of neurons by acting - onmem brane-bound ligand-gated ion channels (McEwen, 1991). One of the primary neurosteroid target receptors is the gamma-aminobutyric acid type A (GABA(A)) receptor complex (Robel and Baulieu, 1995b; Baulieu, 1998). In patients suffering from major depression, neurosteroid concentrations in plasma and cerebrospinal fluid (CSF) differ from control subjects and these concentrations appear to normalise after successful treatment with antidepressants (Romeo et al., 1998; Uzunova et al., 1998; Strohle et al., 1999; Strohle et al., 2000). Avail- able data indicates that neurosteroids are also involved in Premenstrual Dysphoric Disorder (PMDD) (Girdler et al., 2001), eating disorders (Monteleone et al., 2001), Attention Deficit and Hyperactivity Disorder (Strous et al., 2001), Post-Traumatic Stress Disorder (Spivak et al., 2000), generalized anxiety disorder (Jetty et al., 2001; Semeniuk et al., 2001), panic disorder (Strohle et al., 2001; Strohle et al., 2002), sensitivity to benzodiazepines (McAuley et al., 1995; Kroboth and McAuley, 1997), memory modulation (Robel and Baulieu, 1995a), (aggressive behaviour in) Alzheimer’s disease (Pickles, 1994; Legrain et al., 1995; Bernardi et al., 2000), aggression (- liers, 1988; Nicolas et al., 2001), and in the action of alcohol (Young et al., 1996; Morrow et al., 2001). In this review, we will first outline synthesis and effects of neurosteroids and secondly discuss in detail the literature on neurosteroids in depression.

Synthesis of neurosteroids

Several neurosteroids are known: 3alpha,5alpha-tetrahydrodeoxycorticosterone (3alpha,5alpha-TH DOC); the (PREG) and its sulphate ester pregnenolone-sulphate (PREGS), and 3alpha,5beta- tetrahydroprogesterone (3alpha,5beta-TH PROG or pregnanolone) and its epimer 3alpha,5alpha-tetrahydro- progesterone (3alpha,5alpha-TH PROG or allopregnanolone); and the androstanes dehydroepiandrosterone (DHEA), its sulphate ester DHEA-sulphate (DHEAS), and its metabolites and 3alpha-androstanediol (3alpha,5alpha-diol).

42 Chapter 4 | Neurosteroids in depression: a review

Enzymes

The concept of neurosteroids emerged from the following findings. Firstly, brain DHEA(S) (Corpechot et al., 1981) and PREG(S) (Corpechot et al., 1983) levels exceeded their respective concentrations in plasma. Secondly, brain DHEAS was not influenced by adrenal stimulation or inhibition with adrenocorticotropic hormone (ACTH) or respectively and increased two days after the stressful event of adrenalectomy and orchiec- tomy (Corpechot et al., 1981). Thirdly, substantial concentrations of DHEA(S) (Corpechot et al., 1981), PREG(S) (Corpechot et al., 1983) and allopregnanolone (Purdy et al., 1991; Corpechot et al., 1993; Cheney et al., 1995) were detected in the brain of rats after the peripheral production of these steroids was abolished by adrenalect- omization and gonadectomization. Direct evidence for the possible production in the central nervous system of neurosteroids has since been delivered by the demonstration in the brain of most of the enzymes required for steroidogenesis (Stoffel-Wagner, 2001).

Site of synthesis

Steroidogenesis takes place in the mitochondria of glia cells (oligodendrocytes and astrocytes) and in neurons of the central nervous system and peripheral glia (Schumacher et al., 2000). Cholesterol is transported into the mitochondrion by the peripheral type benzodiazepine receptor (PBR) located on the outer mitochon- drial membrane; this receptor is stimulated by the diazepam binding inhibitor, a putative endogenous ligand of the PBR (Papadopoulos and Brown, 1995). It is suggested that the cytoplasmic steroidogenic acute regulatory protein (StAR) targets the cholesterol to the PBR (West et al., 2001). Cholesterol is then converted to PREG, a conversion which constitutes the rate limiting step in the steroidogenesis and which is catalyzed by the P450 side-chain cleavage (P450scc) enzyme located on the inner mitochondrial membrane. PREG is then further metabolized to pregnanes and androstanes (Figure 1).

Neurosteroidogenesis control

Although most of the biochemical pathways involved in the synthesis of neurosteroids have been eluci- dated, the mechanisms that control the activity of the neurosteroid-producing cells are still largely unknown. It has been shown that gamma-aminobutyric acid (GABA), by activating the GABA(A)-receptor complex, inhibits the activity of neurosteroidogenic enzymes (Do-Rego et al., 2000). The fact that neurosteroids can enhance GABA(A) receptor activity (see below) has led to the suggestion of an ultra-short loop in which neurosteroids regulate their own biosynthesis (Do-Rego et al., 2000).

43 Chapter 4 | Neurosteroids in depression: a review

-

-

-dihydrodeoxycorti pregnenolone-sul α

5 PREGS -DH DOC α 5

pregnenolone, PREG

chain cleavage, side

dehydroepiandrosterone-sulphate,

P450 DHEAS P450scc -TH PROG -HH PROG α α ,5 α ,20 dehydroepiandrosterone, 3 α ,5 α DHEA 3

-hydroxysteroid oxidoreductase, α

3

androstanediol, -HSOR α α 3

,5

α ,

3 DHT

-tetrahydrodeoxycorticosterone, -diol α α

,5 ,5 α α

3 3 -TH DOC α ,5 α 3 -dihydroprogesterone, α

5 -DH PROG

α progesterone, -hexahydroprogesterone or , 5 α PROG ,20 -tetrahydroprogesterone or allopregnanolone α α

,5 ,5 α α

Fig. 1 Pathway of (neuro)steroidogenesis.costerone, 3 phate, 3

44 Chapter 4 | Neurosteroids in depression: a review

Binding site

Although neurosteroids may bind to many different kinds of receptors, this review focusses on the interac- tion with the GABA(A) receptor because this interaction appears to be the most relevant one. The binding site of neurosteroids on the GABA(A) receptor complex is not known (Gee et al., 1988; Lambert et al., 1995). Although some neurosteroids positively modulate GABA(A) receptors in a manner similar to benzodiazepines and bar- biturates (Majewska et al., 1986), they seem to act at a site distinct from that of benzodiazepines (Morrow et al., 1990) and barbiturates (Gee et al., 1988), suggesting a specific binding site at the GABA(A) receptor (Rupprecht and Holsboer, 1999). This is supported by the fact that coadministration of the benzodiazepine antagonist fluma- zenil does not diminish the anxiolytic effect of allopregnanolone (Brot et al., 1997). Besides positive modulating neurosteroids there are also negative modulating neurosteroids which act at separate sites and/or have differ- ent mechanisms of action on the GABA(A)-receptor complex (Gee et al., 1989; Zaman et al., 1992; Park-Chung et al., 1999). Other important neurosteroid target receptors are the N-methyl-D-aspartate (NMDA) receptor and sigma receptors. Positive and negative modulation of NMDA receptors by sulphated neurosteroids is mediated by distinct modulatory sites on the NMDA receptor (Park-Chung et al., 1996; Park-Chung et al., 1997). The involve- ment of neurosteroids in the interaction with the sigma receptor is mainly confined to DHEAS and will be further discussed in paragraph 5.3.1. For more general information concerning sigma receptors we refer the reader to the review by Walker (Walker et al., 1990).

Effects of neurosteroids

Neurosteroids bind to membrane-bound receptors and thus exert rapid (non-genomic) effects. In addition, the neurosteroids allopregnanolone and 3alpha,5alpha-TH DOC have the ability to exert an indirect genomic ef- fect.

Non-genomic effects

Allopregnanolone, pregnanolone and 3alpha,5alpha-TH DOC

At nanomolar levels allopregnanolone, pregnanolone and 3alpha,5alpha-TH DOC potentiate GABA-induced chloride influx through the ion-channel by prolonging the duration of opening and increasing opening frequen- cy. This leads to an increased duration of the resulting inhibitory postsynaptic potential (Harrison et al., 1987; Twyman and Macdonald, 1992; Zhu and Vicini, 1997). At micromolar levels, these neurosteroids induce inward

45 Chapter 4 | Neurosteroids in depression: a review

currents even in the absence of GABA (Harrison et al., 1987; Paul and Purdy, 1992; Lambert et al., 1995). These neurosteroids are generally referred to as positive allosteric modulators of the GABA(A) receptor complex. Al- lopregnanolone and 3alpha,5alpha-TH DOC are the most potent and efficacious of known endogenous positive allosteric modulators of GABA(A) receptors (Puia et al., 1990; Paul and Purdy, 1992). Some authors have found pregnanolone and allopregnanolone to have similar potency and efficacy (Gee et al., 1995; Lambert et al., 1995) whereas others have shown that allopregnanolone is more potent than pregnanolone (Norberg et al., 1987; Prince and Simmonds, 1993; Zhu et al., 2001). The non-genomic effects of (allo)pregnanolone and 3alpha,5alpha-TH DOC are benzodiazepine-like. They have sedative/hypnotic/anaesthetic (Norberg et al., 1987), anxiolytic (Crawley et al., 1986; Bitran et al., 1991; Wieland et al., 1991; Bitran et al., 1993; Picazo and Fernandez-Guasti, 1995; Bitran et al., 1999) and anticonvulsive (Belelli et al., 1989; Belelli et al., 1990; Kokate et al., 1994) effects. (Allo)pregnanolone and 3alpha,5alpha-TH DOC produce a benzodiazepine-like EEG profile (Friess et al., 1997; Lancel et al., 1997b; Muller-Preuss et al., 2002). The anxiolytic effect of allopregnanolone is mediated by the amygdala (Akwa et al., 1999), possibly due to an enhanced gene-expression of neuropeptide Y in this brain structure via positive modulation of GABA(A) recep- tors (Ferrara et al., 2001). No data are available on whether (allo)pregnanolone and 3alpha,5alpha-TH DOC cause muscle relaxation and anterograde amnesia, other established effects of benzodiazepines.

DHEA(S)

Dehydroepiandrosterone (DHEA) (Majewska, 1992; Imamura and Prasad, 1998) and its sulphated conju- gated metabolite, DHEAS (Majewska et al., 1990; Demirgoren et al., 1991; Majewska, 1992; Imamura and Prasad, 1998; Park-Chung et al., 1999; Shen et al., 1999) are negative allosteric modulators of the GABA(A) receptor, DHEAS having greater potency than DHEA (Demirgoren et al., 1991; Imamura and Prasad, 1998). Paradoxically, in spite of being negative GABA(A) receptor modulators, DHEAS (when administered in lower doses) and DHEA have anxiolytic properties (Melchior and Ritzmann, 1994). These properties are considered to be the result of a reduction in brain concentrations of PREGS, a negative GABA(A) receptor modulator, as administration of DHEA reduces PREGS concentrations in the brain (Young et al., 1991). The anxiolytic properties may also result from the metabolisation of DHEA to androsterone and 3alpha,5alpha-diol which are positive GABA(A) receptor modulators (Bitran et al., 1996; Frye et al., 1996a; Frye et al., 1996b; Wilson and Biscardi, 1997).

Indirect genomic effects

By entering the cell, allopregnanolone and 3alpha,5alpha-TH DOC can be oxidised to 5alpha-DH PROG and 5alpha-DH DOC, respectively (Figure 1). These substances bind to the cytosolic progesterone (PROG) receptor

46 Chapter 4 | Neurosteroids in depression: a review

and subsequently stimulate protein synthesis by influenceing gene expression (Rupprecht et al., 1993). In this way, allopregnanolone can alter GABA(A) receptor complex subunit composition (Biggio et al., 2000) which de- termines sensitivity to neurosteroids and benzodiazepines (Puia et al., 1990; Puia et al., 1992; Puia et al., 1993). A shortage of allopregnanolone also appears to influence subunit composition as well. Withdrawal from PROG in pseudopregnancy models (i.e., animals with artificially induced high plasma PROG levels), which simulates the rapid decrease in PROG seen in PMDD and postpartum psychosis, results in a decrease in brain content of allopregnanolone and a subsequent increase in the expression of the GABA(A)-receptor alpha-4 subunit. The presence of the alpha-4 subunit appears to be essential for the occurrence of PROG withdrawal effects (e.g., in- crease of anxiety and epileptic insults, tolerance and decreased sensitivity to positive allosteric GABA(A)-recep- tor modulators (e.g., some neurosteroids, benzodiazepines, and ethanol), and increased sensitivity to negative allosteric GABA(A)-receptor modulators) (Costa et al., 1995; Moran et al., 1998; Smith et al., 1998a; Smith et al., 1998b; Follesa et al., 2000; Biggio et al., 2001; Follesa et al., 2001; Gulinello et al., 2001). Other indirect genomic effects of allopregnanolone and 3alpha,5alpha-TH DOC are a decreased gene ex- pression of corticotropin-releasing hormone (CRH) and an attenuation of CRH-induced anxiety, although this latter effect is only partially genomically mediated (Patchev et al., 1994). Another indirect genomic effect of allopregnanolone is enhanced myelin repair in the central and peripheral nervous system by enhancing expres- sion of the genes that code for proteins involved in myelin repair (Baulieu and Schumacher, 2000; Magnaghi et al., 2001; Schumacher et al., 2001). Proposed mechanisms underlying these effects are an increased GABAergic tone, and conversion to 5alpha-DH PROG which subsequently activates the PROG receptor.

Neurosteroids in depression

Allopregnanolone and DHEA(S) in particular are of significance in depression. Both the indirect genomic and non-genomic mediated effects of allopregnanolone are involved in depression while DHEA(S) only exerts non-genomic effects.

Allopregnanolone

Administration of allopregnanolone produces antidepressive effects in the mouse forced swim test model of depression (Khisti and Chopde, 2000; Khisti et al., 2000). These antidepressive effects may be the result of the influence of allopregnanolone on hypothalamic-pituitary-adrenal (HPA) axis activity, and on noradrenaline and serotonin neurotransmission. Increased CRH and arginine-vasopressin (AVP) release is suggested to have a causal relationship with de- pression (Holsboer, 2000; 2001). Allopregnanolone decreases noradrenergic-stimulated CRH release (Patchev

47 Chapter 4 | Neurosteroids in depression: a review

et al., 1994) and decreases adrenalectomy-induced up-regulation of hypothalamic CRH (Patchev et al., 1994) and arginine-vasopressin (Patchev et al., 1996) gene-expression. Conversely, concentration of allopregnanolone increases following administration of CRH (Genazzani et al., 1998). The decremental effect of allopregnanolone on CRH and AVP may result from allopregnanolone-mediated increased GABAergic tone. This suggestion is in line with the effects of benzodiazepines on CRH and AVP. Like allopregnanolone, benzodiazepines are positive allosteric modulators of the GABA(A) receptor. Stimulation of intrahypothalamic GABA(A) receptors by an en- dogenous benzodiazepine ligand (Givalois et al., 1998) or benzodiazepines (Calogero et al., 1988; Kalogeras et al., 1990), decreases CRH mRNA-expression and CRH release, respectively. The benzodiazepine chlordiazepox- ide decreases basal plasma AVP levels (Wible et al., 1985). In depressed patients, CSF GABA levels are reduced (Gold et al., 1980; Gerner and Hare, 1981). In addition to its reducing effect on hypothalamic CRH and AVP levels (Patchev et al., 1994; Patchev et al., 1996), allopregnanolone also reduces plasma ACTH levels (Patchev et al., 1996). This latter effect may be achieved by a direct influence of allopregnanolone on corticotrophs. Thus, allopregnanolone simulates the negative feedback of . In depression, cortisol hypersecretion down-regulates hippocampal receptor (MR) and receptor (GR) (Holsboer, 1995). Allopregnanolone does not, however, down-regulate adrenalectomy-induced up-regulation of hippocampal MR and GR gene-expression (Patchev et al., 1996). It appears that allopregnanolone exerts gene-specificity accord- ing to the substrate (peptides vs. receptors) (Patchev et al., 1996). In addition to these effects on the HPA system, the antidepressive effects of allopregnanolone may also result from the influence of allopregnanolone on neurotransmission. Enhancement of noradrenaline (Delgado and Moreno, 2000) and serotonin (Delgado, 2000) neurotransmission is considered to have an antidepressive effect. Noradrenaline release is stimulated by the activation of GABA(A) receptors located on noradrenergic nerve terminals (Bonanno and Raiteri, 1987a; b; Peoples et al., 1991). The importance of the influence of allo- pregnanolone on serotonin receptors seems to be limited, as allopregnanolone acts only as a relatively weak functional antagonist at the ligand-gated 5-HT3 receptor via an allosteric mechanism and not via the serotonin binding site (Wetzel et al., 1998)

48 Chapter 4 | Neurosteroids in depression: a review

-TH DOC levels in patients with major depression before and following psychopharmacological treatment α

,5 α

: allopregnanolone combined: not assessed with pregnanolone: not significantly in the statistics from controls

Table 1. Allopregnanolone and 3 a b c

49 Chapter 4 | Neurosteroids in depression: a review

In accordance with this suggested allopregnanolone-mediated antidepressive mechanism are the results of the few clinical studies that have reported allopregnanolone levels in depressed patients (Table 1). These patients show significantly lower serum and CSF levels of allopregnanolone and pregnanolone, levels which normalize after clinically successful treatment with fluoxetine or fluvoxamine (Romeo et al., 1998; Uzunova et al., 1998; Strohle et al., 2000) or with non-SSRI antidepressants (Romeo et al., 1998; Strohle et al., 1999). Neither the concentrations of PROG nor those of 3alpha,5alpha,20alpha-hexahydroprogesterone (3alpha,5alpha,20alpha-HH PROG or allopregnanediol), a metabolite of allopregnanolone (Figure 1), were al- tered before or after treatment with antidepressants (Romeo et al., 1998; Uzunova et al., 1998; Strohle et al., 1999). This would suggest that the decreased allopreganolone concentrations in depressed patients are due to changes in the activities of the enzymes involved in the conversion of PROG via 5alpha-DH PROG to allopreg- nanolone (i.e., a decreased 5alpha-reductase activity or a decreased reductase respectively increased oxidative 3alpha-hydroxysteroid oxidoreductase (3alpha-HSOR) activity). Contradictory data have appeared on this sub- ject. It has been suggested on the basis of a mouse model of protracted social isolation, simulating the chronic stress associated with depression, that the reduction in brain allopregnanolone level is due to a decreased brain 5alpha-DH PROG content. This in turn is due to a down-regulation of 5alpha-reductase expression while 3alpha- HSOR activity is not altered (Dong et al., 2001). However, plasma concentrations of 5alpha-DH PROG did not dif- fer significantly between controls and depressed patients either before (Strohle et al., 1999) or after treatment with antidepressants (Romeo et al., 1998; Strohle et al., 1999). These clinical data suggest that the 3alpha-HSOR activity rather than the 5alpha-DH PROG activity is altered in depressive patients. In conclusion, normalisation of the decreased allopregnanolone concentrations following clinically effective treatment with antidepressants in the depressed patients together with the antidepressive effect of allopreg- nanolone in the mouse forced swim test and the allopregnanolone-induced decrease in CRH release suggest a major role for allopregnanolone in the pathophysiology of depression.

3alpha,5alpha-TH DOC

Although 3alpha,5alpha-TH DOC has not been tested in animal models of depression this neurosteroid has also been shown to decrease CRH mRNA levels in the hypothalamus and to counteract the decrease in GR mRNA levels in the hippocampus in a model of stress during early development (Patchev et al., 1997). 3alpha,5alpha-TH DOC is also a positive allosteric modulator of the GABA(A) receptor. One would, therefore expect plasma levels of 3alpha,5alpha-TH DOC to follow the same course as the plasma levels of allopregnanolone during depression. In depressive patients, however, Ströhle et al. found significantly higher plasma 3alpha,5alpha-TH DOC concen- trations compared to control subjects (Strohle et al., 1999; Strohle et al., 2000) (Table 1) and significantly higher plasma levels of its precursor 5alpha-DH DOC (Strohle et al., 1999). This would suggest that both the activities of the 5alpha-reductase and of the 3alpha-HSOR enzymes are elevated in depression. Plasma 3alpha,5alpha-

50 Chapter 4 | Neurosteroids in depression: a review

TH DOC (Strohle et al., 1999) (Table 1) as well as 5alpha-DH DOC (Strohle et al., 1999) concentrations were not significantly influenced by treatment with antidepressants other than SSRI’s and remained higher in remitted patients than in control subjects. 3alpha,5alpha-TH DOC level, however, decreased after 50 days of fluoxetine treatment and was then no different from those of controls (Strohle et al., 2000) (Table 1).

DHEA(S)

Open-label (Wolkowitz et al., 1997) as well as double-blind, randomized, placebo-controlled (Wolkowitz and Reus, 1999) oral administration of DHEA decreases depressive symptoms in patients with major depression, reflected by a significant decrease in Hamilton Depression Rating Scale (HDRS) score. The same effect after oral administration of DHEA is achieved in patients with dysthymia (Bloch et al., 1999). In addition, administration of DHEA has antidepressive effects in the mouse forced swim test model of depression (Prasad et al., 1997; Reddy et al., 1998; Urani et al., 2001). We will discuss below the suggested mechanisms underlying the antidepressive effect of DHEA below.

Interaction of DHEA with sigma receptors

Given that DHEA can be converted to DHEAS in the brain and liver in a reaction that is catalyzed by sul- photransferase (Rajkowski et al., 1997; Aldred and Waring, 1999; Shimada et al., 2001) the antidepressive effect of DHEA can be explained by the interaction between the sigma 1 receptor agonist DHEAS (Maurice et al., 1996) and noradrenaline and serotonin neurotransmission. As has already been mentioned in section 5.1, enhance- ment of noradrenaline (Delgado and Moreno, 2000) and of serotonin (Delgado, 2000) neurotransmission is considered to have an antidepressant effect. In vitro experiments suggest that sigma ligands inhibit noradrena- line presynaptic re-uptake in brain synaptosomes (Kinouchi et al., 1989; Rogers and Lemaire, 1991) and increase noradrenaline release (Kinouchi et al., 1989). Furthermore, treatment with DHEAS increases the number of NMDA receptors (Wen et al., 2001) and potentiates NMDA-evoked noradrenaline release via sigma 1 receptors (Monnet et al., 1995). This is in line with the finding that the SSRI sertraline, which has a high affinity for sigma receptors, enhances NMDA neuronal response in the rat dorsal hippocampus (Bergeron et al., 1993). This effect is mediated by sigma receptors as the sigma antagonist haloperidol blocks this potentiation (Bergeron et al., 1993). In contrast, the SSRI paroxetine, which is devoid of affinity for sigma receptors, has no effects on the NMDA response (Bergeron et al., 1993). With respect to serotonin it has been shown that some sigma 1 receptor ligands increase firing activity of serotonin neurons of the dorsal raphe nucleus by acting on sigma 1 receptors (Bermack and Debonnel, 2001). In support of a positive relationship between DHEA and brain serotonin is the fact that administration of DHEA increases hypothalamic serotonin levels (Abadie et al., 1993; Svec and Porter,

51 Chapter 4 | Neurosteroids in depression: a review

1997). For further information about the relationship between neurosteroids and sigma receptors, we refer to the reviews by Maurice (Maurice et al., 1999; Maurice et al., 2001). In addition to DHEAS, other agonistic ligands of the sigma 1 receptor also have antidepressive effects (Matsuno et al., 1996; Kinsora et al., 1998; Ukai et al., 1998; Akunne et al., 2001; Tottori et al., 2001).

Antiglucocorticoid effects of DHEA

High levels of cortisol impair cognitive function and are associated with depressed mood (Wolkowitz, 1994). DHEA(S) does have antiglucocorticoid effects as it inhibits glucocorticoid induced enzyme activity (Browne et al., 1992) and antagonizes dexamethasone-induced suppression of lymphocytes and thymic involution (Kalimi et al., 1994). In addition, administration of DHEA has been shown to decrease plasma cortisol levels (Wolkowitz et al., 1992; Wolf et al., 1997). It is not known precisely how DHEA(S) antagonizes the effects of cortisol. No nuclear binding of DHEA(S) has been shown in liver cells (Mohan and Cleary, 1992). DHEA(S) does not interact with the cortisol receptor, and the binding to peripheral DHEA(S) receptors or binding sites is, except for a modulatory ef- fect on interleukin-2 production (Meikle et al., 1992), without biological action. Yet a DHEA(S) receptor or binding site in the central nervous system has not been demonstrated (Wolf and Kirschbaum, 1999). The antidepressive effects of other forms of antiglucocorticoid therapy in Cushing’s syndrome and depression have been reviewed elsewhere (Wolkowitz and Reus, 1999).

Cognition enhancing effects

DHEA(S) has memory-enhancing and anti-amnestic properties (Baulieu and Robel, 1996). DHEAS amel- iorates memory impairment induced by dizocilpine, a non-competitive NMDA receptor antagonist, or by sco- polamine, a muscarinic acetylcholine receptor antagonist (Urani et al., 1998; Zou et al., 2000). This can be explained by the antiglucocorticoid action of DHEA(S), and by the interaction of DHEA(S) with the sigma 1 recep- tor, as the amelioration of memory impairment is antagonized by NE-100, a sigma 1 receptor antagonist (Urani et al., 1998; Zou et al., 2000). Furthermore, DHEA(S) antagonizes the memory deteriorating neurotoxic effect of high cortisol levels on the hippocampus (Herbert, 1998). Apart from the interaction with the sigma 1 recep- tor and its antiglucocorticoid effects, the interaction of DHEA with the GABA(A) and NMDA receptors has also been hypothesized as an underlying mechanism for the cognitive-enhancing effects of DHEA. Thus, the cogni- tive-enhancing properties of DHEA(S) may, at least in part, have contributed to the decrease in HDRS scores in depressive patients following DHEA treatment (Wolkowitz et al., 1997; Bloch et al., 1999; Wolkowitz et al., 1999).

52 Chapter 4 | Neurosteroids in depression: a review

DHEA(S) levels and cortisol/DHEA(S) ratios

In the presence of normal cortisol levels, a low DHEA level could result in functional hypercortisolaemia (Goodyer et al., 1996). As there is a wide interindividual variability in the plasma DHEAS levels (Thomas et al., 1994), cortisol/DHEA(S) ratios are more informative than DHEA(S) values alone (Hechter et al., 1997). Indeed, some studies in depressed patients found significantly higher cortisol/DHEA(S) ratios compared to controls (Os- ran et al., 1993; Scott et al., 1999), while other studies failed to find abnormal ratios (Reus et al., 1993; Fabian et al., 2001). Reported values of plasma DHEA(S) levels differed between studies ranging from low (Ferguson et al., 1964; Legrain et al., 1995; Wolkowitz et al., 1995; Goodyer et al., 1996; Wolkowitz et al., 1997; Barrett-Connor et al., 1999; Scott et al., 1999; Morrison et al., 2001) to normal (Osran et al., 1993; Reus et al., 1993; Goodyer et al., 1996; Scott et al., 1999; Fabian et al., 2001) and high levels (Hansen et al., 1982; Tollefson et al., 1990; Heuser et al., 1998; Takebayashi et al., 1998; Maayan et al., 2000; Weber et al., 2000; Morrison et al., 2001). These conflicting data on DHEA(S) levels in depressed patients could be explained by differences in age (as DHEA(S) concentra- tion declines with increasing age (Orentreich et al., 1992)), gender (Orentreich et al., 1984), drug therapy, and concurrent somatic or psychiatric diseases (Berr et al., 1996; Kroboth et al., 1999; Wolf and Kirschbaum, 1999; Salek et al., 2002). Another possible explanation is the time of day samples were obtained (DHEA does but DH- EAS does not follow a (slight) diurnal variation (Rosenfeld et al., 1975) similar to cortisol; in some (Osran et al., 1993; Goodyer et al., 1996) but not all (Heuser et al., 1998) studies a loss of diurnal variation has been found in depressed patients). However, despite all these inconsistencies measured in depression, a 24-hour blood sam- pling study that applied 30-minute sampling intervals in 11 female and 15 male hypercortisolaemic patients with major depression conclusively found significantly higher DHEA levels without loss of diurnal variation (Heuser et al., 1998). Unfortunately, no cortisol to DHEA(S) ratios were reported. One may hypothesize that increased DHEA levels may be part of a, albeit insufficient, compensatory mechanism to ameliorate the depressive state. It should be noted that the depressed patients who benefit from DHEA had low baseline DHEA(S) levels (Wolkow- itz et al., 1995; Wolkowitz et al., 1997). There has been no such 24-hour blood sampling study performed in patients who were in remission after successful antidepressant treatment. This is regrettable because only inconsistent data on DHEA(S) levels and cortisol/DHEAS ratios have been reported following remission from depression achieved by either psychop- harmacological or non-psychopharmacological treatment. Decreased (Wolkowitz et al., 1997) and unchanged (Takebayashi et al., 1998; Maayan et al., 2000; Fabian et al., 2001; Goodyer et al., 2001) ratios compared to baseline have been found following treatment. Increased ratios were found in patients in remission compared to depressed patients (Michael et al., 2000). Specifying the ratios to the respective hormones from which the ra- tios were computed, one can suggest that it is more likely to be the differences in DHEA(S) than in cortisol levels which influence the differences found in cortisol/DHEA(S) ratios in these studies. There is one exception to this: a study in which statistically significantly lower cortisol levels might have influenced the ratios (Michael et al., 2000). To these inconsistent DHEA(S) levels following antidepressant pharmacotherapy, one can add the results

53 Chapter 4 | Neurosteroids in depression: a review

from studies measuring only single DHEA(S) levels, but not computing the ratios of these to cortisol. Non-psy- chopharmacological therapy (i.e. ECT (Michael et al., 2000) or CBT (Tollefson et al., 1990)) is associated with increases in DHEAS. It is remarkable that the data on cortisol in the above-mentioned studies do not support the established cortisol lowering effect of antidepressant pharmacotherapy (Holsboer, 2000). In conclusion, similar to the inconsistent data on DHEA(S) levels in depressed patients, the available studies on DHEA(S) in remitted patients do not conclusively elicit the effect of antidepressant pharmacotherapy on DHEA(S) nor its association with remission of depression.

SSRIs

Although the antidepressive effect of SSRI’s is explained by their ability to increase synaptic serotonin concentration (Maes and Meltzer, 1995), recent findings indicate that another mechanism, independent of the serotonergic effect, may also account for their antidepressant effect. These recent findings show that 1) direct administration of allopregnanolone exerts an antidepressive effect in the mouse forced swim test model of de- pression (Khisti and Chopde, 2000; Khisti et al., 2000); 2) administration of the SSRI’s fluoxetine and paroxetine increases allopregnanolone content in rat brain (Uzunov et al., 1996) and increases the affinity of 3alpha-HSOR for 5alpha-DH PROG (Griffin and Mellon, 1999); and 3) fluoxetine potentiates the antidepressive effect of allo- pregnanolone by enhancing GABAergic tone but, remarkably, not by enhancing serotonergic neurotransmission (Khisti and Chopde, 2000). Is the increase in allopregnanolone levels after administration of SSRI’s due to a specific SSRI effect or to a generally enhanced central serotonergic activity? Factors supporting the latter explanation are the in- creased plasma allopregnanolone concentrations after infusion of the serotonin precursor l-tryptophan (Rasgon et al., 2000) and the findings of increased plasma/brain allopreganolone levels after antidepressant treatment with antidepressants that non-selectively inhibit serotonin re-uptake: amitriptyline, desipramine, clomipramine, nortriptyline, or viloxazine (Romeo et al., 1998; Strohle et al., 1999; Jaworska-Feil et al., 2000). Even drugs with- out serotonin reuptake inhibitory properties like carbamazepine (Biggio et al., 1998; Serra et al., 2000), lithium (Romeo et al., 1998; Strohle et al., 1999) and the antipsychotics olanzapine (Marx et al., 2000) and clozapine (Barbaccia et al., 2001) increase brain allopregnanolone content suggesting that neither a specific SSRI effect nor a central serotonergic effect is responsible for the increased brain content of allopregnanolone. However, a specific SSRI effect that induces increases in allopregnanolone levels has been demonstrated. This specific SSRI effect constitutes a directly altering effect on the activity of the 3alpha-HSOR enzyme (Figure 1). The SSRI’s fluoxetine, paroxetine, and sertraline enhance the affinity of 3alpha-HSOR for 5alpha-DH PROG (Griffin and Mellon, 1999), consequently shifting the activity of 3alpha-HSOR in the direction of reduction. This is not the case for the tricyclic antidepressant imipramine (Griffin and Mellon, 1999), which is equipotent to fluoxetine in inhibiting serotonin re-uptake but is less selective (Barker and Blakely, 1995). In addition, sertraline inhibits

54 Chapter 4 | Neurosteroids in depression: a review

the reduction of allopregnanolone to 5alpha-DH PROG by decreasing the affinity of 3alpha-HSOR for allopreg- nanolone (Griffin and Mellon, 1999). In support of this SSRI-induced and serotonergic-independent increase in allopregnanolone brain content is the normalization of the decreased content of allopregnanolone in the brain in socially isolated mice after the administration of fluoxetine, independent of its action on brain serotonin uptake (Matsumoto et al., 1999). In contrast, the much more potent serotonin re-uptake inhibitor citalopram did not show this effect, which supports the view that the increase in allopregnanolone brain content by fluoxetine is not a serotonergic mediated effect (Matsumoto et al., 1999). How the above-mentioned non-SSRI drugs enhance brain allopreganolone content is not known. With this alternative therapeutic mechanism of SSRI’s, i.e., increase in the affinity of 3alpha-HSOR for 5alpha-DH PROG with a consequent increase in brain allopregnanolone content, the anticonvulsant effect of fluoxetine (Favale et al., 1995; Gigli et al., 1996; Pasini et al., 1996) might be explained by an enhanced GABAergic tone. The involvement of the serotonergic system in PMDD is demonstrated by the fact that serotonin-releasing agents such as fenfluramine and mCPP reduce premenstrual complaints whereas depletion of tryptophan ag- gravates these symptoms (Eriksson, 1999). Furthermore, SSRI’s show a high efficacy in reducing premenstrual complaints in PMDD, often within the first menstrual cycle after the start of treatment and even when applied in an intermittent dosage scheme (Dimmock et al., 2000). The remarkable rapid onset of effect in PMDD (in depres- sion a beneficial effect of SSRI’s is reached only after 2-4 weeks) may reflect that, in PMDD, the mechanism of action of SSRIs is different from that in affective disorders (Wikander et al., 1998). The SSRI citalopram has yet another effect which cannot easily be explained by the increased synaptic serotonin concentration paradigm and which might also be extrapolated to other SSRI’s. This effect contains an enhancement of the decreased sensitivity to pregnanolone in PMDD women as measured by saccadic eye velocity (SEV) (Sundstrom and Back- strom, 1998). Comparable to the reduced SEV responsiveness to benzodiazepines (Sundstrom et al., 1997a; Sundstrom et al., 1997b), PMDD patients also show a reduction in sensitivity to the GABA(A) receptor agonist pregnanolone (Sundstrom et al., 1998). How the demonstrated enhancement of the decreased sensitivity to pregnanolone, and presumably to other neurosteroids as well, may contribute to these effects following treat- ment with SSRI’s, requires further investigation.

Summary, conclusions, and directions for future research

It has become increasingly clear that a deregulation of noradrenergic and serotonergic neurotransmitter systems and an impaired cortisol feedback constitute major features of the pathophysiology of depression. In addition, these features are influenced by effective pharmacological antidepressant treatment. The demonstra- tion that neurosteroids modulate noradrenergic and serotonergic neurotransmission and exert antiglucocorti- coid effects emphasizes the involvement of neuroendocrinological changes in the pathophysiology of depres-

55 Chapter 4 | Neurosteroids in depression: a review

sion. In contrast to the transcriptional regulation through activation of nuclear receptors by their parent steroid hormones (e.g., progesterone and deoxycorticosterone), neurosteroids exert their rapid effects by acting on membrane receptors. With regard to the involvement of neurosteroids in depression, most research has been focused on the PROG metabolite, allopregnanolone and the , DHEA(S). It has been demonstrated that both of these neurosteroids have antidepressant effects in animal models of depression. Plasma and CSF allopregnanolone levels are reduced and plasma 3alpha,5alpha-TH DOC levels are in- creased in depressed patients (Romeo et al., 1998; Uzunova et al., 1998; Strohle et al., 1999; Strohle et al., 2000). Plasma and CSF allopregnanolone levels normalize after improvement of depressive symptoms following treat- ment with SSRI’s (Romeo et al., 1998; Uzunova et al., 1998; Strohle et al., 2000) or other antidepressants (Romeo et al., 1998; Strohle et al., 1999) whereas plasma 3alpha,5alpha-TH DOC levels normalize after treatment with SSRI’s but remain high after treatment with other antidepressants (Strohle et al., 1999). As no studies have been reported on allopregnanolone and 3alpha,5alpha-TH DOC levels following non- pharmacological treatment, it is still unclear whether normalization of these levels is induced pharmacologically or represents a physiological characteristic of improved depression. Therefore, research in depressive patients treated with non-pharmacological therapy, for example with psychotherapy, electroconvulsive therapy or tran- scranial magnetic stimulation, is needed. Activities of the enzymes that are involved in the synthesis of allopreg- nanolone and 3alpha,5alpha-TH DOC, 5alpha-reductase and 3alpha-HSOR, should be concomitantly measured in order to reveal the pathophysiology of the altered levels of these neurosteroids in depression. Future studies should also consist of more homogeneous groups of depressed patients and controls than those used in previ- ous studies. Age and sex should be taken into account. Men, but not women, show an age-related decline in plasma allopregnanolone levels (Genazzani et al., 1998) and brain allopregnanolone levels also decline with age in male rats (Bernardi et al., 1998b). Although no differences exist in plasma allopregnanolone levels of men and women in the folliculair phase, women show higher levels in the luteal phase (Genazzani et al., 1998). The efficacy of SSRI’s in the treatment of depression, PMDD, and anxiety disorders is equal, and some- times superior to other antidepressants even if these also act on the serotonin system (Guidotti and Costa, 1998). This suggests that SSRI’s have an additional beneficial effect independent of augmented serotonergic neurotransmission. This additional effect may be related to the ability of some SSRIs to increase allopregna- nolone in the brain (Guidotti and Costa, 1998). To test this interesting hypothesis, the effect of allopregnanolone should be studied in patients suffering from such disorders. Such studies are, however, hampered by technical difficulties such as the short half-life of allopregnanolone and by the fact that no allopregnanolone prepara- tion is available for human use. However, the epimer of allopregnanolone, pregnanolone, has been successfully solubilized in an albumin solution and administered to women suffering from PMDD (Sundstrom et al., 1999). It appears that PMDD patients are less sensitive to pregnanolone as measured by SEV only in the luteal phase (Sundstrom et al., 1998; Sundstrom and Backstrom, 1998). Equivocal data have been reported on DHEA(S) levels in depressed patients. The hypercortisolaemia seen in these patients and the antiglucocorticoid action of DHEA requires that cortisol to DHEA(S) ratios are reported

56 Chapter 4 | Neurosteroids in depression: a review

rather than cortisol only. These ratios provide a more accurate assessment of functional hypercortisolaemia (Gallagher and Young, 2002). If it proved possible to replicate the successful results of oral DHEA in depressed patients (Wolkowitz et al., 1997; Bloch et al., 1999; Wolkowitz et al., 1999) by reporting the cortisol to DHEA(S) ratios in patients and adjusting for possible cognitive enhancement, DHEA could represent a novel antidepres- sant. The major part of neurosteroid research is performed in animals and the results of these studies are now gradually being translated to the field of clinical psychiatry. Some discrepancies do, however, exist between the results derived from animal studies and those obtained in humans. For example, after long-term treatment with fluoxetine rats show decreased plasma and brain levels of allopregnanolone (Serra et al., 2001), whereas de- pressed patients show increased plasma and CSF levels. Furthermore, while in animals the cognitive enhancing effect of DHEA(S) administration has been clearly established, cognition in humans is almost unaffected after DHEA supplementation (Barnhart et al., 1999). These discrepancies require further elucidation.

57 Chapter 4 | Neurosteroids in depression: a review

Chapter 5 | Effects of PhD examination stress on allopregnanolone and cortisol Plasma levels and peripheral benzodiazepine receptor density

58 Chapter 5 | Effects of PhD examination stress on allopregnanolone and cortisol Plasma levels and peripheral benzodiazepine receptor density

Droogleever Fortuyn HA, Van Broekhoven F, Span PN, Bäckström T, Zitman FG, Verkes RJ. Psychoneuroendocrinology (2004) 29:1341–1344

Chapter 5 | Effects of PhD examination stress on allopregnanolone and cortisol plasma levels and peripheral benzodiazepine receptor density

Abstract

Peripheral benzodiazepine receptor (PBR) density in blood platelets and plasma allopregnanolone concen- tration in humans were determined following acute stress as represented by PhD examination. Fifteen healthy PhD students participated. Heart rate, blood pressure, and plasma allopregnanolone, plasma cortisol, and PBR density, were measured at different time points. Allopregnanolone and cortisol concentration and PBR density were significantly increased during examination. A positive correlation between allopregnanolone and PBR density was found.

61 Chapter 5 | Effects of PhD examination stress on allopregnanolone and cortisol plasma levels and peripheral benzodiazepine receptor density

Introduction

The peripheral benzodiazepine receptor (PBR), which is located in the central and peripheral nervous sys- tem, is involved in steroidogenesis and is considered to be an important potential therapeutic target in psychi- atric disorders (Ansseau et al., 1991). The PBR, located at the outer mitochondrial membrane, plays a role in the translocation of cholesterol from the outer to the inner mitochondrial membrane (Krueger and Papadopoulos, 1990), which is the rate limiting step for the synthesis of (neuro)steroids (Krueger and Papadopoulos, 1992). Allopregnanolone is a metabolite of progesterone and a potent positive allosteric modulator of the gamma- aminobutyric acid receptor type A (GABA(A) receptor) (Majewska et al., 1986). It has anxiolytic, anticonvulsive, and hypnotic/sedative properties (Rupprecht and Holsboer, 1999). The anxiolytic effect appears to depend on the presence or absence of ovarian steroids (Laconi et al., 2001). Because allopregnanolone can both accumu- late in the brain irrespective of supply from peripheral endocrine organs and can be synthesized de novo from sterol precursors, it is called a neurosteroid (Baulieu et al., 1999). Neurosteroids are involved in many central nervous system disorders like depression, anxiety, learning and memory dysfunctions, and epilepsy (Majewska, 1992; Mellon, 1994; Bicikova et al., 2000). In animals, brain and plasma allopregnanolone levels, and in animals and humans, platelet PBR density, both increase following acute stress (Purdy et al., 1991; Drugan, 1996). No conclusive data exist with respect to plasma allopregnanolone levels following acute stress in humans (Altemus et al., 2001; Girdler et al., 2001). Furthermore, the relation between PBR density and allopregnanolone plasma levels following acute stress has not been investigated. The mechanism by which PBR density rapidly increases following acute stress is not known. It could be that there exist PBR containing vesicles that are released by fu- sion of the vesicles with the cell membrane. If this is true and how acute stress causes this exocytosis remains to be elucidated. In the present study, we tried to replicate previous findings of increased PBR density following acute stress as represented by an examination situation (Karp et al., 1989). Furthermore, we examined our hypothesis that in acute stress, plasma allopregnanolone concentration increases and that this increase is correlated with the increase in PBR density. Recently, it has been shown that allopreganolone levels decrease during panic provoca- tion in patients with panic disorder but not in controls (Strohle et al., 2003). Baseline levels in these patients are increased (Strohle et al., 2002), specifically in the early follicular phase of the menstrual cycle (Brambilla et al., 2003a). These findings suggest that during an induced panic attack, patients with panic disorder fail to maintain compensatory increased allopregnanolone levels. Considering the anxiolytic effect of allopregnanolone and the important role of PBR agonists in the biosynthesis of allopregnanolone, a positive correlation between the two during acute stress may support the suggestion that synthetic PBR agonists may prove to be potent and effec- tive new anxiolytic drugs.

61 Chapter 5 | Effects of PhD examination stress on allopregnanolone and cortisol plasma levels and peripheral benzodiazepine receptor density

Methods

Fifteen healthy PhD students (12 men and 3 women), with a mean age of 35 years (range 29-41 years), were recruited by means of advertisements. No abnormalities were found during physical and psychiatric examina- tion. After complete description of the study to the subjects, written informed consent was obtained. In brief, PhD examination in the Netherlands constitutes the presentation and defense of one’s scientific work in front of a board of professors. After the presentation and defense is a break of 15 minutes in which the board retires and reaches a conclusion about the PhD student having passed the examination or not. After the break the board returns and informs the PhD student about its conclusion. PhD examination is considered to be a stressful event and was therefore used in the present study as a model of acute stress (van Rood et al., 1991). Stress was measured by changes in cortisol plasma concentration, blood pressure (BP), and heart rate (HR).

Plasma levels of allopregnanolone, and the maximal number of binding sites in blood platelets (Bmax) for the PBR specific ligand3 H-PK11195 were measured. All measurements took place four weeks before the PhD examination (T1), 45 minutes before the examination (T2), during the examination (T3), and four weeks after the graduation (T4). Blood samples at T2 were drawn between 12.30 h and 15.00 h; blood samples at T3 were drawn at 14.30 or 16.30 h. The preparation of platelets was essentially as described earlier (Gavish et al., 1986). The resulting blood platelet fractions were used for the PBR assay, the (plasma) supernatant was used for the hormone measure- ments. Both were kept at – 80 °C until analysis. The PBR binding assay was performed essentially as described earlier (Gavish et al., 1986) with some modifications. Separation of bound from free 3H-PK11195, was performed as described earlier (Benraad and Foekens, 1990). The inter assay coefficient of variation was 11.7%. Plasma allopregnanolone concentration was measured by radioimmunoassay earlier described in detail (Bixo et al., 1997). The cortisol levels were measured by an immunofluorometric assay (IFMA), the time-resolved fluoroimmunoassay technique (Delfia) using the reagent kit 1244-060 DELFIA® Cortisol Kit (Wallac OY, Turku, Finland) according to the manufacturers instructions. Serum mixed from both women and men were used as controls in the Delfia assay. The intra assay coefficient of variation in the cortisol assay was 5 % and inter assay 4.5%. The analytical sensitivity was 15 nmol/l serum. The inter assay coefficient of variation for the allopregna- nolone assay was 8.5% and the detection limit 18 pg. In the three women, all measurements took place in the follicular phase except for one measurement: baseline measurement (time point 1) in one woman took place in the luteal phase. As a (physiologically) higher allopregnanolone value was found at this time point compared to the other time points (time points 2-4) in this woman, the difference in phase has not confounded the results of this study (high allopregnanolone levels found during the PhD examination (time point 3) in most of the subjects). Differences in means at the different time points and regression coefficients and their corresponding Pear- son correlation coefficients were statistically tested using mixed model analyses of variance (ANOVA) with sub- jects as random factor and time points as fixed factor. Post-hoc comparisons between time points were tested

63 Chapter 5 | Effects of PhD examination stress on allopregnanolone and cortisol plasma levels and peripheral benzodiazepine receptor density

with paired t-tests. In order to adjust for time of blood sampling at T1 and T4, all samples from all time points (T1 –T4) were divided into two groups: samples drawn before or after 12.00 hours.

Results

Results of Bmax of PBR binding, allopregnanolone and cortisol are presented in Figure 1. No effect of time of blood sampling could explain the significant differences in allopregnanolone and cortisol concentrations and Bmax.

There was a statistically significant effect of time for HR (F = 17.6, df = 3, 41, p<0.001) with T3 > T1,T4; systolic BP (F = 28.53, df = 3, 41, p<0.001) with T3 > T1,T4; and diastolic BP (F = 12.38, df = 3, 41, p<0.001) with T3

> T1,T4.

-5 Bmax was significantly correlated with allopregnanolone plasma concentrations (r = 0.35; B = 3.3x10 ; F = 5.92; df = 1, 44; p = 0.02). Cortisol plasma concentrations were significantly correlated with allopregnanolone plasma concentrations (r = 0.40; B = 598; F = 7.97; df = 1, 44; p = 0.007), and systolic (r = 0.40; B = 11; F = 8.08; df = 1, 43; p = 0.007) and diastolic BP (r = 0.38; B=18; F = 6.89; df = 1, 43; p = 0.012).

64 Chapter 5 | Effects of PhD examination stress on allopregnanolone and cortisol plasma levels and peripheral benzodiazepine receptor density

Figure 1. PBR density in blood platelets and plasma allopregnanolone and cortisol levels. Bars indicate means and error bars one standard deviation. Time points on the x-axis (1-4) refer to: four weeks before PhD examination (1); 45 minutes before PhD examina- tion (2); during PhD examination (3); and four weeks after PhD graduation (4). B max (maximal number of binding sites in blood platelets for the PBR specific ligand 3H-PK11195) expressed in 10 -11mol/mg protein, allopregnanolone in nmol/l, and cortisol in μmol/l. Signifi- cant effect of time (H) for PBR density (F=2.94, df=3, 42, p<0.05) with T3 > T 1,T2; plasma allopregnanolone (F=6.38, df=3, 42, p<0.01) with T 3 > T 1,T2,T4; and plasma cortisol (F=5.27, df=3, 42, p<0.01) with T 3 > T 1,T4

65 Chapter 5 | Effects of PhD examination stress on allopregnanolone and cortisol plasma levels and peripheral benzodiazepine receptor density

Chapter 6 | Effects of allopregnanolone Discussion on sedation in men, and in women on oral contraceptives The results showed that allopregnanolone in humans is increased following acute stress. Furthermore, this is the first study to show that, in acute psychological stress, the increase in plasma allopregnanolone concentration is correlated with an increase in PBR density in blood platelets. The fact that during the PhD examination, plasma cortisol concentration, BP, and HR reached their highest scores, supported our assumption that PhD examina- tion is a valid model for acute stress. The increase in PBR density during the examination in the present study is in accordance with previous findings. Karp and collegues have shown increased PBR densities in residents in psychiatry immediately after the acute stress of board examination (Karp et al., 1989). The increase in plasma cortisol during acute stress was not correlated with the increase in PBR density. This suggests that acute stress-evoked increases in cortisol concentrations do not depend on rapid biosynthe- sis. Considering the correlational nature of this study, no causal relationship between the increase in PBR density and allopregnanolone concentration can be drawn from the results. Further research is needed to demonstrate if synthetic PBR agonists enhance the biosynthesis of allopregnanolone and, subsequently, if they may be of use in the treatment of anxiety disorders.

66 Chapter 6 | Effects of allopregnanolone on sedation in men, and in women on oral contraceptives

Van Broekhoven F, Bäckström T, Van Luijtelaar G, Buitelaar JK, Smits P, Verkes RJ. Psychoneuroendocrinology (2007) 32:555-564

Chapter 6 | Effects of allopregnanolone on sedation in men, and in women on oral contraceptives

Abstract

Allopregnanolone is a known GABA(A) receptor agonist not previously given to men, or to women using oral contraceptives (OC). The effects of metabolites of sex hormones on the GABA(A) receptor are different between men and women. OC are known to change GABA(A) receptor subunit composition. These factors might play a role in the differential effect of allopregnanolone in men and women, and in women with or without OC. To study the sedative effect of and sensitivity to allopregnanolone in men and in women with OC, nine healthy men (mean age 24.6 years) and nine healthy women on OC (mean age 21.8 years) were given three, increasing, intravenous dosages (0.015, 0.03, and 0.045 mg/kg) of allopregnanolone. Saccadic eye velocity (SEV), sub- jective ratings, and electroencephalography (EEG) were used to evaluate the response to allopregnanolone. Repeated blood samples for analyses of serum allopregnanolone levels were drawn throughout the study day. Allopregnanolone decreased SEV more in women than in men, and increased subjective ratings of ‘sedation’. The results in women on OC are similar to earlier results in women without OC. Subjective ratings of ‘contented- ness’ decreased in men but increased in women. Serum levels of allopregnanolone were more highly increased in men compared to women. Other pharmacokinetic parameters were not different between sexes. On the EEG, beta power increased in men. In conclusion, men and women on OC reacted differently to allopregnanolone.

69 Chapter 6 | Effects of allopregnanolone on sedation in men, and in women on oral contraceptives

Introduction

Neuroactive steroids like allopregnanolone can be produced through metabolisation of progesterone in the ovaries and adrenals, and also directly in the brain (Baulieu, 1998). Allopregnanolone is a potent positive modulator of the GABA(A) receptor complex (Majewska et al., 1986), more potent than benzodiazepines or barbiturates now known today (Norberg et al., 1999). Consequently, it has sedative, anxiolytic, and anticonvulsive properties, as has been established mainly in rat models (Crawley et al., 1986; Kokate et al., 1994; Norberg et al., 1999). Recently, sedation has also been confirmed in humans (Timby et al., 2006). The role of allopregnanolone in the pathogenesis and treatment of psychiatric disorders is gaining more interest and is becoming progressively clearer (van Broekhoven and Verkes, 2003; Wihlback et al., 2006).

Sedation is a factor very much linked to the GABA(A) receptor. Many anaesthetic drugs used for induction or mainte- nance of anaesthesia act on the GABA(A) receptor (Orser et al., 2002). Sedation can be measured objectively by use of sac- cadic eye velocity (SEV) (Schiller and Tehovnik, 2003; 2005). A high correlation occurs between self-ratings of sedation, and objective measurement of change in SEV (Hommer et al., 1986). Therefore, SEV is a good tool to measure differences in GABA(A) receptor sensitivity between groups of subjects. In the investigation of the differences of GABA(A) receptor sensitivity between the sexes, one has to take into account the fact that sex steroids and their metabolites change the excitability of the nervous system. Therefore, men may considerably differ from women, as women have such large vari- ations in progesterone and allopregnanolone levels during the menstrual cycle. One way to overcome the problem of the cyclic hormone variations in women is to investigate women on oral contraceptives (OC), as the endogenous production of sex steroid hormones will be inhibited by the exogenously given steroids in the OC (Rapkin et al., 2006). As a result, the hormonal milieu will be more stable, since the variation between the individual women due to the day-to-day variation of the sex steroids can be minimized, and is more similar to the situation in men.

There are, however, some drawbacks with the OC approach, since neuroactive steroids and precursors to neuroac- tive steroids are decreased in women taking OC (Paoletti et al., 2004; Kurshan and Neill Epperson, 2006). Further, in animal studies OC can change GABA(A) receptor subunit gene expression (Follesa et al., 2002), and both factors could influence the sedative response to allopregnanolone.

We have the opportunity to compare our data on the effects of allopregnanolone in women with OC to the effects in women without OC, as we collected data in women without OC during the follicular phase, who were given the same al- lopregnanolone dosages, and in whom the serum concentration of allopregnanolone was measured (Timby et al., 2006). It was found that allopregnanolone had the expected sedative effect as measured by decreased SEV, and an increased score on subjective rating of sedation.

Objective electroencephalography (EEG) data may support the subjective central effects: pharmaco-EEG shows distinct patterns of frequency spectra in different classes of drugs. GABA agonists but also benzodiazepines show a typical EEG profile with an increase of sigma and spindle activity commonly reported (Lancel, 1999). In the present open pilot study we have also added behavioural tests to investigate acute effects on behavioral measures by allopregnanolone. Our hypoth- esis was that there was no difference in the effects of allopregnanolone between women on OC and men.

70 Chapter 6 | Effects of allopregnanolone on sedation in men, and in women on oral contraceptives

Materials and methods

Subjects

In this study, allopregnanolone was given intravenously for the first time ever to men. This study was planned in parallel with a study in women without OC, and accordingly used the same protocol (Timby et al., 2006). A stepwise dosage increase was demanded by the regulatory authorities as a safety measure due to fact that the allopregnanolone effect was not known.

Nine healthy men and nine healthy women between the ages of 18 and 40 years were screened for inclusion in the study. See Table 1 for age and body weight. They were recruited through advertisement on the campus of the Rad- boud University Nijmegen. All women used combined monophasic oral contraceptives containing /lev- onorgestrel. Two men and one woman were smokers. The exclusion criteria were any use of benzodiazepines or other psychoactive drugs during the last three months, daily use of any prescribed drug during the last four weeks, use of any prescribed medication within the last week, except paracetamol, use of any illicit drug during the last four weeks (e.g. cannabis, amphetamines, cocaine, morphine mimetic or alcohol of more than 6 units in one day). Further exclusion cri- teria were any current or previous somatic disease, any mental disorder, including premenstrual syndrome (retrospec- tive report), or a history of drug abuse. The presence of psychiatric disorders was evaluated using a semi-structured psychiatric interview, the Mini-International Neuropsychiatric Interview (MINI, version 2.1). Before inclusion, physical examination was performed, as well as electrocardiography and routine blood chemistry and hematology screens.

All subjects had normal electrocardiograms and blood screens and all female subjects had negative pregnancy tests.

Urine was tested on the presence of illicit drugs. All subjects had negative drug tests. No night work or transmeridian trips were allowed during the week before the study day. All subjects gave written informed consent prior to inclusion in the study. The study procedures were in accordance with ethical standards for human experimentation, established by the Declaration of Helsinki of 1975, revised in 1985. The study protocol was approved by the ethics committee of the

Radboud University Nijmegen Medical Centre.

Table 1. Differences in baseline values between men (n=9) and women (n=9).

Age Body weight SEV ‘Sedation’ ‘Contentedness’ ‘Calmness’ Sigma power Beta power

a b Men 24.6±2.14 67.4±2.6 396±4.5 223±24 249±14 33±8 0.817±0.0486 3.167±0.207 Women 21.8±0.86 60.0±1.7 390±9.6 257±48 102±17 40±11 - -

Age, years; body weight, kg; SEV: saccadic eye velocity, degrees/s; ‘Sedation’, ‘Contentedness’, and ‘Calmness’ (visual analogue scales, mm); Sigma and Beta power (power of electroencephalographic frequencies, μvolt2; values are given as mean ± SEM; a p<0.05; b p<0.001

71 Chapter 6 | Effects of allopregnanolone on sedation in men, and in women on oral contraceptives

Study protocol

Subjects were instructed to refrain from consuming food or drinks (except water) from 2200h the day prior to the day of the experiment. Consumption of caffeine, tobacco and chocolate was restricted throughout the study day. Subjects entered the research facility at 0800h, where they received a light standardized breakfast. An intravenous cannula was inserted in each forearm, and blood samples were taken for baseline levels of al- lopregnanolone in serum. To establish baseline, SEV measurements, subjective ratings, and, in men, EEG record- ings, were made and a blood sample was drawn. At 1000h, three intravenous injections of allopregnanolone were given at 30-min intervals, using dosages of 0.015, 0.03, and 0.045 mg/kg, thus giving a cumulative dosage of 0.09 mg/kg. The injections were given slowly, in more than 30 s. The Umeå University Hospital Pharmacy prepared the experimental medications. Intravenous allopregnanolone solution was formulated with purified allopregnanolone, UC 1009 (Umecrine AB, Box 7984, 907 19 Umeå, Sweden) 15 mg dissolved in 100 ml albumin solution (Pharmacia, Stockholm, Sweden,

200 mg/ml) using an ultrasound bath. The solution contained (mean ± SEM) 0.130 ± 0.003 mg/ml of allopregna- nolone in men (n=9) and 0.127 ± 0.001 in women (n=9). There was no difference in concentration between sexes. The allopregnanolone concentration of each batch of solution was determined using HPLC and UV absorbance (Turkmen et al., 2004). After each allopregnanolone injection, SEV recordings, subjective ratings, and, in men, EEG recordings, were made at 5 and 15 min respectively. At the time of SEV measurements, serum for measuring allopregna- nolone levels was drawn from the arm contralateral to that used for drug administration. Additional blood sam- pling was performed at 35, 45, 65, 75, 90, 120, 180, 300, and 330 min. SEV recordings, subjective ratings, and, in men, EEG recordings, were performed until 90 min. Subjects were allowed to walk in the research facility after 120 min.

Saccadic Eye Movements

Saccadic eye movements were measured using electrooculography as a neurophysiological measure for sedation, and were analyzed using a micro computer-based system (Cambridge Electronic Design, Cambridge, England) for sampling and analysis of eye movements according to van Steveninck (van Steveninck et al., 1999). The electrodes were placed 1 cm lateral of the outer canthus of both eyes, with one common electrode on the mastoid. Electrode impedance was less than 5 kΩ. Head movements were restrained with a fixed head support. The target consisted of an array of light emitting diodes on a bar fixed at 50 cm in front of the eyes. For the saccadic test, the target displaced suddenly from left to right in a sequence of 29 targets, from right to left respectively at random intervals between 4 and 8 seconds, corresponding to 30 degrees of eyeball rotation (15 degrees from middle to left–right). A personal computer controlled the target movements and digitized the

72 Chapter 6 | Effects of allopregnanolone on sedation in men, and in women on oral contraceptives

waveform using an analogue-digital converter. The 29 individual EOGs were stored and analyzed off-line. First, the digitized data were processed to locate saccades. To avoid pre-emptive saccades and blinking artifacts, only saccades initiated 50 to 400 ms after target movements were included. Also, to be considered a saccade, the recorded eye movement had to display a velocity of more than 100 deg/s. Second, each saccade was analyzed to determine the peak saccadic velocity (van Steveninck et al., 1999).

Mood Rating Scales

The Mood Rating Scale (MRS) was administered as a subjective rating of drug effects. The MRS consists of 16 visual analogue scales from which three factors are derived: ‘sedation’, ‘calmness’ and ‘contentedness’ (Bond and Lader, 1974a).

EEG recording

EEG frequency spectrum recordings were made only in men, because of no available EEG equipment at the time of testing in women. A Cz electrode was mounted on the head of the subject and referenced to the left mastoid. A ground electrode was placed on the forehead. Skin resistance was reduced to less than 4 kΩ. An Artisan amplifier was used for filtering and for amplification of the EEG signal. The high-pass filter was at 0.1 Hz and the low-pass filter at 120 Hz. A WinDaq (Dataq Instruments) acquisition system was used. The sample frequency was 204.8 Hz. A custom-made software package, Spectranes, was used for the spectral analyses of the EEG (Philip van den Broek, Elecronic Research Group, Radboud University Nijmegen), with corrections for eye movements. During EEG recording, the subject lay in a comfortable chair with his eyes open for 2 min. The following categorization in frequency bands was applied: delta1 (0.5-2 Hz), delta2 (2-4 Hz), theta (4-8 Hz), alfa1 (8-10 Hz), alfa2 (10-12 Hz), sigma (12-15 Hz), beta (15-30 Hz).

Hormone assays

Allopregnanolone was measured by radioimmunoassay (RIA) after diethylether extraction and HPLC pu- rification of samples. The purification of samples has been described earlier (Turkmen et al., 2004). Serum samples were analyzed in duplicates. Recovery was determined for each assay, using 300-500 cpm of tritium- labeled allopregnanolone, [9,11,12-3H(N)]-5α-pregnan-3α-ol-20-one (Perkin Elmer Life Sciences, Boston, USA) added to a serum sample before extraction, and by measuring the amount recovered after HPLC. The recovery of allopregnanolone averaged 98%, and the results are compensated for recovery.

73 Chapter 6 | Effects of allopregnanolone on sedation in men, and in women on oral contraceptives

Extraction

Serum (0.2 – 0.4 ml) was pipetted into a cylindrical flat-bottomed glass vial of 20 ml volume; after which water (0.5 ml) and (3.0 ml) were added. The samples were then allowed to stand on an orbital shaker for ten minutes. Following the liquid-liquid extraction, the vials were transferred into an ethanol/dry ice bath. The water phase was frozen, and the ether phase was decanted and evaporated under a stream of gas.

High Performance Liquid Chromatography (HPLC)

Evaporated samples were resolved in 1 ml ethanol: water 1:1 (V/V) prior to analysis. Our HPLC system consisted of a Waters 1515 Isocratic Pump, delivering the mobile phase ( : water, 60:40, V/V) at a flow rate of 1.0 ml/min. A Waters 717 plus Auto-sampler was used for injection of samples (200µ l) into a Sym- metry C18 3.5 µm 4.6 x 75 mm separation column (Waters), heated to 45°C in a Waters 1500 Column Heater. Detection of retention times of standards and cross-reacting steroids was at 206 nm using a Waters 2487

Dual λ Absorbance Detector. The detector output was recorded on a PC-based Waters Breeze Chromatography Software (version 3.20). In the preparative HPLC, fractions were symmetrically collected around the retention time for allopregnanolone, and retention was found from injection of a standard sample before the start of analysis. A Waters Fraction Collector II was used for collection of samples for further analysis with RIA, and it was possible to separate all cross-reacting steroids. The retention time of the HPLC is given in a previous publication (Turkmen et al., 2004).

Allopregnanolone – radioimmunoassay

All samples were analyzed using a polyclonal rabbit antiserum raised against 3α-hydroxy-20-oxo-5α-preg- nan-11-yl carboxymethyl ether coupled to bovine serum albumin (Purdy et al., 1990). The cross-reactivity profile of the anti-serum was given in the study by Purdy, and none of the cross-reacting steroids was close to allo- pregnanolone in our HPLC (Purdy et al., 1990; Turkmen et al., 2004). The antiserum was used in a dilution of 1/5000 and the antibody solutions were prepared in the same way, as described earlier (Timby et al., 2006). The sensitivity of the assays was 25 pg, with an intra-assay coefficient of variation for allopregnanolone of 6.5% and an interassay coefficient of variation of 8.5%.

74 Chapter 6 | Effects of allopregnanolone on sedation in men, and in women on oral contraceptives

Pharmacokinetic analyses

Before the pharmacokinetic calculations were carried out, the baseline allopregnanolone concentration

(C0) was subtracted from the measured values obtained after the first injection. Only the net concentrations were used for the pharmacokinetic analyses. Baseline concentrations were less than 2% of the maximum con- centration. Pharmacokinetic parameters were calculated by means of the GraphPad Prism version 3.02 program (GraphPad Software, Inc.). The parameter estimates describing the distribution phase slopes of the log-concen- tration of allopregnanolone (λ) were calculated using the best-fit regression line with five concentration/time observations included (measurements at 65, 75, 90, 120, and 180 min). Half-life in the distribution phase (t1/2) was calculated as ln2/λ. Distribution volume was also calculated in the distribution phase. The serum concentra- tion value at the time of the third injection was extrapolated from the serum concentration 5 min later. Areas under the curve (AUC) were calculated with nine concentration/time observations included (measurements at 5, 15, 35, 45, 65, 75, 90, 120, and 180 min).

Statistical methods

SEV, MRS scores, and EEG recordings were calculated as delta scores (difference from baseline at each time point). For normally distributed continuous variables, analysis of variance (ANOVA) with repeated meas- ures was used with time-point as within-subject factor and sex (men and women) as between group factor. As post-hoc test, least significance difference test was used. When the assumption of normal distribution was not appropriate, non-parametric tests were used (the Mann-Whitney U-test for independent observations, and the Wilcoxon Matched-Pair Signed-Rank Test for related observations). The SPSS statistical package (version 11.5 under Windows) was used for all analyses. P-values < 0.05 were considered statistically significant.

75 Chapter 6 | Effects of allopregnanolone on sedation in men, and in women on oral contraceptives

Results

Body weight in men was significantly higher than in women (p<0.05). Age was not statistically significantly different between sexes (Table 1).

Combined analysis of data in men and women

Serum levels of allopregnanolone increased dose-dependently (F(9,144)=39.83; p<0.001) (Figure 1). Post- hoc analysis showed that all dosages led to increased serum levels compared to baseline (p<0.001). The second and third dosage further increased serum levels (p<0.001 and p<0.01 respectively). There was no significant interaction between serum levels and sex, but there was a statistically significant main sex difference: men had higher serum levels than women (F(1,16)=9.68; p<0.01). Allopregnanolone induces a significant dose-dependent reduction in SEV (F(7,112)=13.19; p<0.001) (Figure 1). Post- hoc analysis showed that all dosages gave increased reduction in SEV compared to baseline (p<0.001). There was a further reduction in SEV after the second and third dosage (p<0.05 and p<0.01 respectively). There was a significant interaction between dosage and sex: the effects of the dosage were greater in women as compared to men (F(7,112)=2.29 ; p<0.05) but there was no main sex difference. For the MRS factor ‘sedation’ there was a significant dose-dependent effect of allopregnanolone (F(7,112=9.91; p<0.001) (Figure 1). Post-hoc showed that all dosages gave increased ‘sedation’ compared to baseline (p<0.01, p<0.01, p<0.001). The first and second dosages were not different, but there was a further increase in ‘sedation’ after the third dosage compared to first and second dosage (p<0.01 and <0.001 respectively). There was a sig- nificant interaction between dosage and sex: men scored higher on ‘sedation’ than women after the dosages of allopregnanolone (F(7,112)=3.74; p< 0.01), but there was no main sex difference. For the MRS factor ‘contentedness’ there was no statistically significant dose-dependent effect of the sub- stance (Figure 2). There was a significant interaction between dosage and sex: men showed decreased scores on ‘contentedness’ after the dosages of allopregnanolone while women showed increased scores (F(7,112)=4.64 ; p < 0.001) and there was a significant main sex difference (F(1,16)=11.33 ; p<0.01). For the MRS factor ‘calmness’ no statistically significant changes were found (Figure 2).

76 l / l o m n

, Chapter 6 | Effects of allopregnanolone on sedation in men, and in women on oral contraceptives e n o l o n 200 a n 150 g e r 100 p o 50 l l A (nmol/l) Allopregnanolone 0 0 25 50 75 100

s / g 0 e d -25 , V E -50 ΔS (degrees/s) SEV Δ -75 0 25 50 75 100

m m

, 500 n o i t a 250 d e (mm) Sedation

ΔS Δ 0

0 25 50 75 100 Time, minutes

Figure 1. The serum concentration-time profile of allopregnanolone (top panel) was significantly different between men and women (p<0.05). The third dosage of allopregnanolone induced in women a significantly greater decrease in saccadic eye velocity (ΔSEV) than in men (p<0.05; middle panel). There is a dosage-dependent increase in the Mood Rating Scale factor ‘sedation’ (ΔSedation) related to increased dosage (p<0.001; bottom panel). Values for men (n=9) are shown as solid lines; values for women (n=9) are shown as dashed lines. Values are given as mean ± SEM. Dosages of allopregna- nolone were 0.015, 0.03, and 0.045 mg/kg consecutively, and the times of administration are indicated by arrows.

77 Chapter 6 | Effects of allopregnanolone on sedation in men, and in women on oral contraceptives

s e n 100 d e t 50 n e 0 t n o -50 ΔC Contentedness (mm) Contentedness

Δ -100 0 25 50 75 100 m m

, 25 s s 15 e n 5 m l -5 a C (mm) Calmness Δ -15 Δ

-25 0 25 50 75 100

Time, minutes

Figure 2. Allopregnanolone effect on change in the Mood Rating Scale factors ‘contentedness’ (ΔContentedness, top panel) and ‘calmness’ (ΔCalmness, bottom panel) in men (solid lines; n=9) and women (dashed lines; n=9). In men, there was a significant decrease in ‘contentedness’ after the first dosage (p<0.05), while in women ‘contentedness’ was significantly higher in the meas- urements 5 min after the second until 30 min after the third dosage compared to baseline (p<0.05 – p<0.01). ‘Calmness’ showed no statistically significant changes. Values are given as mean ± SEM. Dosages of allopregnanolone were 0.015, 0.03, and 0.045 mg/kg consecutively, and the times of administration are indicated by arrows.

78 Chapter 6 | Effects of allopregnanolone on sedation in men, and in women on oral contraceptives

Sex differences

Allopregnanolone serum levels significantly increased in both men and women (F(9,72)=18.56; p<0.001 and F(10,80)=27.51; p<0.001 respectively) (Figure 1). In men and women, the post-hoc analyses revealed that there was an effect already with the lowest dosage (p<0.001). The second and third dosages increased levels further (p<0.01 and p<0.05 respectively, in men, and p<0.001 and p<0.01 respectively, in women). Serum allopregnanolo- ne levels at 5 and 15 min. following injections increased significantly more in men than in women after the first (F(1,16)=11.88; p<0.01) and second (F(1,16)=7.434; p<0.05) dosage, and slightly more after the third (F(1,16)=4.11; p=0.060) (Figure 1). Baseline values were significantly higher in men compared to women (p<0.001) (Table 2). There was a significant reduction in SEV in both men and women (F(7,56) = 7.67; p<0.001 and F(11,88) = 6.47; p < 0.001 respectively) (Figure 1). Considering the significant interaction between time and sex, further tests were carried out. In men, the post-hoc test revealed an effect of allopregnanolone already with the lowest dosage (p<0.01). After the second dosage, SEV was further reduced compared to the first dosage (p<0.05). The third dosage was not different from the first or second dosage. In women there was also an effect already with the lowest dosage (p<0.01). The second dosage did not show the expected further reduction on SEV. The third dosage, however, reduced SEV compared to the second and first dosage, and compared to baseline (p<0.05, p<0.001, and p<0.001 respectively). SEV decreased significantly more in women at 5, 15, and 30 min. following the last dosage than in men (F(1,16)=5.24; p<0.05) (Figure 1). Baseline values did not differ between sexes (Table 1). ‘Sedation’ significantly increased in both men and women (F(7,56) = 8.79; p<0.001 and F(11,88) = 6.11; p < 0.001 respectively) (Figure 1). In men, the post-hoc test revealed an effect already with the lowest dosage (p<0.05). After the second dosage ‘sedation’ was increased compared to baseline (p<0.05) and after the third dosage ‘sedation’ further increased compared to the second dosage and to baseline (p<0.01 and p<0.01 respec- tively). In women, there was no significant effect of the first and second dosage. After the last dosage, however, ‘sedation’ was significantly higher than baseline and after first and second dosage (p<0.05, p<0.01, p<0.01 re- spectively). ‘Sedation’ was significantly higher in men than in women at 5 and 15 min. following the first dosage (F(1,16)=5.31; p<0.05) (Figure 1). Following the second and third dosages there were no significant main sex dif- ferences. Baseline scores were not different between men and women (Table 1). ‘Contentedness’ decreased significantly in men (F(7,56) = 6.81; p<0.001) (Figure 2). Post-hoc in men re- vealed that there was a significant decrease already after the first dosage, and it continued to be lower than baseline during the rest of the test period (p<0.01). There was, however, no difference due to increase in dosage. In women, there was a significant change in ‘contentedness’ over the test period (F(11,88) = 3.20; p<0.01). The post-hoc analysis revealed that the ‘contentedness’ in the measurements 5 min after the second dosage until 30 min after the third dosage of allopregnanolone was significantly higher than baseline (p<0.05 – p<0.01). There was a significant difference between sexes for ‘contentedness’ at 5 and 15 min. following the first, second, and third dosage: men had decreased scores of ‘contentedness’ while women had increased scores (F(1,16)=10.77;

79 Chapter 6 | Effects of allopregnanolone on sedation in men, and in women on oral contraceptives

p<0.01, F(1,16)=9.51; p<0.01, and F(1,16)=11.57; p<0.01 respectively) (Figure 2). Baseline scores were significantly higher in men compared to women (p<0.001) (Table 1). Baseline scores of ‘calmness’ were not different between men and women (Table 1).

EEG data

There was a significant dose-dependent increase in the sigma and beta EEG bands in men (F(7,56)=5.54; p<0.001 and F(7,56)=4.50; p<0.001 respectively) (Figure 3). Post-hoc analysis showed that only the last dosage gave a significant increase in sigma power compared to baseline; this increase lasted 15 min (p<0.05). The last dosage also gave a significant increase in beta power, but this effect lasted only 5 min (p<0.05). The sigma power 5 min after the last dosage was also significantly higher compared to all other time points (p<0.01 – p<0.05). 15 Min after the last dosage, sigma power was also significantly higher compared to the first dosage (p<0.05), and significantly lower compared to the third dosage (p<0.05). Next, sigma power further decreased in time: the decrease between 15 and 30 min after the last dosage was significant (p<0.05).

t 2.0 l 2, o 1.5 r2 v μe 1.0 volt wμ -o 0.5 ΔP Power, Power, 0.0

GΔ E -0.5 E EEG - EEG -1.0 0 25 50 75 100 K`d\#d`elk\j

Figure 3. Allopregnanolone effect on sigma (solid lines) and beta (dashed lines) EEG frequency bands in men (n=9). Only the last dosage gave a significant increase in sigma power, which lasted until 15 min (p<0.05), and in beta power, which lasted until 5 min (p<0.05). Values are given as mean ± SEM. Dosages of allopregnanolone were 0.015, 0.03, and 0.045 mg/kg consecutively, and the times of administration are indicated by arrows.

80 Chapter 6 | Effects of allopregnanolone on sedation in men, and in women on oral contraceptives

Pharmacokinetics

Allopregnanolone levels at baseline and 5 min following first (p<0.001) and second (p<0.05) dosage were significantly higher in men compared to women (Table 2). Maximum serum allopregnanolone level wasalso higher in men. Distribution volume, t1/2, and AUC adjusted for body weight did not differ between sexes. No serious adverse effects were observed. Three men reported sleepiness and tiredness, three reported feeling as if they had taken alcohol, and one reported flushing. In the women, two reported sleepiness, four reported feeling as if they had taken alcohol, and one complained of a mild frontal headache.

Table 2. Differences in pharmacokinetic parameters between men (n=9) and women (n=9) after three injections of allopregnalone, representing a cumulative dosage of 0.09 mg/kg.

Men (mean±SEM) Women (mean±SEM)

3 C0 (nmol/l) 2.38 ± 0.18 0.41 ± 0.08

1 C5 (nmol/l) 35.44 ± 4.68 24.34 ± 1.60

1 C35 (nmol/l) 95.83 ± 10.57 61.57 ± 6.08

C65 (nmol/l) 145.31 ± 22.50 100.02 ± 10.68

1 C120 (nmol/l) 50.6 ± 7.2 34.9 ± 4.8

2 C180 (nmol/l) 43.9 ± 10.3 25.2 ± 9.3

C300 (nmol/l) 10.1 ± 2.0 13.8 ± 3.1

1 Cmax (nmol/l) 149.9 ± 20.7 100.6 ± 10.3

Time to Cmax (min) 66.1 ± 1.1 61.6 ± 3.3

t1/2 (min) 18.9 ± 5.3 21.9 ± 4.7 Vd (l) 13.9 ± 2.3 16.7 ± 2.7

-1 1 AUC (nmol l min) 11067 ± 1187 7580 ± 692 -1 AUC/kg (nmol l min)/kg 164.9 ± 18.1 126.2 ± 10.8

C0, endogenous (baseline) serum concentration; C5, C35, C65, C120, C180, C300, serum concentrations at 5, 35, 65, 120, 180, and 300 min after the first allopregnanolone injection (second and third allopregnanolone injection were at 30 and 60 min after the first allopregnanolone injection); Cmax, maximum serum allopregnanolone concentration; t1/2, half-life during the first 2 hours of distribution; Vd, distribution volume; AUC, area under the serum concentration/time curve; AUC/kg, AUC adjusted for body weight; 1 p<0.05, 2 p<0.01, 3 p<0.001

81 Chapter 6 | Effects of allopregnanolone on sedation in men, and in women on oral contraceptives

Discussion

The results of this study show that intravenously-administered allopregnanolone decreases SEV more in women than in men after three increasing dosages. Conversely, allopregnanolone increases subjective ratings of ‘sedation’ more in men than in women, but only after the first dosage. Following the administration of the second and third dosage, ratings of ‘sedation’ were similar. In men, SEV was dose dependently reduced after the first and second dosage but not after the third, while in women SEV was dose dependently reduced after the first and third dosage. Men rated their feelings of ‘sedation’ as significantly increased after each of the three dosages, while women rated them only after the last, highest, dosage as significantly increased. The results of the difference in allopregnanolone levels are discussed below. The reasons for the difference in SEV and ‘sedation’ in men and women are not known and are not further investigated in this paper. However, one can speculate that the accumulation in the brain of allopregnanolone does differ between men, and women using OC. The accumulation in the brain is not diffuse, but varies between the different regions in the brain, and the allopregnanolone variance in the menstrual cycle is reflected in the brain concentrations (Bixo et al., 1997). One can further speculate that the higher concentration of testosterone in men could decrease allopreg- nanolone content, and competes with allopregnanolone on the GABA(A) receptor on the allopregnanolone site (Hosie et al., 2006), as the 5 alpha–reduced metabolite of testosterone is less potent as a GABA(A) receptor agonist compared to allopregnanolone and thus would act as a partial antagonist on the GABA(A) receptor (Rah- man et al., 2006). However, ‘sedation’ showed opposite effect in men and women, making the result difficult to interpret. The results in the present study suggest that there can be a dissociation between SEV and sedation, and that this dissociation differs between men and women. A dissociation between SEV and sedation has been shown earlier, for instance concerning thyroid hormone and diphenhydramin (Glue et al., 1992; Hopfenbeck et al., 1995). The dissociation of the different reactions between men and women is interesting and deserves further research. The decrease in SEV in the women on OC in the present study was similar to the decrease in SEV in women on normal menstrual cycles investigated during the follicular phase (Timby et al., 2006). The maximum effect was exactly the same while the second dosage of 0.03 mg/kg gave a slightly larger effect in women without OC compared to the effect seen in the present study. In a previous study in which the same cumulative dosage of 0.09 mg/kg of allopregnanolone was used in women as in the present study, a maximum percent decrease in SEV of 13.2 was found (Timby et al., 2006). This finding is in line with the present study, in which a decrease of 16.6% was found in women (in men 9.34%). In the present study there was no decrease in SEV between 0.015 and 0.03 mg/kg dosages. This suggests that in the present study an acute tolerance was obtained at the second injection, and that the tolerance was released to the third injection. That allopregnanolone can induce acute tol- erance has earlier been shown in rat experiments, and tolerance arrives within 90 min of injections (Birzniece et al., 2006). Acute tolerance is also seen after midazolam infusion in women (Sundstrom et al., 1997b) and in rats after allopregnanolone infusion (Birzniece et al., 2006). Subjective sedation was rated slightly more in the study

82 Chapter 6 | Effects of allopregnanolone on sedation in men, and in women on oral contraceptives

of Timby et al. (2006) than in the present study (450 and 300 mm, respectively). Plasma allopregnanolone levels 5 min after each injection reached slightly higher levels in the present study as compared to Timby et al. (2006): 20, 50 and 90 nmol/l and 20, 50 and 74 nmol/l, respectively. A possible explanation for the lower effect of allopregnanolone on ‘sedation’ in women on oral contraceptives compared to women without oral contracep- tive use is that oral contraceptives can induce tolerance to GABA(A) receptor substances and thereby a decrease in the GABA(A) receptor sensitivity (Rapkin et al., 2006). In men, there was a decrease in SEV after the first and second dosage but no further decrease after the third dosage. The fact that in men, the third dosage gave no dose-dependent decrease in SEV, indicates again development of acute tolerance. The subjective ratings of ‘sedation’ show a similar response with a strong ‘seda- tion’ after the first dosage but no further increase after the second. Women were more sensitive than men, as shown by a greater decrease in SEV, although women had lower serum allopregnanolone concentrations. This is in line with previous findings of higher sensitivity in female rats for the anticonvulsive effect of allopregnanolo- ne, compared to male rats (Finn and Gee, 1994). According to our knowledge, there is no other study concerning the effect of allopregnanolone on SEV and sedation in men. There is, however, one study in which the effect of progesterone on sedation and smooth pursuit eye movements was measured (Soderpalm et al., 2004). The results of that study are similar to the ones obtained in this paper, and as it is known that progesterone activity is mainly due to the conversion to allopregnanolone, we consider this as support for our findings. There is a significantly lower baseline level of allopregnanolone in women compared to men. The level is similar to the level seen in women without OC during the follicular phase (Timby et al., 2006), but higher than the concentration seen in postmenopausal women (Andreen et al., 2006a). The levels of allopregnanolone are, however, lower than those seen in the luteal phase of women (Wang et al., 1996). It has earlier been shown that OC decrease the levels of allopregnanolone during the premenstrual phase in women to the levels seen during the follicular phase in women with normal cycles (Follesa et al., 2002; Paoletti et al., 2004). A lower level of al- lopregnanolone in brain tissue has also been shown during OC treatment in animal studies (Follesa et al., 2002). In men, baseline levels have been reported to be between 1 to 4 nmol/l, showing that our levels are within the earlier published range (Droogleever Fortuyn et al., 2004; Brambilla et al., 2005; Holdstock et al., 2006). The allopregnanolone levels differ beween men and women, even though allopregnanolone was given adjusted for body weight. This could be due to difference in body distribution or fat to body weight ratio. In addition, the use of OC in women may contribute to the lower serum allopregnanolone levels. OC are known to increase the of steroids in the liver (Rang et al., 2003). The glucuronisation is the last step of me- tabolism of both the 3alpha-hydroxy metabolites of progesterone and testosterone. These two steroids compete for the same enzyme before the steroids are excreted to the urine (Ganong, 1991). As women have much lower levels of testosterone than men, allopregnanolone may be glucuronised and excreted more easily in women. This competition might also explain in part the difference in allopregnanolone levels between women and men. Concerning the comparison between women with and without OC, a similar situation occurs when the steroids are metabolised with glucuronisation and the OC use the same pathway to be excreted to urine. As the OC are

83 Chapter 6 | Effects of allopregnanolone on sedation in men, and in women on oral contraceptives

resorbed by the gut, the liver receives a very high concentration during the passage, which will be of importance for the metabolic capacity of steroids in women taking OC. Men had lower scores on ratings of ‘contentedness’ than women following allopregnanolone. In men, ‘con- tentedness’ decreased already after the first dosage, and remained decreased throughout the test period. Prob- ably this finding reflects the statistically significant higher ‘contentedness’ in the men at baseline. In women, ‘contentedness’ increased already after the second dosage and stayed up for almost one hour. At the same time women show a decrease in ‘calmness’, giving a picture difficult to interpret. This could be an effect related to sedation. Allopregnanolone increased beta and sigma frequencies on the EEG in men, in a dose-dependent manner, an effect which lasted for at least 5 and 15 min respectively. The decrease in alpha power on the EEG recordings was lacking, the increase in sigma and beta power that was seen in this study is in line with the pharmaco-EEG profiles that were previously found in rats after the administration of benzodiazepines such as diazepam (Coenen and van Luijtelaar, 1989), allopregnanolone (Lan- cel et al., 1997b) and other GABA-mimetics (Lancel et al., 1997a), supporting the GABA–mimetic effects of al- lopregnanolone. It could be possible that certain types of behaviour, the type of mental activity or the vigilance state might contribute to whether all drug effects on the spectral content of the EEG can be easily noticed. Our subjects were sitting with open eyes, which might be less suitable for observing alpha activity, or for a putative reduction in alpha and or delta power. The drug-induced changes in beta (and sigma) activity are very typical for hypnotic drugs of the benzodiazepine-type, and these changes occur independently of the behavioural state of the subject (Saletu et al., 1987; Romano-Torres et al., 2002). In conclusion, our data on women with OC are in line with previously-published data in women without OC (Timby et al., 2006), i.e. allopregnanolone causes sedation, as measured with SEV and subjective sedation. In addition, our data shows that allopregnanolone can also be safely administered to men. There are, however, differences in response between women and men. Women are more sensitive to allopregnanolone than men, according to SEV results. Subjective ratings of sedation were different between men and women, and subjective ratings of contentedness were higher in women. Power of beta and sigma EEG frequencies increased in line with human and animal data. Further research is needed before conclusions can be drawn with respect to the clinical implications of the present findings.

Acknowledgements

Support for the study has been obtained from Radboud University Nijmegen, the Swedish Research Coun- cil, Medicine, (product number 11198), and from EU an objective 1 grant. We would like to thank drs. Miriam Kos and Elisabeth Zingmark for their technical assistance.

84 Chapter 7 | Oral progesterone decreases saccadic eye velocity and increases sedation in women

Van Broekhoven F, Bäckström T, Verkes RJ. Psychoneuroendocrinology (2006) 31:1190-1199

Chapter 7 | Oral progesterone decreases saccadic eye velocity and increases sedation in women

Abstract

The aim of this study was to investigate the neurophysiological and behavioural effects of a single dose of progesterone in women. Allopregnanolone is a metabolite of progesterone and a potent positive modulator of the GABAA receptor and produces sedative and anxiolytic effects. This study was designed to examine the ef- fect of oral progesterone and the metabolite allopregnanolone in women. Women (n=15) in their follicular phase received oral progesterone (400 mg) or placebo. Dependent measures included plasma levels of progesterone and allopregnanolone, saccadic eye velocity (SEV), subjective ratings (visual analogue scales), and reaction time. Administration of progesterone decreased SEV and increased sedation. This effect is probably due to enhanced GABA activity.

87 Chapter 7 | Oral progesterone decreases saccadic eye velocity and increases sedation in women

Introduction

Allopregnanolone is a metabolite of progesterone and enhances inhibitory neurotransmission in the brain by binding to the GABAA receptor complex (Majewska et al., 1986). Allopregnanolone has sedative and anxiolytic properties in animals (Wieland et al., 1991; Wang et al., 1995), and is increased following stress (Droogleever et al., 2004) and changes during the menstrual cycle (Wang et al., 1996). Orally administered progesterone is substantially converted to allopregnanolone. In this way, progesterone can be regarded as a pro-drug for allo- pregnanolone (Andreen et al., 2006a). As allopregnanolone appears to be involved in some psychiatric disorders (van Broekhoven and Verkes, 2003), it would be of interest to know if progesterone can be used as a medica- tion treatment of these disorders. So far, however, data on the relationship between plasma allopregnanolone levels induced by oral progesterone and pharmacodynamic effects have been inconsistent (Arafat et al., 1988; Freeman et al., 1993; de Lignieres et al., 1995; Vanselow et al., 1996). However, a benzodiazepine like effect of progesterone on sleep in males has been shown earlier (Friess et al., 1997; Grön et al., 1997; Lancel et al., 1996). The absorption and metabolism of progesterone varies significantly between individuals (Arafat et al., 1988; Freeman et al., 1993). Several speculations on the reasons for the variations have been presented but so far no evidence for a single reason is available. Consequently, differences between women in plasma allopregna- nolone levels and pharmacodynamic effects after oral progesterone are to be expected. Because progesterone is a female reproductive hormone, only women were recruited for the present study. Also, affective and anxi- ety disorders have a higher prevalence in women than in men, making progesterone more likely as a potential medication treatment in this group. Saccadic eye velocity (SEV) is related to sedation and strongly controlled by the GABA(A) system (Schiller and Tehovnik, 2003; 2005). There is still some controversy on whether acute progesterone and its metabolite allopregnanolone changes emotional responses in humans. By combining an CNS effect that has been proven to be influenced by allopregnanolone (SEV) and measurements of mood factors we can more rely on the results obtained in the psychometric tests. In the present study, the aim is to investigate the effects of oral progesterone on plasma levels of pro- gesterone and allopregnanolone and on saccadic eye velocity and behavioural measurements. As secondary aim, we wanted to know the degree and timing of progesterone rise after progesterone administration and the latency to allopregnanolone rise. A third aim was to find out if the stress of saccade measurements and behav- ioural testing would be enough for a stress response in progesterone and allopregnanolone concentrations. For this aim the placebo treatment was analysed separately. By dividing the group on the basis of the progesterone plasma levels after oral progesterone, concentration-effect relationships can be obtained while using only one dosage. Women were tested in the follicular phase because in this phase endogenous levels of progesterone and allopregnanolone are low and stable compared to the luteal phase.

88 Chapter 7 | Oral progesterone decreases saccadic eye velocity and increases sedation in women

Materials and methods

Subjects

Fifteen healthy women between the ages of 18 to 40, 21.3 ± 2.1 years, (mean ± standard deviation) were recuited for the study. They were recruited through advertisement at the campus of the Radboud University Nijmegen. Women who used hormonal forms of contraception were excluded. Further exclusion criteria were any use of benzodiazepines or other psychopharmaca during the last three months, daily use of any prescribed drug during the last four weeks, use of any prescribed medication within the last week, except paracetamol, use of any illicit drug during the last four weeks (e.g. cannabis, amphetamines, cocaine, morphine mimetic or alcohol of more than 6 units in one day). Women planning to become pregnant were excluded. Other exclusion criteria were any current or previous somatic disease, any mental disorder, in- cluding premenstrual syndrome, or a history of drug abuse and they had to refrain from alcohol use during the last five days prior to test. Signs of malnutrion was an exclusion criterion. The presence of psychiatric disorders was evaluated using a semi-structured psychiatric interview, the Mini-International Neuropsychi- atric Interview (MINI, version 2.1). Before inclusion, physical examination was performed, as well as elec- trocardiography and routine blood chemistry and hematology screens. All subjects had normal electrocar- diograms and blood screens and all female subjects had negative pregnancy tests. Urine was tested on the presence of illicit drugs. All subjects had negative drug tests. No night work or transmeridian trips were allowed during the week before the study day. All subjects gave written informed consent prior to inclusion in the study. The study procedures were in accordance with ethical standards for human experimentation, established by the Declaration of Helsinki of 1975, revised in 1985. The study protocol was approved by the ethics committee of the Radboud University Nijmegen Medical Centre.

Study protocol

Subjects were tested in a double-blind, placebo-controlled, randomised, crossover study, in two consecu- tive menstrual cycles. Time of testing was between day 6 and 12 of the follicular phase. Subjects were instructed to refrain from taking food or drinks (except water) from 2200h the day prior to the experimental day. Caffeine, tobacco and chocolat use was restricted throughout the study day. Subjects entered the research facility at 0800h, where they received a light standardized breakfast. An intravenous cannula was inserted in the forearm for blood sampling. To establish baseline, SEV measurements, Mood Rating Scale scores, and reaction times were recorded and a blood sample for baseline levels of progesterone and allopregnanolone was drawn. There- after, at 1000h, subjects received orally either 400 mg of oral micronized progesterone (Progestan (Organon)) or placebo, by four identical capsules. After ingestion of the study drug, plasma sampling, SEV recordings, Mood

89 Chapter 7 | Oral progesterone decreases saccadic eye velocity and increases sedation in women

Rating Scale scores, and reaction time tasks were performed at 30, 60, 90, 120, 150, 180, 240, and 360 min, respectively. Subjects were allowed to ambulate in the research facility after the 240 min measurements.

Saccadic Eye Movements

Saccadic eye movements were recorded as a neurophysiological measure for sedation and were assessed using a micro computer-based system (Cambridge Electronic Design, Cambridge, England) for sampling and analysis of eye movements according to Van Steveninck (1999). Electrode impedance was less than 5 kΩ. Head movements were restrained with a fixed head support. The target consisted of an array of light emitting diodes on a bar fixed at 50 cm in front of the eyes. For the saccadic test, the target displaced suddenly from left to right, respectively, from right to left at random intervals between 4 and 8 seconds, corresponding to 30 degrees of eyeball rotation to both sides (15 degrees from middle to left–right). As outcome parameter, we calculated the mean peak velocity of the saccadic eye movements, which is a highly sensitive parameter for sedative benzodi- azepine effects (van Steveninck et al., 1999).

Mood Rating Scales

The Mood Rating Scale was administered in order to measure subjective drug effects. The Mood Rating Scale consists of 16 visual analogue scales from which three factors are derived: ‘sedation’, ‘calmness’ and ‘con- tentment’ (Bond and Lader, 1974a).

Reaction Time Tasks

Patients had to perform a simple and a complex reaction time task. The reaction time is considered as a motor performance index, but it gives also an indication of attention and concentration. The simple reaction time task consists of pressing a button as fast as possible when a round bull at the computer screen changed into a square. In the complex reaction time task, subjects had to press the button with their left and right index fingers, respectively, according to the message on the screen. Each session consisted of 20 trials at random intervals. The mean reaction time per session was calculated. Reaction times of individual trials above two standard deviations of the mean of a session were excluded, as well as anticipations, defined as responses faster than 125 ms for the simple reaction time task and 175 ms for the complex reaction time task.

90 Chapter 7 | Oral progesterone decreases saccadic eye velocity and increases sedation in women

Hormone assays

Allopregnanolone was measured by radioimmunoassay (RIA) after diethylether extraction and HPLC pu- rification of samples. The purification of the sample has been described earlier (Turkmen et al., 2004). Plasma samples are analyzed in duplicates. Recovery was determined for each assay, using 300-500 cpm of tritium- labeled allopregnanolone, [9,11,12-3H(N)]-5α-pregnan-3α-ol-20-one (Perkin Elmer Life Sciences, Boston, USA) added to a plasma sample before extraction, and by measuring the amount recovered after HPLC. The recovery of allopregnanolone averaged 98% and the results are compensated for recovery.

Extraction

Plasma (0.2 – 0.4 ml) was pipetted into a cylindrical flat bottom glass vial of 20 ml volume, where after water (0.5 ml) and diethyl ether (3.0 ml) were added. The samples were then placed on an orbital shaker for ten minutes. Following the liquid-liquid extraction, the vials were transferred into an ethanol/dry ice bath. The water phase was frozen, and the ether phase was decanted and evaporated under a stream of nitrogen gas.

High Performance Liquid Chromatography (HPLC)

Evaporated samples were resolved in 1 ml ethanol: water 1:1 (V/V) prior to analysis. Our HPLC system con- sisted of a Waters 1515 Isocratic Pump, delivering the mobile phase (methanol : water, 60:40, V/V) at a flow rate of 1.0 ml/min. A Waters 717 plus Auto-sampler was used for injection of samples (200 µl) into a Symmetry C18 3.5 µm 4.6 x 75 mm separation column (Waters), heated to 45°C in a Waters 1500 Column Heater. Detection of retention times of standards and cross-reacting steroids was at 206 nm using a Waters 2487 Dual λ Absorbance Detector. The detector output was recorded on a PC-based Waters Breeze Chromatography Software (version 3.20). In the preparative HPLC fractions were symmetrically collected around the retention time for allopregna- nolone, retention was found from injection of a standard sample before the start of analysis. A Waters Fraction Collector II was used for collection of samples, for further analysis with RIA and all cross-reacting steroids were possible to separate.

Allopregnanolone – radioimmunoassay

All samples were analyzed using a polyclonal rabbit antiserum raised against 3α-hydroxy-20-oxo-5α-preg- nan-11-yl carboxymethyl ether coupled to bovine serum albumin, (Purdy et al., 1990). The antiserum was used in

91 Chapter 7 | Oral progesterone decreases saccadic eye velocity and increases sedation in women

a dilution of 1/5000 and the antibody solutions were prepared in the same way, as described earlier (Timby et al., 2006). The sensitivity of the assays was 25 pg, with an intra-assay coefficient of variation for allopregnanolone of 6.5% and an interassay coefficient of variation of 8.5%.

Progesterone

Measurements of plasma progesterone were made by using Delfia kits (Wallac Oy, Turku, Finland), a fluor- oimmunoassay, according to the manufacturer’s instructions. Intra-assay coefficient of variation is 3.6% and inter-assay coefficient of variation 5.2%.

Statistical methods

SEV, Mood Rating Scale scores, and reaction time tasks were calculated as delta scores (difference from baseline at each time point). For normally distributed continuous variables, analysis of variance (ANOVA) with repeated measures was used with time (time-points) as within-subjects factor and drug (progesterone and pla- cebo) as between group factor. In the analysis between high and low absorption groups, absorption (high and low) was the between group factor. The division was made using the technique of median split on the basis of the indiviual progesterone plasma level area under the curve. When appropriate, non-parametric tests were used as Mann-Whitney U-test for independent observations and the Wilcoxon Matched-Pair Signed-Rank Test for related observations. The SPSS statistical package (version 11.5 under Windows) was used for all analyses. P-values < 0.05 were considered statistically significant.

92 Chapter 7 | Oral progesterone decreases saccadic eye velocity and increases sedation in women

Results

There were no statistically significant differences between the progesterone and placebo condition with respect to baseline values for progesterone or allopregnanolone plasma levels (Table 1), SEV, Mood Rating Scale scores, and reaction time tasks (Table 2). The plasma progesterone-time relationship is shown in Figure 1 (bottom panel). Plasma levels in combined analysis were different between placebo and progesterone condition (F(1,14)=19.17; p=0.001). There was a sig- nificant interaction between drug and time (F(8,112)=2.06; p=0.046). Post hoc analysis shows that ingestion of progesterone increased serum levels compared to placebo 90 minutes after ingestion and remained increased compared to placebo till the end of the study day, i.e. 360 min after ingestion (p<0.05). The peak of the plasma concentration of progesterone arrived at 120 min which was 60 min earlier than the allopregnanolone peak. The metabolism to allopregnanolone showed a great variance between the individuals and some had a delay of up to 120 min before allopregnanolone started to rise. The allopregnanolone levels in samples from 180 min to 360 min were significantly higher than in the placebo treatment (F(1,10)=10.11; p=0.025). When the progesterone and placebo condition were investigated separately the analysis revealed a signifi- cant difference in both placebo and progesterone conditions (F(8,112)=2.16; p=0.036 and F(8,112)=2.06; p=0.046, respectively). In the placebo condition, the post hoc analysis revealed that progesterone plasma levels were significantly higher compared to baseline at 30 en 90 minutes and 180 and 360 min after ingestion of the pla- cebo capsule. In the progesterone condition, progesterone plasma levels were significantly higher compared to baseline at 90 minutes after ingestion and remained increased till the end of the study day (p=0.029 to p<0.001). Plasma level at the end of the study day (i.e., 360 min after ingestion) was significantly lower compared to that at 240 min after ingestion (p=0.011). The plasma allopregnanolone-time relationship after progesterone ingestion is shown in Figure 1 (bot- tom panel). Plasma levels showed no variance during the placebo condition (F(8,24)=0.361; p=0.931). Plasma levels increased after ingestion of progesterone (F(8,112)=4.13; p<0.001). Post hoc analysis shows that ingestion of progesterone increased serum levels compared to baseline 120 min after ingestion and remained increased compared to baseline till the end of the study day, i.e. 360 min after ingestion (p=0.042 to p<0.001). With respect to SEV, there was a significant interaction between drug and time. Progesterone decreased SEV more than placebo (F(8,112)=2.54; p=0.014) (Figure 1, top panel). Post hoc analysis shows that ingestion of progesterone decreased SEV compared to placebo at 90 min, 150 and 180 min after ingestion (p<0.05). When the progesterone and placebo condition were investigated separately the analysis revealed a significant differ- ence in both placebo and progesterone conditions (F(8,112)=3,36; p=0.002 and F(8,112)=7.93; p<0.001, respec- tively). In the placebo condition, the post hoc analysis revealed that SEV was significantly lower compared to baseline at 90 minutes after ingestion of the placebo capsule and remained lower till the end of the study day (p=0.028 to p=0.003). In the progesterone condition, the post hoc analysis revealed that SEV was significantly lower compared to baseline at 30 minutes already after ingestion of the progesterone and remained lower till

93 Chapter 7 | Oral progesterone decreases saccadic eye velocity and increases sedation in women

Table 1. Baseline and maximum progesterone and allopregnanolone plasma levels (mean±SEM, nmol/l) in placebo and progester- one conditions in follicular phase.

Progesterone Progesterone Allopregnanolone Allopregnanolone baseline maximum baseline maximum

Placebo 0.638±0.106 1.65±0.20 0.76±0.23 0.95±0.15 Progesterone 1.008±0.413 414.05±90.50 0.89±0.18 236.59±55.95

Maximum progesterone and allopregnanolone concentration were significantly higher during progesterone treatment (P<0.03).

Table 2. Baseline measurements (mean±SEM) in placebo and progesterone conditions in follicular phase.

SEV ‘sedation’ ‘contentedness’ ‘calmness’ SRT CRT

Placebo 387.9±11.6 352±33 183±20 67±7 261±7 316±11 Progesterone 386.1±8.3 303±29 135±15 57±7 265±11 327±14

There were no statistically significant differences between drug conditions. SEV: saccadic eye velocity (degrees/s); ‘sedation’, ‘contentedness’, and ‘calmness’ (mm); SRT: simple reaction time task (ms); CRT: complex reaction time task (ms)

94 Chapter 7 | Oral progesterone decreases saccadic eye velocity and increases sedation in women

) s) / l 0 g / e l d (o -25 m Vn E( S -50 Δn Placebo o i Progesterone

SEV (degrees/s) SEV -75

t Δ a 0 100 200 300 400 r t 300 n e c progesterone n 200 treatment o c 100 a m s placebo a 2 treatment l 1

P (nmol/l) Plasma concentration 0 0 100 200 300 400 Time (min) Allopregnanolone Progesterone

Figure 1. Change in saccadic eye velocity from baseline (ΔSEV) and progesterone and allopregnanolone plasma concentration-time profile after ingestion of progesterone or placebo on t=0 (mean±SEM). ΔSEV was significantly increased following ingestion of progesterone compared to placebo ((F(1,14)=16.22; p=0.001). There was no difference in SEV at baseline between progesterone (filled triangles) and placebo (open triangles) conditions (top panel). There was no difference in progesterone or allopregnanolone plasma concentrations at baseline between progesterone and placebo conditions (bottom panel).

95 Chapter 7 | Oral progesterone decreases saccadic eye velocity and increases sedation in women

the end of the study day (p=0.017 to p<0001). Lowest SEV occurred from 150 to 240 min after ingestion. SEV increased at the end of the study day (360 min after ingestion) compared to 150, 180, and 240 min after inges- tion (p=0.049, 0.009, and 0.024, respectively). For the MRS factor ‘sedation’ there was a significant drug effect (F(1,14)=7,75; p=0.015) (Figure 2, top panel). Post hoc analysis showed a significant increase in ‘sedation’ after progesterone ingestion (p=0.015). When the progesterone and placebo condition were investigated separately the analysis revealed a significant difference in both placebo and progesterone conditions (F(8,112)=6,87; p<0.001 and F(8,112)=2.58; p=0.013, respectively). In the placebo condition, sedation was decreased from 90 minutes after ingestion and remained decreased till the end of the study day (p=0.029 to 0.003). In the progesterone condition, sedation increased to a maximum at 60 min and thereafter decreased and became significantly different at 180 min after ingestion (p=0.031). For the MRS factor ‘contentedness’ there was no statistically significant drug effect in combined analysis of placebo and progesterone condition (Figure 2, middle panel). When the progesterone and placebo condition were investigated separately the analysis revealed a significant difference in both placebo and progesterone conditions (F(8,112)=3,90; p<0.001 and F(8,112)=2.18; p=0.034, respectively). In the placebo condition, ‘content- edness’ was decreased from 90 minutes after ingestion and remained decreased till the end of the study day (p=0.029 to 0.004). In the progesterone condition, ‘contentedness’ increased to a maximum at 60 min (not significant) and thereafter decreased and became significantly different from 60 min after ingestion at 120 min after ingestion and remained decreased till the end of the study day (p=0.037 to 0.004). For the MRS factor ‘calmness’ there was no statistically significant drug effect in combined analysis of placebo and progesterone condition (Figure 2, lower panel). When the progesterone and placebo condition were investigated separately the analysis revealed a significant difference in the placebo (F(8,112)=3,22; p=0.002) but not in the progesterone condition. In the placebo condition, ‘calmness’ was decreased from 30 minutes after ingestion till the end of the study day, except for 120 and 180 min after ingestion (p=0.048 to 0.006).

For simple reaction time and for complex reaction time there was no statistically significant drug effect in combined analysis of placebo and progesterone condition (data not shown). When the progesterone and placebo condition were investigated separately the analysis revealed no significant difference in both placebo and pro- gesterone conditions.

96 Chapter 7 | Oral progesterone decreases saccadic eye velocity and increases sedation in women

m m

, 150 n placebo o 100 progesterone i t 50 a 0 dm em -50

ΔS , -100 Sedation (mm) Sedation

sΔ s -150 e 0 100 200 300 400 n 50 d e 25 t 0 n e -25 t n -50 o

C (mm) Contentedness -75 ΔΔ m m -100 0 100 200 300 400 , s 20 s e 10 n 0 m l -10 a C -20 (mm) Clmness Δ Δ -30

0 100 200 300 400 Time, minutes

Figure 2. Changes in the Mood Rating Scale factors ‘Sedation’ (ΔSedation), ‘Contentedness’ (ΔContentedness), and ‘Calmness’ (ΔCalmness) following ingestion of progesterone and placebo on t=0 (top, middle, and lower panel, respectively). ‘Sedation’ increased significantly after ingestion of progesterone compared to placebo (F(1,14) = 7,75; p = 0.015). Post hoc showed that ‘Contentedness’ decreased 90 min after ingestion of placebo (p = 0.029-0.004) whereas it increased to a maximum at 1 h after ingestion of progesterone (not significant). ‘Calmness’ was decreased 30 min after ingestion of placebo (p = 0.048-0.006). Values are given as mean±SEM.

97 Chapter 7 | Oral progesterone decreases saccadic eye velocity and increases sedation in women

Low absorption compared with high absorption

There was no statistically significant difference in baseline progesterone and allopregnanolone plasma levels (Table 3). Baseline SEV was significantly higher in the low absorption group (p=0.037) while baseline Mood Rating Scale scores and reaction time tasks were not different (Table 4). The plasma progesterone and allopregnanolone-time relationships are shown in Figure 3 (bottom panel). Progesterone plasma levels were significantly higher in the high absorption group compared to the low absorp- tion group in combined analysis (F(1,13)=21,39, p<0.001). Allopregnanolone plasma levels were not different between low and high absorption groups in combined analysis. Maximum progesterone and allopregnanolone plasma levels were significantly higher in the high ab- sorption group (p=0.002 and 0.04, respectively) (Table 3). SEV was significantly decreased in the high absorption group compared to the low absorption group in combined analysis (F(1,13)=13.52; p=0.003) (Figure 3 (top panel)). For the MRS factors ‘sedation’, ‘contentedness’, and ‘calmness’ there was no significant effect in combined analysis between the low and high absorption group. Also, for the simple and complex reaction time tasks there were no significant effects in combined analysis between the low and high absorption group.

Table 3. Baseline and maximum progesterone and allopregnanolone plasma levels (mean±SEM, nmol/l) in progesterone condi- tion in follicular phase divided in low (n=7) and high (n=8) absorption.

Progesterone Progesterone Allopregnanolone Allopregnanolone baseline maximum1 baseline maximum2

Low absorption 0.693±0.218 108.51±31.35 0.86±0.10 120.13±33.76 High absorption 1.284±0.763 681.40±90.00 0.91±0.34 338.50±87.79

1: p=0.002 2: p=0.04

Table 4. Baseline measurements (mean±SEM) in progesterone condition in follicular phase divided in low (n=7) and high (n=8) absorption.

SEV1 ‘sedation’ ‘contentedness’ ‘calmness’ SRT CRT

Low absorption 400.9±10.5 285±34 127±24 65±11 257±12 310±21 High absorption 373.3±11.1 319±46 143±20 51±8 272±19 342±17

SEV: saccadic eye velocity (degrees/s); ‘sedation’, ‘contentedness’, and ‘calmness’ (mm); SRT: simple reaction time task (ms); CRT: complex reaction time task (ms) 1: p=0.037

98 Chapter 7 | Oral progesterone decreases saccadic eye velocity and increases sedation in women

) s / 0 low absorption g) el high absorption d/ (l -25 o Vm En S( -50 Δ

n o SEV (degrees/s) SEV -75 iΔ t 0 100 200 300 400 a r t 500 n prog high abs. e c 400 prog low abs. n o 300 allo high abs. c

a 200 allo low abs. m s 100 a l P 0 (nmol/l) Plasma concentration 0 100 200 300 400 Time (min)

Figure 3. Change in saccadic eye velocity from baseline (ΔSEV) (top panel) and progesterone and allopregnanolone plasma con- centrations (bottom panel) in subjects with high (n=8) and low (n = 7) progesterone absorption (ingestion of progesterone on t = 0). ΔSEV was significantly decreased in the high compared to the low absorption group in combined analysis (F(1,13) = 13.52; p = 0.003). Baseline SEV was significantly higher in the low absorption group (p = 0.037). Maximum progesterone and allopregnanolone levels were significantly higher in the high absorption group (p = 0.002 and 0.04, respectively). There was no statistically signifi- cant difference in baseline progesterone and allopregnanolone plasma levels.

99 Chapter 7 | Oral progesterone decreases saccadic eye velocity and increases sedation in women

Discussion

The present study shows that 400 mg of oral micronized progesterone produces a significant decrease in SEV with a concomittant increase in subjective rating of sedation compared to placebo in healthy women in the follicular phase. Reaction time, contentedness and other MRS factors were not different between progesterone and placebo treatments. The results of this study show that in women progesterone has a sedative effect. This is similar to what has earlier been shown in men (Friess et al., 1997). Our results confirm in women that proges- terone does not change reaction time which has earlier been shown in men (Grön et al., 1997). The study setting shows an increase in progesterone concentration that could be interpreted as a stress response similar to what is seen for allopregnanolone in humans and animals (Droogleever et al., 2004; Purdy et al., 1991). An increase in progesterone due to stress has earlier been shown in men and women taking oral contraceptives but not in cycling women (Wirth et al., 2006). Our results are in contrast to this observation as we see a reaction during the placebo treatment in our study group of women during the follicular phase. During the follicular phase the ovar- ian contribution to the progesterone level is minute and the adrenals have a greater impact. In the progesterone treatment one can expect a latency between the progesterone peak and allopregnanolone peak as progesterone has to be metabolised with two enzymatic steps before allopregnanolone is formed. In this study the latency between progesterone peak and allopregnanolone peak was 60 min and fits with the effect curve of sedation and SEV. When the data were divided in a high and low absorption group, depending on the height of the proges- terone plasma level after oral progesterone, SEV was significantly decreased in the high absorption group com- pared to the low absorption group. Subjective ratings of sedation, contentedness, and calmness, and reaction times, did not differ between groups. One important transmittor system for the regulation of SEV is the GABA system. Enhancement of the

GABA activity will decrease the SEV (Schiller and Tehovnik, 2003; 2005). The lack of effect on sedation in the women who showed a high absorption of oral progesterone might be due to a decreased capability of perception of sedation, regardless of GABA activity. This in turn, can be caused by the development of acute tolerance to allopregnanolone. Acute tolerance to allopregnanolone has been shown to occur in animals during anaesthesia (Zhu et al., 2004). As has been shown earlier, intravenous injections of small dosages of allopregnanolone in women produce decreased SEV and increased scores in subjective rating of sedation (Timby et al., 2006). The levels of allopregnanolone achieved in the present study are much higher (up to 338.5 nmol/l) than those in the orally administered allopregnanolone studies (up to 100.6 nmol/l). This difference in maximum plasma levels is probably due to enhanced metabolisation of progesterone to allopregnanolone in the liver after oral ingestion. Freeman et al. measured allopregnanolone in 18 healthy premenopausal women following an oral dose of 1200 mg micronized progesterone in a double-blind, placebo-controlled study (Freeman et al., 1993). Plasma allopregnanolone was significantly negatively correlated with measures of tension/anxiety and positively cor- related with fatigue and confusion (measured by the Profile Of Mood State) and immediate recall. Significant

100 Chapter 7 | Oral progesterone decreases saccadic eye velocity and increases sedation in women

changes compared to placebo in fatigue, delayed verbal recall and symbol copying were experienced by subjects who achieved high levels of allopregnanolone. Söderpalm administered 200 mg progesterone intramuscularly in

7 women (n=7; age (mean ± standard deviation) 23.3 ± 6.2 years) in the early follicular phase and found modest effects on subjective rating of fatigue and smooth pursuit eye movement with an maximum allopregnanolone plasma level of (mean ± standard deviation) 13.5±4.3 ng/ml, 42 nmol/l (Soderpalm et al., 2004). Both in the progesterone and placebo condition there was a decreased sedation, contentedness, and calm- ness. For sedation there was a significant difference between progesterone and placebo, however not for con- tentedness, and calmness. This indicates that sedation is related to the progesterone treatment but not content- edness and calmness. Although there was a difference in effect between progesterone and placebo on SEV and subjective rating of sedation, there was no difference in effect for the other effect parameters. This might be due to the devel- opment of another progesterone metabolite which is the 3beta variant of allopregnanolone and which is an antagonist of allopregnanolone (Turkmen et al., 2004). That the 3beta-steroid is produced as a conversion of progesterone is also shown (Stromstedt et al., 1993). The fact that the women in the present study did not fall asleep while having tremendously high allopregnanolone plasma levels, support this view. In conclusion, this study is the first that measured the effects of progesterone on SEV, which decreased significantly, when compared to placebo. This effect is most likely due to enhanced GABA activity by the pro- gesterone metabolite allopregnanolone. In spite of very high plasma levels of allopregnanolone, no difference in effect was seen on other outcome measures. This could be due to concomittantly increased plasma levels of the 3beta-variant of allopregnanolone. This should be investigated in another study.

Acknowledgements

Support for the study has been obtained from Radboud University Nijmegen, the Swedish Research Coun- cil, Medicine, (product number 11198), and from EU an objective 1 grant. Elisabeth Zingmark is acknowledged for her technical assistance.

101 Chapter 7 | Oral progesterone decreases saccadic eye velocity and increases sedation in women

Chapter 8 | Increased sensitivity to oral Progesterone in the luteal phase in healthy women

102 Chapter 8 | Increased sensitivity to oral Progesterone in the luteal phase in healthy women

van Broekhoven F, Bäckström T, Buitelaar JK, Verkes RJ (submitted for publication)

Chapter 8 | Increased sensitivity to oral progesterone in the luteal phase in healthy women

Abstract

Knowledge about the degree of sensitivity to progesterone and allopregnanolone in healthy women across the cycle is important in further elucidating the pathophysiology of PMDD. The aim of this study was to investigate whether the effect of a single dose of progesterone on GABA(A) receptor sensitivity and mood differed between the follicular and luteal phase of the normal menstrual cycle. Women (n=9) received oral progesterone (400 mg) or placebo in their follicular and luteal phase in a double- blind, randomized, crossover design. Dependent measures were saccadic eye velocity (SEV), subjective ratings of sedation, contentedness, calmness (visual analogue scales), and reaction times on a simple and complex at- tentional task. Administration of progesterone decreased SEV and increased sedation more in the luteal phases then in the follicular phase. Complex reaction time was more increased in the luteal phase compared to the fol- licular phase after progesterone but this difference was not statistically significant. The effects are probably due to enhanced GABA activity by the progesterone metabolite allopregnanolone. In conclusion: the effect of oral progesterone on GABA(A) receptor sensitivity as measured by SEV is dependent on the phase of the menstrual cycle.

105 Chapter 8 | Increased sensitivity to oral progesterone in the luteal phase in healthy women

Introduction

Mood changes during the menstrual cycle can be a source of distress in some women and lead to clini- cal syndromes such as the premenstrual dysphoric disorder (PMDD) (American Psychiatric Association, 1994). The development of symptoms is related to the presence of progesterone during the luteal phase (Sanders et al., 1983). Progesterone has been shown to increase symptoms of negative mood (Andreen et al., 2006b) and sedation (Merryman et al., 1954). Symptoms of PMDD disappear when the production of progesterone by the corpus luteum, which is formed after ovulation, is prevented by administration of gonadotropin releasing hormone (GnRH) agonists or ovariectomy. In the past, therapeutic interventions consisted of administration of progesterone through various routes, e.g. vaginal, oral, transdermal, to restore assumed dysregulated plasma progesterone concentrations. However, it is currently believed that absolute levels of plasma progesterone con- centrations do not play an important role in the pathophysiology of PMDD. This view is supported by the lack of effect of the progesterone suppletion therapies aforementioned (Freeman, 2004). Nowadays, focus in research in PMDD is laid on a dysregulation of the sensitivity of the GABA(A) receptor complex. Progesterone is metabolized to very potent GABA(A) receptor modulators like allopregnanolone and in animals most of the central nervous system effects by progesterone have been shown to be mediated via this metabolite (Majewska et al., 1986; Mok and Krieger, 1990). Women with PMDD show decreased sensitivity to GABAergic substances like diazepam, pregnanolone and alcohol in the luteal phase compared to the follicular phase and/or compared to healthy control women (Sundstrom et al., 1997a; Sundstrom et al., 1997b; Sundstrom et al., 1998; Nyberg et al., 2004). Decreased sensitivity was shown by reduced saccadic eye velocity (SEV). It thus seems that women with PMDD develop tolerance to the increasing plasma progesterone levels in the luteal phase followed by a withdrawal effect. Premenstrually, the plasma concentration of progesterone already starts to decrease while symptoms of irritability and depressive mood further increase in women with PMDD (Back- strom et al., 2008a). Besides a decreased sensitivity of the GABA(A) receptor complex there seems to exist also a shift of the sensitivity of the receptor complex toward reverse agonism in PMDD: women with PMDD react to the GABA(A) receptor antagonist flumazenil in the luteal phase with increased anxiety while healthy controls show no effect (Le Melledo et al., 2000). One can draw a paralell between PMDD and panic disorder in that pa- tients with panic disorder also respond less to benzodiazepines and show a paradoxical effect to flumazenil (Nutt et al., 1990; Roy-Byrne et al., 1990). In studies investigating the difference in sensitivity to GABAergic substances across the menstrual cycle, healthy control women responded with increased sensitivity to pregnanolone and diazepam in the luteal phase (Sundstrom et al., 1997a; Sundstrom et al., 1998). Monkeys show an increased response to allopregnanolone in this phase of their cycle (Green et al., 1999). No data exist, however, on the influence of the cycle on the sensitiv- ity to progesterone or allopregnanolone in healthy women. We have earlier studied the effect of progesterone and placebo on SEV in the follicular phase but not during the luteal phase in the same patients (van Broekhoven et al., 2006). Knowledge about the degree of sensitivity to progesterone and allopregnanolone in healthy wom-

106 Chapter 8 | Increased sensitivity to oral progesterone in the luteal phase in healthy women

en across the cycle may help in further elucidating the pathophysiology of PMDD. Thus, we conducted a placebo-controlled, double-blind, crossover study in healthy women by administer- ing oral progesterone during both follicular and luteal phases. It was our aim to investigate if the effect of oral progesterone on GABA(A) receptor sensitivity differs between the phases of the menstrual cycle. Our hypothe- sis is that women are more sensitive to oral progesterone in the luteal phase compared to the follicular phase. Sedation is strongly linked to the GABA system and is objectively measured by saccadic eye velocity (SEV) (Schiller and Tehovnik, 2003; 2005). Self-ratings of sedation correlate with SEV (Hommer et al., 1986). SEV and sedation are easy to study and very sensitive to GABA active drugs. Therefore we used SEV and sedation as tools to investigate the sensitivity of the GABA system to oral progesterone. Feelings of contentedness and calmness were rated to gain more information on the background of the possible induction of negative mood changes by progesterone.

Materials and methods

Subjects

Subjects were recruited through advertisements at the campus of the Radboud University Nijmegen. They had to have ovulatory cycles controlled by progesterone values during the luteal phase and regular cycles to be eligible to participate in the present study. Therefore subjects were excluded from participation if they used hormonal forms of contraception. Further exclusion criteria were any use of benzodiazepines or other psycho- tropic medication during the last three months, daily use of any prescribed drug during the last four weeks, use of any prescribed medication within the last week, except for paracetamol, and use of any illicit drug during the last four weeks (e.g. cannabis, amphetamines, cocaine, morphine mimetic or alcohol of more than six units in one day). Women planning to become pregnant were excluded. Other exclusion criteria were any current or previ- ous somatic disease, any mental disorder, including premenstrual syndrome, or a history of drug abuse. Signs of malnutrition were an exclusion criterion. The presence of psychiatric disorders was evaluated using a semi- structured psychiatric interview, the Mini-International Neuropsychiatric Interview (MINI, version 2.1) (Sheehan et al., 1998). Before inclusion, physical examination was performed, as well as electrocardiography and routine blood chemistry and hematology screens. All subjects had normal electrocardiograms and blood screens and all female subjects had negative pregnancy tests. Urine was tested on the presence of illicit drugs. All subjects had negative drug tests. Nine healthy women (age (mean ± standard deviation) 20.9±0.7 years), were included in the study. All subjects gave written informed consent prior to inclusion in the study. The study procedures were in accordance with ethical standards for human experimentation, established by the Declaration of Helsinki of 1975, revised in 1985. The study protocol was approved by the ethics committee of the Radboud University Nijmegen Medical Centre.

107 Chapter 8 | Increased sensitivity to oral progesterone in the luteal phase in healthy women

Study protocol

No night work or transmeridian trips were allowed during the week before the study day and subjects had to refrain from alcohol use during the last five days prior to test. Subjects were tested in a double-blind, placebo-controlled, randomised, crossover study, in separate men- strual cycles. Time of testing was in the follicular phase between day 6 and 12 and in the luteal phase within 10 days prior to the expected onset of menstruation. Subjects were instructed to refrain from taking food or drinks (except water) from 22h the day prior to the experimental day. Caffeine, tobacco and chocolat use was restricted throughout the study day. Subjects entered the research facility at 08h, where they received a light standardized breakfast. An intra- venous cannula was inserted in the forearm for blood sampling. To establish baseline, SEV measurements, Mood Rating Scale scores, and reaction times were recorded and a blood sample for baseline level of progesterone was drawn. Thereafter, at 1000h, subjects received orally either 400 mg of oral micronized progesterone (Pro- gestan (Organon)) or placebo, by four identical capsules. After ingestion of the study drug, plasma sampling, SEV recordings, Mood Rating Scale scores, and reac- tion time tasks were performed at 30, 60, 90, 120, 150, 180, 240, and 360 min, respectively. Subjects were al- lowed to ambulate in the research facility after the 240 min measurements.

Saccadic Eye Movements

Saccadic eye movements were recorded as a neurophysiological measure for sedation and were assessed using a micro computer-based system (Cambridge Electronic Design, Cambridge, England) for sampling and analysis of eye movements according to Van Steveninck (van Steveninck et al., 1999). Electrode impedance was less than 5 kΩ. Head movements were restrained with a fixed head support. The target consisted of an array of light emitting diodes on a bar fixed at 50 cm in front of the eyes. For the saccadic test, the target displaced suddenly from left to right, respectively, from right to left at random intervals between 4 and 8 seconds, cor- responding to 30 degrees of eyeball rotation to both sides (15 degrees from middle to left–right). As outcome parameter, we calculated the mean peak velocity of the saccadic eye movements, which is a highly sensitive parameter for sedative benzodiazepine effects (van Steveninck et al., 1999).

108 Chapter 8 | Increased sensitivity to oral progesterone in the luteal phase in healthy women

Mood Rating Scales

The Mood Rating Scale was administered in order to measure subjective drug effects. The Mood Rating Scale consists of 16 visual analogue scales from which three factors are derived: sedation, contentedness, and calmness (Bond and Lader, 1974a).

Reaction Time Tasks

Patients had to perform a simple and a complex reaction time task. The reaction time is considered as a motor performance index, but it gives also an indication of attention and concentration (Hindmarch, 1980). The simple reaction time task consists of pressing a button as fast as possible when a round bull at the computer screen changed into a square. In the complex reaction time task, subjects had to press the button with their left and right index fingers, respectively, according to the words ‘left’ or ‘right’ at random appearing on the screen. Each session consisted of 20 trials at random intervals. The mean reaction time per session was calculated. Reaction times of individual trials above two standard deviations of the mean of a session were excluded, as well as anticipations, defined as responses faster than 125 ms for the simple reaction time task and 175 ms for the complex reaction time task (Voshaar et al., 2005).

Hormone assay

Plasma progesterone levels were assessed by using progesterone Immulite® solid-phase, competitive chemiluminescent enzyme immunoassay kits. These analysis kits were provided by Diagnostic Products Cor- poration, Corporate Offices, 5210 Pacific Concourse Drive, Los Angeles, CA 90045-6900, USA. The assay was performed according to the manufacturer’s instructions. Intra-assay coefficient of variation is 3.6% and inter- assay coefficient of variation 5.2%.

Statistical methods

Differences in responses on progesterone as compared to placebo between the two menstrual phases were analyzed using linear mixed models for fixed and random effects. In order to account of the dependency of the repeated assessments of the outcome variables within individuals ‘subject’ was taken as a random vari- able. Treatment (progesterone versus placebo), phase (follicular versus luteal), and time (time-points) were fixed factors. The relationships between outcome measures and plasma progesterone levels were also analyzed by

109 Chapter 8 | Increased sensitivity to oral progesterone in the luteal phase in healthy women

means of mixed linear models with menstrual phase and progesterone as fixed independent variables and a random intercept and random slope for each subject . When appropriate, a paired-samples T test was used for related observations. The SPSS statistical package (version 15.0.1 under Windows) was used for all analyses. P- values < 0.05 were considered to be statistically significant.

110 Chapter 8 | Increased sensitivity to oral progesterone in the luteal phase in healthy women

Results

Baseline progesterone concentration was higher in the luteal phase than in the follicular phase (p<0.01). Baseline score for contentedness was significantly higher in the follicular compared to the luteal phase (p<0.05). There were no differences in baseline scores for the other effect variables: SEV, sedation, calmness, and reaction times (Table 1).

Table 1. Baseline measurements (mean±SEM) in follicular and luteal phase.

prog. SEV sedation contentedness calmness SRT CRT

Follicular 0.5 388.99±9.73 351.43±29.22 172.78±17.62 64.56±6.47 260.24±10.42 327.13±15.04 Luteal 16.2 380.75±11.05 311.28±36.88 135.44±24.3 61.72±8.57 260.73±11.07 317.46±16.52

Values represent the means of baseline values of both drug conditions (placebo and progesterone) within phases. There were no differences between these baseline values within phases. There was a significant difference in progesterone plasma concen- tration and in contentedness between follicular and luteal phase (p<0.01 and <0.05, respectively). Prog: progesterone plasma concentration (nmol/l); SEV: saccadic eye velocity (degrees/s); sedation, contentedness, and calmness (visual analogue scale scores) (mm); SRT: simple reaction time task (ms); CRT: complex reaction time task (ms)

After administration of progesterone, there was no difference in the time course over the day of the plas- ma progesterone concentration between the phases (session by phase interaction (F(8,136 ) = 0.416; p>0.05)). Peak concentrations were achieved at 240 min after progesterone ingestion (207 nmol/l in the follicular phase and 264 nmol/l in the luteal phase).

Effects of progesterone versus placebo

SEV was significantly more decreased after administration of progesterone than after placebo (main ef- fect of drug: (F(1,288 ) = 31.94; p<0.001). SEV was more decreased in the luteal phase than in the follicular phase (main effect of phase: (F(1,288) = 20.28; p<0.001). The progesterone effect on SEV was larger in the luteal phase than in the follicular phase (drug by phase interaction: (F(1,288) = 6.20; p=0.013) (Figure 1).

111 Chapter 8 | Increased sensitivity to oral progesterone in the luteal phase in healthy women

m m

, n 200 o 150 i t 100 a 50 d 0 e -50 ΔS

Sedation (mm) Sedation -100 Δ -150 0 100 200 300 400

placebo, follicular phase progesterone, follicular phase

) placebo, luteal phase progesterone, luteal phase s / g 0 e d ( -25

V )E S lΔ -50

/ (degrees/s) SEV

l Δ o -75 m 0 100 200 300 400 n (

e n o 400 r e 300 t 200 s e 100 g o 50

r (nmol/l) Progesterone 25 P 0 0 100 200 300 400 Time, minutes

Figure 1. Changes in sedation (ΔSedation; top panel) and saccadic eye velocity (ΔSEV; (middle panel) after intake on time = 0 of placebo or progesterone in the follicular and luteal phase (mean±SEM). The progesterone effect on sedation was larger in the luteal phase than in the follicular phase (drug by phase interaction: (F(1,288) = 5.05; p=0.025). The progesterone effect on SEV was also larger in the luteal phase than in the follicular phase (drug by phase interaction: (F(1,288) = 6.20; p=0.013). Progesterone plasma levels are shown to indicate follicular and luteal phase (lower panel).

112 m Chapter 8 | Increased sensitivity to oral progesterone in the luteal phase in healthy women m

, s s e n 50 d 25 e t 0 n e -25 t -50 n o -75

C (mm) Contentedness Δ -100 Δ

0 100 200 300 400 placebo, follicular phase progesterone, follicular phase m m placebo, luteal phase progesterone, luteal phase

, 10 s s 0 e n -10 m l -20 a C Δ -30 Calmness (mm) Calmness Δ -40 0 100 200 300 400

60 s 50 m 40 , 30 T 20 R 10 C Δ 0

(ms) CRT -10 Δ -20

0 100 200 300 400

Time, minutes

Figure 2. Changes in contentedness (ΔContentedness; top panel), calmness (ΔCalmness; middle panel), and complex reaction time (ΔCRT; lower panel) after intake on time = 0 of placebo or progesterone in the follicular and luteal phase (mean±SEM). Progester- one induced a significant dose-related increase in complex reaction time (F(1,312) = 19.06; p<0.001). There were no differences in contentedness or calmness.

113 Chapter 8 | Increased sensitivity to oral progesterone in the luteal phase in healthy women

Sedation was significantly more increased after administration of progesterone than after placebo (main effect of drug: (F(1,288 ) = 15.76; p<0.001). There was no difference between the phases (main effect of phase: (F(1,288) = 0.45; p>0.05). The progesterone effect on sedation was larger in the luteal phase than in the follicu- lar phase (drug by phase interaction: (F(1,288) = 5.05; p = 0.025) (Figure 1). Complex reaction time was more increased after administration of progesterone than after placebo, but this effect was not statistically significant (main effect of drug: (F91,288) = 3.64; p = 0.057) (Figure 2). Contentedness, calmness and simple reaction time showed no statistically significant differences.

Effects related to progesterone concentrations

Overall, progesterone induced a significant dose-related decrease in SEV (F(1,146) = 67.49; p<0.001) and a significant dose-related increase in subjective sedation (F(1,147) = 32.89; p<0.001), simple reaction time (F(1,313) = 3.95; p=0.048), and complex reaction time (F(1,312) = 19.06; p<0.001). Progesterone had no effect on content- edness (F(1,311) = 0.07; p>0.05) and calmness (F(1,307) = 3.13; p>0.05). The regression coefficients for SEV and subjective sedation were significantly different in the luteal phase as compared to those in the follicular phase. For SEV the regression coefficient was -0.14 (95% confidence interval -0.18, -0.10) versus -0.07 (-0.15, 0.02), p<0.01. For subjective sedation the regression coefficient was 0.46 (0.31, 0.61) versus 0.10 (-0.24, 0.44), p<0.001. The regression coefficients for simple and complex reaction time did not differ between the phases.

114 Chapter 8 | Increased sensitivity to oral progesterone in the luteal phase in healthy women

Discussion

The main findings of the present study are that administration of oral progesterone led to a decrease of SEV and an increase of subjective ratings of sedation in both phases and that both effects were significantly larger in the luteal phase as compared to the follicular phase. This suggests sensitivity to oral progesterone to be increased in the luteal phase of the menstrual cycle in healthy premenopausal women. Our findings are in line with results from other studies in which healthy premenopausal women showed increased sensitivity to preg- nanolone (Sundstrom et al., 1998) and diazepam (Sundstrom et al., 1997a) in the luteal phase compared to the follicular phase. Comparable to the effects found in these conditions of physiologically increased plasma proges- terone concentrations, i.e. the luteal phase in premenopausal women, is the increased sensitivity to in postmenopausal women when given in combination with progesterone (McAuley et al., 1995). Our results are not due to different pharmacokinetics of progesterone between the phases as the course of the progesterone concentrations over the day in our study did not differ (i.e. there was no session by phase interaction effect). Contentedness was higher at baseline during the follicular phase. A lower baseline level of contentedness in the luteal phase is in line with what is a common finding in this phase of the cycle namely negative mood symptoms (Sveindottir and Backstrom, 2000). There was no drug effect on contentedness and calmness within the phases. This shows that a single acute intake of progesterone does not influence these aspects of mood. This is in line with previous studies which failed to find either an effect of acute progesterone intake (Arafat et al., 1988; Freeman et al., 1993; de Wit et al., 2001). However, after long-term (days) treatment with oral progesterone a negative mood appears (Andreen et al., 2006b). Reaction times were not influenced by progesterone administration. This finding is in line with observa- tions in men that showed progesterone not to affect reaction time (Gron et al., 1997). The increased sensitivity of GABA(A) receptors in the luteal phase explains why withdrawal effects do not occur after the steep decrease in progesterone levels at the end of the luteal phase. The results from the present study support the hypothesis of a difference in GABA(A) receptor sensitivity in the luteal phase between women with and without PMDD. Women with PMDD react differently on GABA(A) receptor active compounds compared to women without PMDD. They show reduced sensitivity to GABAergic substances and inverse agonism to the GABA(A) receptor antagonist flumazenil (Le Melledo et al., 2000). The state of art treatment for PMDD is the prescribing of a selective serotonin reuptake inhibitor (SSRI). The SSRI may be administered continuously or only in the luteal phase (Freeman, 2004). Effects are already present in the first cycle of treatment and there is no withdrawal effect when administered intermittently. The mechanism of action of SSRIs is not known but it has been shown that the reduced sensitivity to a GABAergic compound in women with PMDD is normalized after treatment with a SSRI (Sundstrom and Backstrom, 1998). SSRIs also increase allopregnanolone concentration in plasma and in the brain by influencing the enzyme that is involved in the synthesis of allopregnanolone (van Broekhoven and Verkes, 2003).

115 Chapter 8 | Increased sensitivity to oral progesterone in the luteal phase in healthy women

In conclusion, this study shows that healthy premenopausal women are more sensitive to the sedative effect Chapter 9 | Progesterone selectively increases of oral progesterone in the luteal phase than in the follicular phase. Previous studies have noted the same amygdala reactivity in women difference across the cycle for other GABAergic substances. Future studies should examine whether women with PMDD show decreased sensitivity to oral progesterone in the luteal phase as they do to other GABAergic substances.

116 Chapter 9 | Progesterone selectively increases amygdala reactivity in women

Van Wingen GA, van Broekhoven F, Verkes RJ, Petersson KM, Bäckström T, Buitelaar JK, Fernández G. Mol. Psychiatry (2008) 13:325-333

Chapter 9 | Progesterone selectively increases amygdala reactivity in women

Abstract

The acute neural effects of progesterone are mediated by its neuroactive metabolites allopregnanolone and pregnanolone. These neurosteroids potentiate the inhibitory actions of γ-aminobutyric acid (GABA). Proges- terone is known to produce anxiolytic effects in animals, but recent animal studies suggest that pregnanolone increases anxiety after a period of low allopregnanolone concentration. This effect is potentially mediated by the amygdala and related to the negative mood symptoms in humans that are observed during increased allo- preganonolone levels. Therefore, we investigated with functional MRI whether a single progesterone administra- tion to healthy young women in their follicular phase modulates the amygdala response to salient, biologically relevant stimuli. The progesterone administration increased the plasma concentrations of progesterone and allopregnanolone to levels that are reached during the luteal phase and early pregnancy. The imaging results show that progesterone selectively increased amygdala reactivity. Furthermore, functional connectivity analy- ses indicate that progesterone modulated functional coupling of the amygdala with distant brain regions. These results reveal a neural mechanism by which progesterone may mediate adverse effects on anxiety and mood.

119 Chapter 9 | Progesterone selectively increases amygdala reactivity in women

Introduction

Changes in the blood concentration levels of progesterone and are associated with changes in mood regulation (Steiner et al., 2003) and cognition (Kimura, 1999). For instance, explicit memory (Sharp et al., 1993; Keenan et al., 1998; Buckwalter et al., 1999) and mood (Keenan et al., 1998; Buckwalter et al., 1999; Bennett et al., 2004) de- teriorate during pregnancy when progesterone and estradiol levels reach their highest physiological concentrations.

Mood is also influenced by the more subtle changes in these hormone levels during the menstrual cycle. The negative mood symptoms in premenstrual dysphoric disorder (PMDD) and premenstrual syndrome (PMS) occur during the luteal phase, when progesterone and estradiol concentrations are high, and disappear a few days after the onset of menstrua- tion, when these hormone concentrations are lowest (Backstrom et al., 1983). However, these negative mood symp- toms develop in the absence of abnormal gonadal steroid hormone concentrations (Backstrom et al., 1983; Rubinow et al., 1988). Furthermore, the administration of progesterone or estradiol after gonadal hormone suppression produces negative mood symptoms in women with PMS, but not in women without PMS (Schmidt et al., 1998). The negative mood in PMDD/PMS therefore appears as an abnormal response to normal hormonal changes (Rubinow and Schmidt, 2006).

The neural mechanism by which these hormones influence mood are, however, unknown.

Animal research has shown that progesterone modulates anxiety, and that these effects are likely mediated by its metabolite allopregnanolone (Bitran et al., 1995b; Smith et al., 1998a; Reddy et al., 2005). This neurosteroid potentiates the inhibitory actions of γ-aminobutyric acid (GABA) by modulation of the GABAA-receptor (Majewska et al., 1986). Con- sistent with this GABAergic mechanism, progesterone and allopregnanolone administration usually produce an anxi- olytic effect (Bitran et al., 1995b; Reddy et al., 2005) by action on the amygdala (Akwa et al., 1999). Also pregnanolone, the less potent 5β-stereoisomer of allopregnanolone that has a similar effect (Paul and Purdy, 1992; Zhu et al., 2001b), can produce anxiolysis (Smith et al., 2006). However, it induces anxiogenic effects to aversive stimuli after a period of relative allopregnanolone suppression (Smith et al., 2006), which mimics the low allopregnanolone levels during the human follicular phase (Wang et al., 1996). As the amygdala is involved in a wide range of emotional behavior, and local infusions of GABA antagonists induce anxiogenic effects (Davis and Whalen, 2001b), the amygdala may mediate the anxiogenic effects of allopregnanolone as well.

In the present study, we investigated whether a single progesterone administration to healthy women during their follicular phase increases amygdala reactivity. The amygdala is part of a larger emotion circuitry, which is important for the identification of the emotional significance of stimuli, the generation of an affective response, and emotion regula- tion (Phillips et al., 2003a). We therefore explored whether progesterone’s modulation of amygdala activity influences the functional integration of this system. We used an oral administration of micronized progesterone, which increases allopregnanolone and pregnanolone concentrations to a similar extent (de Lignieres et al., 1995). By administering progesterone instead of comparing neural activity during the luteal (high progesterone and estradiol) and the follicular phase (low progesterone and estradiol), the results of this study provide a mechanistic account for a hormone-brain interaction that is not confounded by (spurious) effects linked to the menstrual cycle.

120 Chapter 9 | Progesterone selectively increases amygdala reactivity in women

Materials and Methods

Subjects

Eighteen naturally cycling women (mean age 24 years; range: 19-39) participated in this study after sign- ing informed consent. They were healthy as determined by routine physical and laboratory examinations, and had no current psychiatric disorder as measured by a structured interview (Sheehan et al., 1998). They were right handed, free of medication, did not use hormonal contraceptives, and reported no history of psychiatric or somatic disease potentially affecting the brain. This study was approved by the local ethics committee (CMO Regio Arnhem-Nijmegen, The Netherlands).

Design and procedure

The subjects participated twice during the early follicular phase (day 2-7) of different menstrual cycles, when endogenous hormone levels are low. They arrived at ~8.15 AM after overnight fast to receive a standard- ized light breakfast, after which 400 mg of micronized progesterone or placebo was administered orally, in a double-blind, cross-over fashion. A venous blood sample was collected prior to drug administration and the scanning session, which started ~90 minutes after drug intake. In addition, subjects completed a mood rating questionnaire (MRS) (Bond and Lader, 1974a) just prior to scanning and a state anxiety questionnaire (STAI) (Spielberger et al., 1970) about 190 minutes after drug intake.

Behavioral task

The experimental paradigm consisted of a blocked design, including an emotion condition and a visuo-mo- tor control condition. This paradigm has been used previously to investigate drug effects on amygdala reactivity (Hariri et al., 2002; Paulus et al., 2005). It robustly engages the amygdala, by contrasting the response to simul- taneously presented angry and fearful face stimuli (http://www.macbrain.org) with the response to ellipses (that consisted of scrambles of the same face stimuli). The results therefore do not show emotion specific effects, but rather a general response to salient, biologically relevant stimuli. Two emotion blocks were interleaved with three control blocks, and each 30 s block consisted of six 5 s trials. Each trial consisted of three simultaneously presented stimuli, with the cue stimulus presented above the target and distractor (Hariri et al., 2002). In the emotion condition, an angry or fearful face was presented on top as cue, and subjects had to indicate by an appropriate button press which of the bottom two faces (one angry

121 Chapter 9 | Progesterone selectively increases amygdala reactivity in women

and one fearful) matched the cue in emotional expression. The three simultaneously presented faces per trial were from different persons from the same sex. Half the trials presented faces of men and half of women, half of each target emotion (angry or fearful). In the control condition, a horizontally or vertically oriented ellipse was presented as cue above two ellipses (one vertical and one horizontal), and subjects had to select the identically oriented ellipse.

Data acquisition

Scanning was performed with a 1.5 T Siemens Sonata MR scanner (Siemens, Erlangen, Germany), equipped with a standard head coil. Seventy-six T2*-weighted EPI-BOLD images were acquired per subject with an echo time of 30 ms to reduce signal dropout, with each image volume consisting of 33 axial slices (3 mm, 0.5 mm slice-gap, TR = 2.290 s, 64 × 64 matrix, FOV = 224 mm, FA = 90°). In addition, a high resolution T1-weighted structural MR image was acquired for spatial normalization procedures (MP-RAGE, TR = 2250 ms, TE = 3.93 ms,

176 contiguous 1 mm slices, 256 × 256 matrix, FOV = 256 mm).

FMRI data analysis

Image analysis was performed with SPM2 (Wellcome Department of Imaging Neuroscience, London, UK). The first five EPI-volumes were discarded to allow for T1 equilibration, and the remaining images were realigned to the first volume. Images were then corrected for differences in slice acquisition time, spatially normalized to the Montreal Neurological Institute T1 template, super-sampled into 2 × 2 × 2 mm3 voxels, and spatially smoothed with a Gaussian kernel of 10 mm FWHM. Statistical analysis was performed within the framework of the general linear model (Friston et al., 1995). For each drug condition, the two experimental conditions were modeled as box-car regressors convolved with the canonical hemodynamic response function of SPM2. In addition, the realignment parameters were included to model potential movement artifacts as well as a high-pass filter (cut-off at 1/128 Hz). In order to account for various global effects, the EPI-data was proportionally scaled. Temporal autocorrelation was modeled with an AR(1) process and the parameter estimates were obtained by maximum likelihood estimation (Friston et al., 2002) to allow departures from sphericity. Single-subject parameter images contrasting the emotion and control condition were obtained and entered into subsequent random-effects analyses. The main effect of task was assessed with a one-sample t-test, and the effect of drug (i.e., drug-by-task interaction) was assessed with a paired t-test. In addition, the effects of drug on the functional integration of the amygdala was investigated using a functional connectivity analysis (Friston et al., 1997). The time-course of amygdala activity was extracted (i.e., the first eigenvariate of a sphere with 6 mm radius around the peak voxel of the right amygdala as identified

122 Chapter 9 | Progesterone selectively increases amygdala reactivity in women

by the main effect of task), and added as a covariate of interest to the model. Single-subject parameter images containing the regression coefficients were obtained and entered into subsequent random-effects analyses. One-sample t-tests were used to identify brain regions with a correlated time-course, and paired t-tests were used to identify the effects of drug on functional connectivity with the amygdala. Statistical tests were family- wise error (FWE) rate corrected for multiple comparisons across the whole brain or the region-of-interest (ROI), using a small volume correction (Worsley et al., 1996), unless stated otherwise. The bilateral amygdala ROI was anatomically defined using the WFU Pickatlas (Maldjian et al., 2003), and the bilateral face responsive regions in the fusiform gyri were defined as spheres with 14 mm radius around previously reported Talairach coordinates (Kanwisher et al., 1997) that were transformed into MNI-space (http://www.mrc-cbu.cam.ac.uk/Imaging/Com- mon/mnispace.shtml). Furthermore, mean amygdala activity in both drug conditions was extracted from the bilateral amygdala ROI for additional correlation analyses (Brett et al., 2002).

Serum analysis

Progesterone was measured with Delfia progesterone kits (Wallac Oy, Turku, Finland) according to the manufacturer’s instructions. Allopregnanolone was measured with radioimmunoassay (RIA) after diethylether extraction and celite chromatography (recovery 78%). The antiserum was directed against 3α-hydroxy-20-oxo- 5α-pregnan-11-yl carboxymethyl ether-BSA with low cross reactivity and a gift from Dr. Robert H. Purdy (Depart- ment of Psychiatry, College of Medicine, University of California, San Diego, CA, USA). The intra-assay coefficient of variation 6.5% and inter-assay coefficient of variation of 8.5% (Andréen et al., 2005).

123 Chapter 9 | Progesterone selectively increases amygdala reactivity in women

Results

The baseline plasma concentrations of progesterone and allopregnanolone did not differ significantly be- tween the progesterone and placebo condition (both p > 0.2). The progesterone administration increased the progesterone concentration from the follicular to the luteal phase range (mean ± SEM, from 1.94 ± 0.29 to 16.83 ± 6.17 nmol/L, t(16) = 2.39, p = 0.03), and increased the concentration of allopregnanolone from 0.60 ± 0.05 to 24.98 ± 7.30 nmol/L (t(16) = 3.35, p = 0.004). This allopregnanolone level is similar to that during early pregnancy (Parizek et al., 2005; Paoletti et al., 2006). However, progesterone had no significant effect on the mood (Bond and Lader, 1974b) or state anxiety (Spielberger et al., 1970) questionnaires (all p > 0.2) and did not significantly modulate response accuracy or reaction times in the behavioral task (all p > 0.05). These factors therefore appear to have little explanatory value in relation to the observed differences in brain activity between the progesterone and placebo conditions. The fMRI results show that the emotion condition yielded larger responses than the control condition in the amygdala, ventral visual stream (ranging from the primary visual cortex to the fusiform gyrus), inferior fron- tal gyri, cerebellar vermis, and supplementary motor area (p < 0.05, corrected; see Table 1 and Fig. 1). Critically, progesterone increased the right amygdala response (p < 0.05, corrected, see Table 1 and Fig. 2-3). A similar effect was also present in the left amygdala, although at a less conservative threshold (see Table 1). In contrast, progesterone did not significantly influence neural activity in other brain areas.

To explore the relation between amygdala reactivity and allopregnanolone levels, state anxiety and mood across subjects, additional correlation analyses were performed. While progesterone reliably increased amy- gdala reactivity in the within-subjects comparison, no significant correlations between allopregnanolone levels and amygdala activity were observed across subjects during the progesterone or placebo conditions. However, progesterone modulated the relation between amygdala activity and mood. While amygdala activity was posi- tively correlated with the MRS subscales alertness (r = 0.66, p = 0.003) and contentedness (r = 0.53, p = 0.025) during the placebo condition, no such relation was observed during the progesterone condition (p > 0.1). The functional connectivity analysis identified various brain regions that were functionally coupled with the right amygdala. Brain regions with a positive correlation with amygdala activity included regions in the medial temporal lobe (including the left amygdala and both hippocampi), the basal ganglia, orbitofrontal cortex, temporal pole, thalamus, inferior temporal gyri, pons, and fusiform gyri. Brain regions in the midbrain and cer- ebellar vermis showed a negative correlation with amygdala activity (p < 0.05, corrected; see Table 2). Progesterone decreased functional connectivity of the fusiform gyrus with the amygdala (p < 0.05, cor- rected). Furthermore, the results indicate that progesterone increased functional connectivity of the dorsal anterior cingulate cortex with the amygdala at a less conservative threshold (p < 0.001, uncorrected; see Table 2 and Fig. 4).

124 Chapter 9 | Progesterone selectively increases amygdala reactivity in women

Table 1. Cluster maxima for brain regions with a larger response during the emotion than the control condition (main effect of task), and for brain regions that are modulated by progesterone (drug-by-task interaction).

MNI coordinates Cluster size t p1 x y z

Main effect of task R inferior occipital gyrus 26 -90 -10 3613 17.9 < 0.001 L inferior occipital gyrus -26 -92 -12 3138 14.1 < 0.001 R amygdala 22 -4 -16 125 9.0 < 0.001 L amygdala -22 -10 -16 164 8.5 < 0.001 R inferior frontal gyrus 44 10 32 575 8.4 < 0.001 L inferior frontal gyrus -44 10 28 240 8.4 < 0.001 Cerebellar vermis 2 -74 -32 63 7.0 0.002 Supplementary motor area 0 10 60 41 6.9 0.002 L inferior frontal gyrus -56 22 -4 22 6.7 0.005

Drug-by-task interaction (progesterone > placebo) R amygdala 26 -6 -22 14 5.8 0.0012 L amygdala -28 -6 -22 - 2.9 0.0053

1 P-values are corrected for multiple comparisons across the whole brain (p < 0.05, k ≥ 10). 2 P-value corrected for multiple comparisons across the amygdala (p < 0.05). 3 Peak voxel (p-value is uncorrected for multiple comparisons).

125 Chapter 9 | Progesterone selectively increases amygdala reactivity in women

Figure 1. Larger amygdala responses to salient, biologically relevant stimuli than the control condition in the placebo (A) and the progesterone condition (B). The sagittal slices on the left (x = 26) and the maximum intensity projections on the right are thresh- olded at p < 0.001, uncorrected for multiple comparisons.

Figure 2. A single progesterone administration increases the amygdala response. A coronal slice (A) shows the increased right amy- gdala reactivity (y = -6) in the progesterone condition. The maximum intensity projection (B) shows the selectivity of the effect. The figures are thresholded at p < 0.001, uncorrected for multiple comparisons. Note that the same effect was observed in the left amygdala at a less conservative statistical threshold.

126 Chapter 9 | Progesterone selectively increases amygdala reactivity in women

Table 2. Cluster maxima for brain regions that are functionally connected with the right amygdala, and for brain regions whose con- nectivity shows an interaction with progesterone.

MNI coordinates Cluster size t p1 x y z

Positive correlation MTL, basal ganglia, OFC, temporal pole, thalamus4 -24 0 -16 14025 18.8 < 0.001 R inferior temporal gyrus 62 -54 -8 98 8.6 < 0.001 R inferior occipital gyrus 40 -88 -2 82 7.6 0.001 Pons 12 -32 -42 479 7.1 0.002 L inferior temporal gyrus -52 -64 -8 39 6.3 0.012 L middle temporal gyrus -62 -46 -8 10 6.0 0.026 L fusiform gyrus -48 -64 -10 93 5.5 0.0012 R fusiform gyrus 52 -58 -10 56 5.1 0.0022

Negative correlation Midbrain -4 -26 -14 151 8.2 < 0.001 Cerebellar vermis 0 -52 -12 57 5.4 0.003

Progesterone > placebo Anterior cingulate cortex 10 24 40 9 4.7 < 0.0013

Placebo > progesterone R fusiform gyrus 42 -60 -6 10 5.5 0.0102

1 P-values are corrected for multiple comparisons across the whole brain (p < 0.05, k ≥ 10). 2 P-values corrected for multiple comparisons across the region-of-interest (p < 0.05). 3 P-value uncorrected for multiple comparisons (p < 0.001). 4 MTL: medial temporal lobe; OFC: orbitofrontal cortex

Figure 3. Mean (± SEM) amygdala reactivity in the progester- Figure 4. Progesterone modulates functional connectivity of one and placebo condition (26, -6, -22). the (A) dorsal anterior cingulate cortex (10, 24, 40) and (B) right fusiform gyrus (42, -60, -6) with the amygdala (mean ± SEM).

127 Chapter 9 | Progesterone selectively increases amygdala reactivity in women

Discussion

The results show that a single progesterone administration to healthy women in their follicular phase selectively increases amygdala reactivity. The progesterone administration significantly increased the plasma concentration of progesterone as well as that of its metabolite allopregnanolone, which potentiates the inhibi- tory actions of GABA (Majewska et al., 1986). Because it is this neurosteroid that mediates the acute effects of progesterone (Bitran et al., 1995; Reddy et al., 2005), the observed activity increase of the amygdala is likely mediated by the neuroactive metabolites allopregnanolone and/or pregnanolone. The amygdala is critically involved in a wide range of emotional behavior (Phelps and LeDoux, 2005), including fear and anxiety (Davis and Whalen, 2001a), and the amygdala displays a pathological pattern of re- sponsiveness in certain anxiety (Rauch et al., 2003) and mood disorders (Drevets, 2003). A previous animal study, in which a relatively large dose of allopregnanolone was infused into the central nucleus of the amygdala, has shown that allopregnanolone acts on the amygdala to mediate anxiolysis (Akwa et al., 1999). Although the present study does not show an anxiogenic effect of allopregnanolone, the results suggest a neural mechanism by which allopregnanolone could increase anxiety and worsen mood (i.e., by increasing amygdala activity). Progesterone decreased functional connectivity of the amygdala with the fusiform gyrus, a brain region preferentially involved in the processing of faces (Kanwisher et al., 1997; Tsao et al., 2006). The amygdala projects back to all levels of the ventral visual stream (Amaral et al., 2003), and the amygdala influences activity in this brain region during face perception (Vuilleumier et al., 2004), suggesting that progesterone decreases the back-projections of the amygdala to the fusiform gyrus. However, because functional connectivity is a corre- lational method, this result may also reflect the influence of the fusiform gyrus on the amygdala by feed-forward projections. Furthermore, although the functional integration of the emotion circuitry with the amygdala was largely unaltered by progesterone, the results indicate that progesterone increased functional coupling of the dorsal anterior cingulate gyrus (dACC) with the amygdala. The dACC is activated during the cognitive evalua- tion of threatening stimuli which reduces amygdala activity (Hariri et al., 2003), and is thought to support the cognitive control of emotion (i.e., reappraisal) (Ochsner and Gross, 2005). This suggests either a progesterone induced change in dACC regulation of amygdala activity, or conversely, the influence of the amygdala on regu- latory processes. However, this result should be interpreted with caution, because the effect did not remain significant after correction for multiple comparisons. Although progesterone consistently increased amygdala reactivity in a within-subject comparison, no sig- nificant correlations between allopregnanolone levels and amygdala activity were observed across subjects. However, serum allopregnanolone levels do not reflect the (regionally specific) availability of this and other neu- roactive metabolites in the brain, and individual differences in neurosteroid sensitivity may obscure a relation between neurosteroid levels and amygdala reactivity when assessed across individuals. Furthermore, while amy- gdala activity was positively correlated to alertness and contentedness in the placebo condition, which are two subscales of the mood rating scale that was used, no correlations with state anxiety and mood were observed in

128 Chapter 9 | Progesterone selectively increases amygdala reactivity in women

the progesterone condition. The absence of a relation between amygdala reactivity and state anxiety and mood in the progesterone condition could reflect the limitation of this study to assess across subject correlations, because the study was designed as a within-subjects investigation. In addition, the effects of progesterone on mood are likely mediated by interactions within a larger emotion circuitry (Pezawas et al., 2005). The absence of a large change in the functional integration of the emotion circuitry may therefore explain the absence of a significant change in mood or anxiety. Two related mechanisms could explain the paradoxical allopregnanolone induced amygdala activity in- crease. First, the hormonal milieu appears to determine the acute effects of the neuroactive progesterone me- tabolites. A study in female mice showed that an otherwise anxiolytic dose of pregnanolone increases anxiety to an aversive stimulus after a period of relative allopregnanolone suppression, which was accomplished by admin- istering a 5α-reductase inhibitor for three days that attenuates the metabolic transformation of progesterone to allopregnanolone (Smith et al., 2006). This approach was used to mimic the prolonged low allopregnanolone concentration during the human follicular phase (Wang et al., 1996), the menstrual cycle phase during which the women in this study were investigated. Second, allopregnanolone induces a nonlinear behavioral response (Bäckström et al., 2003; Miczek et al., 2003; N-Wihlbäck et al., 2006). Low doses of allopregnanolone increase aggressive behavior in mice, whereas higher doses decrease aggression (Fish et al., 2001). Moreover, the addition of a low dose of progesterone to the treatment with estradiol during hormonal replacement therapy increases negative mood, but higher doses do not (Andréen et al., 2006). These negative mood symptoms are related to allopregnanolone levels in a bimodal fashion, as these symptoms are only present at moderate, but not low or high allopregnanolone concentrations (Andréen et al., 2005; Andréen et al., 2006). Furthermore, allopregnanolone levels are lower in PMS patients that respond to antidepressant or placebo treatment than non-responders (Freeman et al., 2002). In the present study, an increased responsiveness of the amygdala was observed after increasing the progesterone and al- lopregnanolone concentrations from the follicular phase to levels that are naturally occurring during the luteal phase and early pregnancy, respectively (Parizek et al., 2005; Paoletti et al., 2006). The increased amygdala response might therefore be specific for this concentration change, while higher supra-physiological concen- trations might eventually decrease amygdala activity (Akwa et al., 1999), in parallel with increased sedation in humans (Timby et al., 2006). These results therefore show that changes in progesterone level that are within the physiological range can increase amygdala activity. These two mechanisms are not mutually exclusive but may interact, because the probable progesterone dose-dependent amygdala response is likely modulated by the hormonal milieu. In this context, it is important to note that during natural hormone fluctuations during the menstrual cycle, high progesterone levels occur only in the presence of increased estradiol levels. Thus, the threshold for amygdala reactivity likely changes over the menstrual cycle. This suggestion is supported by studies showing that the response of healthy women to a preg- nanolone dose is larger in the luteal than follicular phase (Sundström et al., 1998). Furthermore, women with PMS experience more severe symptoms during cycles with high luteal phase progesterone and estradiol levels

129 Chapter 9 | Progesterone selectively increases amygdala reactivity in women

(Hammarbäck et al., 1989), and higher doses increase negative mood symptoms during the progestin phase of hormonal replacement therapy (Björn et al., 2003). The mechanism by which allopregnanolone could increase amygdala activity is not well-understood. How- ever, the paradoxical allopregnanolone induced amygdala activity increase could reflect the disinhibition of the principal neurons of the amygdala via inhibition of the inhibitory interneurons (N-Wihlbäck et al., 2006), a mechanism that has been demonstrated for the disinhibitory actions of dopamine (Marowsky et al., 2005). Neurosteroid action is brain region and neuron specific, and neurons of the central nucleus of the amygdala appear relatively insensitive to allopregnanolone (Belelli et al., 2006). This heterogeneity in neurosteroid sensi- tivity could be mediated by regional differences in receptor subunit composition, phosphorylation, and steroid metabolism (Belelli et al., 2006). Specifically, allopregnanolone inhibitsα 4β2δ GABAA-receptors, which reduces tonic inhibition and thereby increases excitability. Because this receptor is highly sensitive to neurosteroids, inhibition of this receptor at relatively low neurosteroid concentrations can induce paradoxical effects (Shen et al., 2007), and may contribute to amygdala disinhibition. The amygdala has many (largely reciprocal) connections to brain areas that are involved in sensory processing, visceral functioning, emotional behavior, and mood (Price, 2003). The progesterone induced amy- gdala activity increase could therefore affect the processing in many other brain regions relevant for mood regulation. Studies with depressed patients suggest that negative mood is generated by increased activity of brain structures that support the identification of the emotional significance of stimuli and the generation of an affective response (e.g., the amygdala, insula, ventral striatum, and ventral prefrontal cortex), and decreased activity of brain structures supporting emotion regulation (e.g., the dorsal prefrontal cortex and hippocampus) (Phillips et al., 2003b). The menstrual cycle has been shown to modulate activity in several of these brain regions (Fernández et al., 2003; Goldstein et al., 2005; Protopopescu et al., 2005; Amin et al., 2006), suggesting that progesterone and estradiol may influence mood by modulating this emotion circuitry. The results of the present study show that progesterone increases amygdala activity and indicate that it increases functional coupling with the dACC, without affecting the response amplitude in other brain regions. The results of our study therefore suggest that progesterone and/or allopregnanolone modulates amygdala activity, and may thereby change the functionality of a larger emotion circuitry. The amygdala is involved in a wide range of emotional behavior, including social cognition (Adolphs, 2006) and the stress response (Herman et al., 2005), which appear to be modulated by progesterone. For example, the modulation of the preference for faces during the menstrual cycle appears to be mediated by progesterone (Penton-Voak et al., 1999; Jones et al., 2005), as well as the enhancement of the hypothalamus-pituitary-adrenal (HPA) axis response to stressors during the luteal phase (Kirschbaum et al., 1999; Roca et al., 2003). Our results suggest that progesterone could modulate these processes by increasing the responsiveness of the amygdala. The present study shows that allopregnanolone can have paradoxical effects on amygdala activity in healthy women, without influencing anxiety or mood. However, this mechanism could mediate the negative effects of allopregnanolone on anxiety and mood in susceptible women. Whereas the administration of progesterone or

130 Chapter 9 | Progesterone selectively increases amygdala reactivity in women

estradiol after gonadal hormone suppression does not produce mood changes in healthy women, it produces negative mood symptoms in women with PMS (Schmidt et al., 1998). Furthermore, PMS patients have a differ- ent sensitivity to pregnanolone during the high progesterone luteal phase (Sundström et al., 1998), suggesting that allopregnanolone could affect amygdala activity differently in patients with PMDD/PMS. Because the amy- gdala is pathologically activated in certain anxiety (Rauch et al., 2003) and mood disorders (Drevets, 2003), we speculate that allopregnanolone might increase amygdala activity to a larger extent in women with PMDD/PMS than in healthy women as observed in this study, which could explain part of the negative mood symptoms in PMDD/PMS and possibly other gonadal hormone related mood disorders.

Acknowledgements

This work was supported by an internal grant from Radboud University Nijmegen Medical Center, an EU structural fund objective 1 program, and the Swedish Research Council project 11198.

131 Chapter 9 | Progesterone selectively increases amygdala reactivity in women

chapter 10 | discussion

132 chapter 10 | discussion chapter 10 | Discussion

In this chapter the results of the studies described in the previous chapters are discussed and the ap- plied methodology considered. At the end directions for further research are given. The primary aim was to investigate the GABAergic effects of allopregnanolone in men and women. Secondary aims were to investi- gate the influence of the menstrual cycle on these effects in women and to test the hypothesis that in acute stress plasma allopregnanolone concentrations increase in men and women together with an increase in PBR density.

GABAergic effects of allopregnanolone in men and women

We investigated the GABAergic effects of allopregnanolone in humans by measuring SEV, sedation, EEG frequency spectra (in men), and amygdala activation, after intravenous administration of allopregnanolone (chapter 6) and oral administration of progesterone (chapters 7 and 9). The results show that allopregnanolone decreases SEV (chapters 6 and 7), increases self-ratings of sedation (chapters 6 and 7), and increases sigma and beta power on EEG frequency spectra recordings (chapter 6). These findings are in line with earlier findings obtained in animal research and indicate that allopregnanolone exerts a GABAergic effect in humans. Allopregnanolone, however, increased amygdala activation during emotional processing of angry and fearful face stimuli in women (chapter 9) and this is in contrast with the decreased amygdala activity found after lorazepam (Paulus et al., 2005). In this latter study, however, the subjects consisted of both men and women. Further, no data was given about hormonal contraceptive use or the phase of their cycle the women were tested. The effect of neuroactive progesterone metabolites depends on the hormonal milieu of the subject, i.e. the concentrations of endogenous and exogenous progesterone and estradiol. Pregnanolone has, for example, an anxiogenic effect after a period of suppression of allopregnanolone (Smith et al., 2006). Low allopregnanolone concentrations are also present in the follicular phase, the time in which the women in our study were tested for emotional processing. Another explanation for the activation of the amygdala in our study is the fact that positive GABA(A) receptor modulators like allopregnanolone, benzodiazepines, and alcohol, show paradoxical effects consisting of anxiety, depression, irritability, or aggressiveness, at low dose, and anxiolysis and sedation at higher doses (Beauchamp et al., 2000; Fish et al., 2001; Miczek et al., 2003). These paradoxical effects are observed in 3-10% of the people (similar but more moderate effects are observed in 20-30% of the people) (Backstrom et al., 2008a). Instead of speaking of a paradoxical effect, it is more precise to speak of a bimodal effect of GABA(A) receptor positive modulators in certain individuals (Andreen et al., 2006a). In support of the view of this bimodal effect of allopregnanolone is the fact that we have found a decreased amygdala response

135 Chapter 10 | Discussion

during higher allopregnanolone concentrations in an other study (van Wingen et al., 2007). This latter study is not part of the present thesis. A third explanation for the demonstrated increased amygdala activity in our study might be found in the subunit composition of the GABA(A) receptor in our subjects. GABA(A) receptors consist- ing of subunits that cause increased instead of decreased activity in brain regions have been shown. They re- duce tonic inhibition of hippocampal pyramidal cells when activated by allopregnanolone resulting in increased anxiety in pubertal female mice (Shen et al., 2007b). Although our subjects did not show increased anxiety on the clinical level following progesterone administration, they were young and increased neural excitability of the amygdala might be present due to this subunit compostion without increased anxiety. Not only age influences subunit composition, also exposure to allopregnanolone itself changes subunit composition of GABA(A) recep- tors and consenquently may modulate its functional activity and resulting behaviour of the subject (Follesa et al., 2004). A limitation of the intravenous allopregnanolone challenge (chapter 6) is the fact that this study was not placebo-controlled and that the women used oral contraceptives. On the other hand, the use of oral contracep- tives might have made the group of women more homogeneous. Hormonal contraceptives decrease concentra- tions of endogenous neuroactive progesterone metabolites and thus limit the influence of cyclic variation in the endogenous production of allopregnanolone and synchronise GABA(A) receptor subunit composition (Follesa et al., 2002; Rapkin et al., 2006). Although women showed a larger response on SEV to the allopregnanolone injections than men, they showed less response on subjective ratings of sedation. Subjective ratings of sedation were only significantly increased after the last and highest dosage of allopregnanolone and this is in accordance with the pattern that was shown earlier in women without oral contraceptives (Timby et al., 2006). It could be suggested that women show tolerance to some effects of progesterone leading to lower scores on subjective ratings of sedation, inde- pendently of the fact if they use or not use oral contraceptives.

The influence of the menstrual cycle on effects of allopregnanolone in women

A secondary aim of this thesis was to investigate the influence of the menstrual cycle on the sensitivity to progesterone in women. Healthy women show increased sensitivity to GABAergic substances like pregna- nolone and diazepam in the luteal phase while women with PMDD show decreased sensitivity to pregnanolone, diazepam and alcohol in the luteal phase. No data exist, however, on the influence of the cycle on the sensitivity to progesterone in both groups. Therefore, we compared the effects of oral progesterone in healthy premeno- pausal women between their follicular and luteal phase (chapter 8). There appeared to be an effect of the phase of the cycle that consisted of an increased sensitivity to progesterone in the luteal phase as compared to the follicular phase. This increased sensitivity was shown by an enhanced SEV and sedation response to adminis- tration of progesterone. We concluded that the effect of oral progesterone on GABA(A) receptor sensitivity as

136 Chapter 10 | Discussion

measured by SEV and subjective ratings of sedation is larger in the luteal then in the follicular phase and that this effect is probably due to enhanced GABA activity by the progesterone metabolite allopregnanolone. Future studies should examine whether women with PMDD show decreased sensitivity to oral progesterone in the luteal phase. A limitation of the study is the small number of subjects that were included. This was due to problems with the determination of the luteal phase. We applied hormone concentrations tests in urine to detect the start of the luteal phase. The test consisted of the measurement of luteinizing hormone (LH) in urine. LH shows an increase in concentration about 24 hours before ovulation. Women received test kits and performed the test at home. When the test was positive they made an appointment for the study day. Plasma concentrations of progesterone were drawn at the study day to confirm luteal phase. It turned out that not all women with assumed luteal phase on the basis of the positive test had high enough progesterone concentrations that corresponded with luteal phase levels. These women were left out of the statistical analysis.

Relationship between plasma allopregnanolone concentrations and PBR density following acute stress in men and women

Our hypothesis that plasma allopregnanolone concentrations and PBR density in humans will increase in acute stress was confirmed (chapter 5). This finding is in line with previous results from animal studies. There was also a small but significant correlation between the plasma allopregnanolone concentration and PBR den- sity. We also assumed that PhD examination would lead to stress and measured clinical signs that could prove this assumption. Plasma cortisol concentration, blood pressure, and heart frequency increased significantly and therefore it can be concluded that PhD examination can be used as a model for stress in humans.

Methodological considerations

In this thesis measurement of the peak velocity of saccadic eye movements and EEG has been used as effect parameters for GABA function. The outcome has been evaluated against the plasma concentrations. This gives a good insight into the GABAergic effect of allopregnanolone. One has to consider, however, that plasma levels reflect brain concentrations only to a certain extent. Lumbar punctures to collect cerebral liquor may overcome this shortcoming partially. Concentration in the brain differs substantially between regions both in humans and in animals (Wang et al., 1995; Bixo et al., 1997). Neither plasma nor cerebral liquor can adequately mirror the tissue concentration in brain regions like the amygdala or hippocampus. In this thesis lumbar punc- ture was not performed because of the extra burden for the subject.

137 Chapter 10 | Discussion

In addition to the bioavailability of allopregnanolone another relevant parameter is the receptor sensitivity and density. While SEV and EEG as used in this thesis reflect the sensitivity of the receptor, positron emission tomography (PET) gives information about the receptor density. In essence, in a PET study a radioactive tracer is administered intravenously and the level of the tracer in the brain is measured thereafter. Differences in GABA(A) receptor density between phases and between healthy and not-healthy women could reveal possible pathophysiological mechanisms. A PET study in women with PMDD revealed differences in binding of the sero- tonergic receptor between the phases (Jovanovic et al., 2006). So far, no PET study has been performed of the GABA(A) receptor in PMDD. Reaction time was measured to give an indication of attention and concentration. The Mood Rating Scale was administered as a subjective rating of the allopregnanolone effect. As method to establish phase of the cy- cle of the women, we applied hormone concentration tests in urine and plasma. Ultrasonosgraphy of the ovaries is an alternative but more inconvenient for the subject. The procedure around the intravenous allopregnanolone challenge (chapter 6) deserves some more atten- tion. We started the PhD project with the idea of administering allopregnanolone as the best way to investigate its role in psychiatric disorders. As allopregnanolone itself had not been administered to humans, effects of allo- pregnanolone in man were only retrievable from studies in which its precursor progesterone had been adminis- tered to humans, regardless of the purpose of the study (Arafat et al., 1988; Freeman et al., 1993; de Lignieres et al., 1995; Vanselow et al., 1996). None of these studies used neurophysiological assessments as, for example, the saccadic eye movements as a measure for sedation. However, this measure is more reliable in dose-effect stud- ies (van Steveninck et al., 1999). Although a lot of research was done with direct infusion of allopregnanolone in animals in vivo, we quickly came to the conclusion that there was no pharmaceutical solution for human use available. Therefore, we inventorised the possibilities to make a solution for human use. As we came across the literature about the development of the isomer of allopregnanolone, pregnanolone, to an anesthetic, we became aware of the difficulties of finding a solvent that had no adverse effects. The development of pregnanolone to an anesthetic had been stopped because of adverse effects to the solvent Cremophor. Pregnanolone (3alpha- OH-5beta-ol-20-one) and allopregnanolone (3alpha-OH-5alpha-ol-20-one) are each others cis- and transisomers determined by the alpha or beta position of the hydrogen atom bound to the fifth carbon (Morris et al., 1999). Both are 3alpha-reduced metabolites of progesterone. Both can be formed peripherally but production within the brain (via de novo production of progesterone or by metabolization of blood-borne progesterone) is only established for allopregnanolone. This is due to the fact that the 5alpha-reductase (both isoenzymes, type I and II) is found to be present in the brain but no research has been done on the presence of the 5beta-reductase. Be- cause allopregnanolone seemed to play an important role in psychiatric disorders more than pregnanolone, our aim was still to elicit the effects of allopregnanolone. The solvent 2-hydroxypropyl-beta-cyclodextrin instead of Cremophor was appropriate in our opinion to dissolve allopregnanolone because it had been standardly used for many years to dissolve the itraconazol for intravenous use in human. With cyclodextrin as proposed solvent for allopregnanolone, we made contact with the Departments of Clinical Pharmacy and Pharmacology

138 Chapter 10 | Discussion

and Toxicology of the Radboud University Nijmegen Medical Centre. It became clear then, however, that it would take a lot of time to make such a solution (laboratory facilities had to be set up to be able to assess concentra- tion, stability, purity of the solution, and so on). We then turned to the group of professor Bäckström in Sweden which had administered a solution of pregnanolone in albumin to humans before. The knowledge of how to dis- solve pregnanolone in albumin could be used for the preparation of an allopregnanolone solution. Although it was of course our wish that the solution would be prepared at the Clinical Pharmacy Department of our hospital, this was not possible due to the need of special equipment. Therefore we agreed the solution to be prepared at the Clinical Pharmacy Department of the Univerity Hospital in Umea. There the solution was frozen and at the same day sent to us. As this was a first use in human study, special conditions had to be fulfilled. Permission of the Netherlands Health Care Inspectorate had to be obtained. A pharmaceutical report showing stability of the frozen allopregnanolone-albumin solution after thawing after different periods of time was demanded by the Clinical Pharmacy Department of the Radboud University Nijmegen Medical Centre, attention had to be paid to the safety of the subjects, the ethics committee had to be convinced of the need of the callenge, and so on. The important role of allopregnanolone in several psychiatric disorders and the possibility to use allopregnanolone as a drug, convinced the ethics committee of the need to further investigate this neuroactive steroid in humans. Later, allopregnanolone challenges in humans were also performed in Sweden (Timby et al., 2006; Kask et al., 2008). Stability tests of allopregnanolone before and after thawing were performed and the results were good. As for the clinical safety of the subjects, the Department of Anaesthesiology was informed about our project. During study days there was always an anaesthesiologist standby. Accidents with bottles of allopregnanolone solution that were broken during transportation from Sweden at the first day of testing had to be overcome. Volunteers had to be found and had to be reliable as study days were neatly planned.

Directions for further research

Plasma levels and even cerebro spinal fluid levels may not reflect concentrations in specific brain regions that are the subject of the current study. Neuroactive steroids can easily cross the blood brain barrier, but this does not exclude the possibility of local alterations in brain concentrations. Therefore, further research is needed to develop highly sensitive and specific technologies of neuroactive steroid determination, such as magnetic resonance spectroscopy (MRS). With MRS, for example, the GABA concentration in the brain has been studied in PMDD patients and controls and it was shown that PMDD patients have lower concentrations in the follicular phase than the controls (Epperson et al., 2002). This supports the suggestion that in PMDD there is a dysfunction of the GABA system. PET studies of the GABA(A) receptor could reveal differences in binding pro- files between phases and between subjects suffering from PMDD and controls.

139 Chapter 10 | Discussion

Neuroactive steroids as biomarkers

In future, it could be suggested that neuroactive steroids could serve as biomarkers in the differential diagnosis of psychiatric disorders, in the prediction of risk of relapse after remission, and as a risk factor for psychiatric disorders. In depression it has been shown that concentrations of allopregnanolone are decreased (Romeo et al., 1998; Strohle et al., 1999), and in panic disorder they are increased (Strohle et al., 2002). But more research is needed before neuroactive steroids can reliably be used in clinical psychiatry as biomarkers: pro- spective studies containing a great number of participants in which plasma values before, at the start, during, and after remission of a psychiatric disorder are assessed, together with studies that establish reference values in healthy subjects.

Neuroactive steroids and alcohol

High alcohol intake increases allopregnanolone plasma concentrations in men and women (Torres and Ortega, 2003; 2004) while low alcohol intake of 0.2 g/kg (one glas of wine) decreases the allopregnanolone concentration in women (Nyberg et al., 2005). In animals it has been shown that alcohol also increases allo- pregnanolone brain content (Khisti et al., 2003). It is suggested that allopregnanolone mediates the anxiolytic, rewarding and other behavioural effects of alcohol in humans (Pierucci-Lagha et al., 2006). These findings are derived mostly from acute alcohol exposure. More chronical intake of alcohol does not change allopregnanolone plasma concentrations, but brain levels, however, could be increased (Holdstock et al., 2006). Alcohol induced changes in allopregnanolone brain levels are time and region dependent (Khisti et al., 2004). It is suggested that allopregnanolone increases due to alcohol induced increased HPA-axis activity as adrenolectomy prevents allopregnanolone increase in rats (Khisti et al., 2002). In our allopregnanolone challenge (chapter 6) subjects reported alcohol-like feelings like drowsiness, sleepiness and a funny mood. Based on the findings described above it seems likely that new drugs can be developed to treat alcoholism, e.g. drugs that inhibit the formation of allopregnanolone or that antagonise its activity at the GABA(A) receptor, thus preventing the desired effects of alcohol. When intending to stop taking alcohol, people may find benefit from allopregnanolone suppletion. It has been demonstrated, namely, that during the first days of withdrawal from alcohol, allopregnanolone plasma levels are decreased while scores of depression and anxiety are increased (Romeo et al., 1996). A few days later, levels are normalised.

140 Chapter 10 | Discussion

The influence of neuroactive steroids on benzodiazepine dependence

As both benzodiazepine and allopregnanolone positively modulate the GABA(A) receptor, and the majority of benzodiazepine users are women (Herings, 1994), information about the influence of progesterone on the sensitivity to benzodiazepines is needed. A few studies have addressed this issue (McAuley et al., 1995; de Wit and Rukstalis, 1997; Kroboth and McAuley, 1997; McAuley and Friedman, 1999; Rukstalis and de Wit, 1999). No difference has been observed in sensitivity to orally administered , a benzodiazepine, between the follicular (low serum level of progesterone) and luteal (high serum level of progesterone) phase of the menstrual cycle in premenopausal women (McAuley and Friedman, 1999). In this study, sensitivity has been measured us- ing neuropsychological tests and nurse-rated sedation. This lack of influence of cyclical changes in endogenous progesterone applies also to triazolam, another benzodiazepine, when administered in the follicular, periovula- tory, and luteal phase (de Wit and Rukstalis, 1997; Rukstalis and de Wit, 1999). The pharmacodynamic effects of triazolam across these phases were also measured by means of neuropsychological tests. The sensitivity of postmenopausal women to intravenously administered triazolam, however, is enhanced when 300 mg of micro- nized progesterone is orally administered 2.5 hours earlier (McAuley et al., 1995). Also in this study effect was measured using neuropsychological tests and nurse-rated sedation. This discrepancy in effect on sensitivity to benzodiazepines between administration of progesterone and fluctuating concentrations of endogenous pro- gesterone across the cycle (de Wit and Rukstalis, 1997; McAuley and Friedman, 1999; Rukstalis and de Wit, 1999), can be explained by the much higher progesterone levels that are achieved following progesterone administra- tion (maximum concentrations up to 626 ng/ml versus 19 ng/ml and 20 ng/ml during normal cycles). A limitation of all studies is that the pharmacodynamic effects have been investigated by subjective self-report measures. These measures are sensitive to other, non-pharmacological, influences. It might be that by using neurophysi- ological measurements, subtle differences in sensitivity to benzodiazepines induced by fluctuating endogenous progesterone levels are detected. Such a neurophysiological measurement, and one that is highly sensitive to the effects of centrally acting drugs, is the measurement of saccadic eye movements (van Steveninck et al., 1999). An enhancing effect of progesterone on the sensitivity to benzodiazepines is likely because progesterone and its metabolites have been shown to facilitate benzodiazepine binding to the GABA(A) receptor (Prince and Simmonds, 1993)

The therapeutic potential of neuroactive steroids

Neuroactive steroids could have therapeutic potential. Development of synthetic derivates of endogenous neuroactive steroids that act as positive or negative modulators of the GABA(A) receptor or that have antago- nistic effects to neuroactive steroids that exert, for example, negative mood effects, may add new psychophar- macological possibilities. Another way by which neuroactive steroids could be part of new treatment modalities

141 Chapter 10 | Discussion

is by modulation of neuroactive steroid synthesis. In this way the equilibrium of endogenous neuroactive ster- oids can be influenced. Modulation of synthesis could happen, for example, by development of new substances that bind to the peripheral benzodiazepine receptor (Kita et al., 2004). When endogenous steroids are to be developed into new drugs, then the pharmacokinetic properties have to be payed special attention to. Steroids are known to be poorly solubilized in plasma, have a short half-life and are metabolized into other substances. Neuroactive steroids may turn out to be the mediating key factor of the clinical efficacy of some psychotropic drugs and may give new insight in the psychopharmacological treatment of psychiatric disorders (Locchi et al., 2008)

PMDD and negative mood on oral contraceptives

In PMDD allopregnanolone is suggested to contribute to the negative mood, probably due to an altered sensitivity of the GABA(A) receptor in these women. Development of an antagonist to allopregnanolone for these patients is rational (Backstrom et al., 2008b). The 3beta variant of allopregnanolone has already been shown to inhibit the impairment in learning and memory induced by allopregnanolone (Turkmen et al., 2004). The same may apply for women who experience negative mood complaints when using oral contraceptives. This is a relevant clinical problem. Plasma concentrations of neuroactive steroids are decreased in women taking oral contraceptives (Paoletti et al., 2004; Kurshan and Neill Epperson, 2006), as well as in brain of mice treated with oral contraceptives (Follesa et al., 2002), and can induce tolerance to GABA(A) receptor substances leading to decreased GABA(A) receptor sensitivity (Rapkin et al., 2006). Further, in animal studies oral contraceptives can change GABA(A) receptor subunit gene expression (Follesa et al., 2002). Low endogenous concentrations of allopregnanolone (bimodal effect) or a shift of the GABA(A) receptor sensitivity to inverse agonism, may finally lead to negative mood and irritability in certain women who use oral contraceptives.

Antipsychotics

Post-mortem studies have shown that allopregnanolone levels are decreased in parietal cortex in subjects with schizophrenia compared to control subjects adding further support for the hypothesis that neuroactive steroids are involved in psychiatric disorders (Marx et al., 2006). Allopregnanolone induces catalepsy like an- tidopaminergic agents as haloperidol do and can be reversed by dopamine agonists and GABA(A) receptor antagonists (Khisti et al., 1998; Mandhane et al., 1999). In an animal model of schizophrenia with, among other agents, apomorphine, the allopregnanolone-induced catalepsy was blocked by prior treatment with the GABA(A) receptor antagonist picrotoxin (Khisti et al., 2002a). This neuroactive-steroid-induced catalepsy effect is been attributed to the enhanced GABAergic tone. Selective serotonin reuptake inhibitors like fluoxetine potentiate the catalepsy effect of allopregnanolone probably by enhancing brain content of allopregnanolone and this may contribute to the atypical profile of newer antipsychotic drugs (Deshpande et al., 2001). And although it had

142 Chapter 10 | Discussion

been shown that the antipsychotics olanzapine (Marx et al., 2000a; Marx et al., 2003) and clozapine (Barbaccia et al., 2001; Marx et al., 2003) increase allopregnanolone brain content, until now only one study investigated the role of allopregnanolone in mediating the antipsychotic effect of olanzapine (Ugale et al., 2004). The results from this study suggest that the antipsychotic effect of olanzapine is depended on allopregnanolone induced enhanced GABAergic transmission. That the effect of olanzapine is mediated by allopregnanolone can also be suggested from another perspective. It has been shown that allopregnanolone induces hyperphagia (Chen et al., 1996). Accordingly, obese patients have higher plasma allopregnanolone levels than controls and react with faster allopregnanolone increase upon administration of CRH (Menozzi et al., 2002). Taken these findings together, it is not surprising that olanzapine, which increases allopregnanolone brain content, induces weight gain, more than most other atypical antipsychotics. A fact which is commonly known by psychiatrists and leads to poor drug compliance in patients. The allopregnanolone increasing effects of olanzapine and clozapine were shown in animals. In humans, clozapine did not lead to increased allopregnanolone levels despite good clinical effect (Monteleone et al., 2004). This could, however, be due to the time of blood collection, i.e. the time point of acute increased plasma concentration was missed. Apart from the antipsychotic effect of olanzapine, its anxiolytic effect is also suggested to be mediated by allopregnanolone (Locchi et al., 2008). Another interesting feature from allopregnanolone is that it can reduce orofacial dyskinesia, an important side effect of antipsychot- ics (Bishnoi et al., 2008). The underlying mechanism that is responsible for this beneficial effect is the allopreg- nanolone induced increased GABA(A) receptor activity in the basal ganglia.

Antidepressants

Women are twice as likely to suffer from major depression than men and onset or exacerbation of depres- sive episodes are more frequent during periods of hormonal fluctuations such as puberthy, menstrual cycles, the postpartum period and menopause. It can thus be suggested that progesterone and allopregnanolone play an important role in mood disorders in women. As allopregnanolone has been shown to have antidepressive effects in an animal model of depression and allopregnanolone are decraesed during depression, allopregna- nolone could be developed as a new antidepressant. , which is a slightly modulated form of the allopregnanolone molecule, is already under investigation in clinical trials in patients with epilepsy and shows antidepressive effects on the neurophysiological level (Robichaud and Debonnel, 2006). It increases, like allo- pregnanolone, the firing rate of serotonergic neurons and prevents the initial decrease of the firing rate induced SSRI’s. SSRI’s increase the serotonin content in the synapse but in the beginning of the treatment the 5HT1A autoreceptors are activated and the firing rate is decreased. Later desensitisation develops and firing rate is increased again. Adjuvant administration of allopregnanolone or ganaxolone could speed up the recovery proc- ess of depressive patients as the antidepressive effect of antidepressants alone takes 4-6 weeks.

143 Chapter 10 | Discussion

The role of neuroactive steroids in side effects of drugs

Neuroactive steroids could also play a role in the understanding of side effects of drugs. Finasteride, for ex- ample, has been linked to increased frequency of epileptic insults due to its blockage of the forming of allopreg- nanolone (Herzog and Frye, 2003) and depressive symptoms have been noticed (Ciotta et al., 1995; Altomare and Capella, 2002; Rahimi-Ardabili et al., 2006). Non-steroidal anti-inflammatory drugs (NSAIDs) also prohibit allopregnanolone synthesis by blocking an- other enzyme than finasteride does, namely the 3alpha-HSOR. It has been shown that daily use ofNSAIDs in elderly people lowers the prevalence of Alzheimer’s disease (in ‘t Veld et al., 2002). Explanations for this association include anti-inflammatory effects and suppression of the formation of amyloid. It could, however, also be suggested that by blocking the negative effect of allopregnanolone on learning the risk of dementia is diminished (Johansson et al., 2002; Brinton and Wang, 2006). The risk of inducing depression and suicide by lowering serum cholesterol by means of drug treatment, has been subject of debate for many years and is still going on (Manfredini et al., 2000). Although the effect seems to be clinically not significant overall, attention is still needed. Research findings focus on group out- come but the individual patient may react differently. All neuroactive steroids arise from cholesterol and drugs that lower the serum level of cholesterol, such as the frequently prescribed statins, will also lower neuroactive steroid concentrations. Bearing in mind that allopregnanolone is decreased in depressive patients (Romeo et al., 1998) and that allopregnanolone has antidepressive effects in an animal model of depression (Khisti et al., 2000), patients using statin with a vulnerability for depression, emerging for example from a positive psychiatric history, should be monitored carefully. It is remarkable that from the huge amount of studies on the investiga- tion of neurosteroidogenesis, in which all kinds of compounds are used to block or enhance biosynthesis, only one study used a statin to inhibit neuroactive steroid biosynthesis (Guarneri et al., 1994). And although one study assessed changes in hormonal profile due to cholesterol-lowering drug treatment, only parent hormones like testosterone, cortisol and pregnenolone were measured and not the neuractive metabolites (Ormiston et al., 2004). In a large case-control analysis a lower number of depression was found in the statin using group compared with the high cholesterol serum group that not used statins (Yang et al., 2003). However, the study was sponsored by many pharmaceutical companies and as it concerned a case-control study, no causal relation- ship can be drawn and prospective studies are still needed. A prospective study that showed decreased rates of depression and anxiety in statin using patients involved patients with coronary artery disease who besides the use of statins also received health advices probably resulting in more health consciousness and improved quality of life (Young-Xu et al., 2003). The fact that the antidepressive-like effect of the statins was independent of the cholesterol reduction, support this suggestion. Another limitation of the study was the fact that researchers had not been blinded for treatment modality. In another, randomized double-blind placebo-controlled, prospective study in coronary artery disease patients, pravastatin did not induce more depression or anxiety (Stewart et al., 2000). , however, resulted in a statistically significant increase in depression scores in a randomized

144 Chapter 10 | Discussion

double-blind placebo-controlled crossover trial in 120 men with hypercholesterolemia (Hyyppa et al., 2003). In these studies, however, depression was rated with questionnaires and no correlations were calculated between depressive symptoms and steroid concentrations. A prospective study consisting of defined groups of patients prescribed statins for different indications together with a healthy control group should be performed and psy- chiatric evalation by a psychiatrist should be added to assessment by questionnaires. Serum profile of neuroac- tive steroids should be correlated with outcome measures of depression and anxiety and cognitive functions. Cognitive functions should also be part of the outcome measures. Statins are namely increasingly prescribed in patients suffering from dementia while possible deteriorating effects due to changes in neuroactive steroid brain concentrations on already weak cognitive functions have not yet been definitively ruled out (Xiong et al., 2005).

Conclusion

Neuroactive steroids are strong candidates for being developed as new psychopharmacotropic drugs (Red- dy, 2002; Uzunova et al., 2006). With the exact binding site on the GABA(A) receptor discovered (Hosie et al., 2006), focused pharmacological research may eventually lead to a new class of drugs. Depending on the type of the disease, the involved neuroactive steroid itself or the antagonist of it should be modulated to meet phar- macokinetic demands. Interfering in the biosynthesis of neuroactive steroids is another possibility. Allopregna- nolone can have both positive and negative effects depending on dose and presence of psychiatric disorder in the subject. It can cause memory and learning impairment and mood deterioration in women subsceptible for it but it has also sedative, anxiolytic and anticonvulsive effects with a potency higher than any known benzo- diazepine. Important to mention but outside the scope of this thesis is the promising therapeutic value of allo- pregnanolone in brain injury due to trauma or stroke. Here, allopregnanolone has been demonstrated to reduce cell loss, edema and inflammation (Brinton and Wang, 2006; Sayeed et al., 2006; VanLandingham et al., 2006; VanLandingham et al., 2007).

145 Chapter 10 | Discussion

Chapter 11 | summary

146 Chapter 11 | summary

Chapter 11 | Summary

In this thesis the effect of the neuroactive steroid allopregnanolone is investigated. Allopregnanolone is an endogenous metabolite of progesterone and a positive allosteric modulator of the GABA(A) receptor. It is called a neuroactive steroid because it modulates neuronal excitability via a membrane-bound neurotransmitter recep- tor. Most data in literature comes from animal research but in this thesis human subjects were investigated. In this chapter the results are summarized.

Chapters 1 - 3

In these chapters a general introduction to the concept of neuroactive steroids is given, as well as pharma- cological data on progesterone and allopregnanolone and an outline of the serotonin system.

Chapter 4

In this chapter the available literature on the role of neuroactive steroids in depression is reviewed. Al- lopregnanolone is a positive allosteric modulator of the GABA(A) receptor and has anxiolytic, sedative, and anticonvulsive properties. Although levels of allopregnanolone in plasma and cerebro spinal fluid are decreased in depressive patients and are normalized following effective psychopharmacological treatment, it is not clear whether this normalization is only an epiphenomenon of the psychopharmacological treatment or that it is obligatory for remission of the disease. Research in which patients are treated by non-psychopharmacological modalities, such as sleep deprivation, repetitive transcranial magnetic stimulation or electroconvulsive therapy, are therefore needed. Allopregnanolone is suggested to have antidepressive properties as shown by a short- ened time of immobilization in the forced swim test model of depression. A negative feed-back on the HPA-axis and a serotonin and noradrenalin enhancing effect in brain synapses are given as an explanation for the anti- depressive property. Selective serotonin re-uptake inhibitors enhance the allopregnanolone content in brain in animals and this mechanism is supposed to explain the immediate positive effect of these drugs seen in PMDD. The neuroactive steroid dehydroepiandrosterone (DHEA) has cognitive enhancing effects in animal studies but this effect could not be replicated in humans. DHEA has also antidepressive effects and clinical trials in humans show positive results.

149 Chapter 11 | Summary

Chapter 5

Although the increase in allopregnanolone levels after acute stress has been repeatedly shown in animals, no data exists about humans. Therefore, we applied a new model of acute stress in humans, the defending of one’s PhD thesis in front of a board, and measured plasma levels of allopregnanolone. It appeared that also in humans (n = 15) acute stress leads to an increase in allopregnanolone. In addition, the density of the PBR which is an important step in the biosynthesis of allopregnanolone, was investigated. This was done because it had been shown that PBR density increases following acute stress but no data existed on the correlation between the increased plasma levels of allopregnanolone and PBR density. In our study there was a correlation, although small. As vital signs, blood pressure and heart frequency, increased together with an established endocrinologi- cal stress parameter, plasma level of cortisol, our applied model of acute stress in humans can be assumed to be valid. The results give support to the idea of developing a specific PBR ligand that regulates the synthesis of anxiolytic neuroactive steroids.

Chapter 6

Current data from human research has been restricted to collection of blood samples and administration of the precursor of allopregnanolone, progesterone. We applied a new pharmaceutical formulation for intrave- nous administration in humans. Nine healthy men and nine healthy women were administered allopregnanolone in three, increasing, intravenous dosages. Saccadic eye velocity (SEV), subjective ratings (visual analogue scales), and electroencephalography (EEG) were used to evaluate the response to allopregnanolone. Repeated blood samples for analyses of serum allopregnanolone levels were drawn throughout the study day. Allopregna- nolone decreased SEV more in women than in men, and increased subjective ratings of sedation. Subjective rat- ings of contentedness decreased in men but increased in women. Serum levels of allopregnanolone were more increased in men compared to women. Other pharmacokinetic parameters were not different between sexes. On the EEG, beta power increased in men. In conclusion, men and women reacted differently to allopregnanolone.

Chapter 7

This study was designed to examine the effect of oral progesterone and the metabolite allopregnanolone in women. Women (n = 15) in their follicular phase received oral progesterone (400 mg) or placebo. Dependent measures included plasma levels of progesterone and allopregnanolone, SEV, subjective ratings (visual ana- logue scales), and reaction time. Progesterone decreased SEV and increased sedation. This effect is probably due to enhanced GABA activity.

150 Chapter 11 | Summary

Chapter 8

Knowledge about the degree of sensitivity to progesterone and allopregnanolone in healthy women across the cycle is important in further elucidating the pathophysiology of PMDD. Therefore, we investigated whether the effect of a single dose of progesterone on GABA(A) receptor sensitivity and mood differed between the follicular and luteal phase of the normal menstrual cycle. Women (n=9) re- ceived oral progesterone (400 mg) or placebo in their follicular and luteal phase in a double-blind, randomized, crossover design. Dependent measures were saccadic eye velocity (SEV), subjective ratings of sedation, con- tentedness, calmness (visual analogue scales), and reaction times on a simple and complex attentional task. Administration of progesterone decreased SEV and increased sedation more in the luteal phases then in the follicular phase. Complex reaction time was more increased in the luteal phase compared to the follicular phase after progesterone but this difference was not statistically significant. The effects are probably due to enhanced GABA activity by the progesterone metabolite allopregnanolone. In conclusion: the effect of oral progesterone on GABA(A) receptor sensitivity as measured by SEV is dependent on the phase of the menstrual cycle.

Chapter 9

We investigated, placebo-controlled, with functional magnetic resonance imaging (fMRI) whether a single progesterone administration to healthy young women (n = 18) in their follicular phase modulates the amygdala response to salient, biologically relevant stimuli. The oral progesterone administration (400 mg) increased the plasma concentrations of progesterone and allopregnanolone to levels that are reached during the luteal phase and early pregnancy. The imaging results show that progesterone selectively increased amygdala reactivity. It is suggested that the increased activity is due to disinhibition of inhibitory interneurons. Furthermore, functional connectivity analyses indicate that progesterone modulated functional coupling of the amygdala with distant brain regions. These results reveal a neural mechanism by which progesterone may mediate adverse effects on anxiety and mood.

151 Chapter 11 | Summary

Chapter 12 | references

152 Chapter 12 | references

Chapter 12 | References

Abadie J.M., Wright B., Correa G., Browne E.S., Porter J.R., Svec F., 1993. Effect of dehydroepiandrosterone on neurotransmitter levels and appetite regulation of the obese Zucker rat. The Obesity Research Program. Diabetes 42, 662-669. Adlercreutz H., Martin F., 1980. Biliary excretion and intestinal metabolism of progesterone and in man. J Steroid Biochem 13, 231-244. Adolphs R., 2006. How do we know the minds of others? Domain-specificity, simulation, and enactive social cognition. Brain Res 1079, 25-35. Akunne H.C., Zoski K.T., Whetzel S.Z., Cordon J.J., Brandon R.M., Roman F., Pugsley T.A., 2001. Neuropharmaco- logical profile of a selective sigma ligand, igmesine: a potential antidepressant. Neuropharmacology 41, 138-149. Akwa Y., Purdy R.H., Koob G.F., Britton K.T., 1999. The amygdala mediates the anxiolytic-like effect of the neuros- teroid allopregnanolone in rat. Behav. Brain. Res. 106, 119-125. Aldred S., Waring R.H., 1999. Localisation of dehydroepiandrosterone sulphotransferase in adult rat brain. Brain Res. Bull. 48, 291-296. Altemus M., Redwine L.S., Leong Y.M., Frye C.A., Porges S.W., Carter C.S., 2001. Responses to laboratory psycho- social stress in postpartum women. Psychosom. Med. 63, 814-821. Amaral D.G., Behniea H., Kelly J.L., 2003. Topographic organization of projections from the amygdala to the visual cortex in the macaque monkey. Neuroscience 118, 1099-1120. American Psychiatric Association, 1987, 3rd edition-revised. Diagnostic and statistical manual of mental disor- ders. American Psychiatric Association, Washington, D.C. American Psychiatric Association, 1994, 4th edition. Diagnostic and statistical manual of mental disorders. American Psychiatric Press, Washington, D.C. Amin Z., Epperson C.N., Constable R.T., Canli T., 2006. Effects of estrogen variation on neural correlates of emo- tional response inhibition. Neuroimage 32, 457-464. Andreen L., Spigset O., Andersson A., Nyberg S., Backstrom T., 2006a. Pharmacokinetics of progesterone and its metabolites allopregnanolone and pregnanolone after oral administration of low-dose progesterone. Maturitas 54, 238-244. Andréen L., Sundström-Poromaa I., Bixo M., Andersson A., Nyberg S., Bäckström T., 2005. Relationship between allopregnanolone and negative mood in postmenopausal women taking sequential hormone replace- ment therapy with vaginal progesterone. Psychoneuroendocrinology 30, 212-224.

155 Chapter 12 | References

Andréen L., Sundstrom-Poromaa I., Bixo M., Nyberg S., Backstrom T., 2006b. Allopregnanolone concentration and mood - a bimodal association in postmenopausal women treated with oral progesterone. Psychop- harmacology (Berl) 187, 209-221. Andréen L., Sundström-Poromaa I., Bixo M., Nyberg S., Bäckström T., 2006. Allopregnanolone concentration and mood-a bimodal association in postmenopausal women treated with oral progesterone. Psychopharma- cology (Berl) 187, 209-221. Angold A., Costello E.J., Erkanli A., Worthman C.M., 1999. Pubertal changes in hormone levels and depression in girls. Psychol. Med. 29, 1043-1053. Ansseau M., von Frenckell R., Cerfontaine J.L., Papart P., 1991. Pilot study of PK 11195, a selective ligand for the peripheral-type benzodiazepine binding sites, in inpatients with anxious or depressive symptomatology. Pharmacopsychiatry 24, 8-12. Arafat E.S., Hargrove J.T., Maxson W.S., Desiderio D.M., Wentz A.C., Andersen R.N., 1988. Sedative and hypnotic effects of oral administration of micronized progesterone may be mediated through its metabolites. Am. J. Obstet. Gynecol. 159, 1203-1209. Association A.P., 1994, 4th. Diagnostic and statistical manual of mental disorders. American Psychiatric Associa- tion, Washington D.C. . Bäckström T., Andersson A., Andree L., Birzniece V., Bixo M., Bjorn I., Haage D., Isaksson M., Johansson I.M., Lindblad C., Lundgren P., Nyberg S., Odmark I.S., Stromberg J., Sundstrom-Poromaa I., Turkmen S., Wahl- strom G., Wang M., Wihlback A.C., Zhu D., Zingmark E., 2003. Pathogenesis in menstrual cycle-linked CNS disorders. Ann N Y Acad Sci 1007, 42-53. Bäckström T., Andreen L., Bixo M., Bjorn I., Fernandez G., Johansson I.M., Lundgren P., Lofgren L., Nyberg S., Ragagnin G., Poromaa-Sundstrom I., Stromberg J., van Broekhoven F., van Wingen G., Wang M., 2008a. Neuroactive steroids in brain and mood. In: Ritsner M.S., Weizman A. (Eds.), Neuroactive Steroids in Brain Functions, and Mental Health: New Perspectives for Research and Treatment. Springer, New York, pp. 423-433. Bäckström T., Birzniece V., Fernandez G., Johansson I.M., Kask K., Lindblad C., Lundgren P., Nyberg S., Ragagnin G., Sundstrom Poromaa I., Stromberg J., Turkmen S., Wang C., van Broekhoven F., van Wingen G., 2008b. Neuroactive steroids: effects on cognitive functions. In: Ritsner M.S., Weizman A. (Eds.), Neuroactive Steroids in Brain Functions, and Mental Health: New Perspectives for Research and Treatment. Springer, New York, pp. 103-121. Bäckström T., Sanders D., Leask R., Davidson D., Warner P., Bancroft J., 1983. Mood, sexuality, hormones, and the menstrual cycle. II. Hormone levels and their relationship to the premenstrual syndrome. Psychosom. Med. 45, 503-507. Baghai T.C., di Michele F., Schule C., Eser D., Zwanzger P., Pasini A., Romeo E., Rupprecht R., 2005. Plasma con- centrations of neuroactive steroids before and after electroconvulsive therapy in major depression. Neuropsychopharmacology 30, 1181-1186.

156 Chapter 12 | References

Barbaccia M.L., Affricano D., Purdy R.H., Maciocco E., Spiga F., Biggio G., 2001. Clozapine, but not haloperidol, increases brain concentrations of neuroactive steroids in the rat. Neuropsychopharmacology 25, 489- 497. Barbaccia M.L., Roscetti G., Trabucchi M., Purdy R.H., Mostallino M.C., Concas A., Biggio G., 1997. The effects of inhibitors of GABAergic transmission and stress on brain and plasma allopregnanolone concentrations. Br. J. Pharmacol. 120, 1582-1588. Barker E.L., Blakely R.D., 1995. Norepinephrine and and serotonin transporters: molecular targets of antidepres- sant drugs. In: Bloom F.E., Kupfer D.J. (Eds.), Psychopharmacology: the fourth generation of progress. Raven Press, New York, pp. 321-333. Barnes N.M., Sharp T., 1999. A review of central 5-HT receptors and their function. Neuropharmacology 38, 1083-1152. Barnhart K.T., Freeman E., Grisso J.A., Rader D.J., Sammel M., Kapoor S., Nestler J.E., 1999. The effect of de- hydroepiandrosterone supplementation to symptomatic perimenopausal women on serum endocrine profiles, lipid parameters, and health-related quality of life. J. Clin. Endocrinol. Metab. 84, 3896-3902. Barrett-Connor E., von Muhlen D., Laughlin G.A., Kripke A., 1999. Endogenous levels of dehydroepiandrosterone sulfate, but not other sex hormones, are associated with depressed mood in older women: the Rancho Bernardo Study. J Am Geriatr Soc 47, 685-691. Baulieu E., Robel P., 1996. Dehydroepiandrosterone and dehydroepiandrosterone sulphate as neuroactive neu- rosteroids. J. Endocrinol. 150, S221-S239. Baulieu E., Robel P., Schumacher M., 1999. Neurosteroids: from definition and biochemistry to physiopathologic function. In: Baulieu E., Robel P., Schumacher M. (Eds.), Neurosteroids: a new regulatory function in the nervous system. Humana Press, Totowa, New Jersey, pp. 1-25. Baulieu E.E., 1998. Neurosteroids: a novel function of the brain. Psychoneuroendocrinology 23, 963-987. Belelli D., Herd M.B., Mitchell E.A., Peden D.R., Vardy A.W., Gentet L., Lambert J.J., 2006. Neuroactive steroids and inhibitory neurotransmission: mechanisms of action and physiological relevance. Neuroscience 138, 821-829. Benraad T.J., Foekens J.A., 1990. Hydroxyapatite assay to measure epidermal growth factor receptor in human primary breast tumors. Ann Clin Biochem 27 ( Pt 3), 272-273. Bergeron R., Debonnel G., De Montigny C., 1993. Modification of the N-methyl-D-aspartate response by antide- pressant sigma receptor ligands. Eur. J. Pharmacol. 240, 319-323. Bermack J.E., Debonnel G., 2001. Modulation of serotonergic neurotransmission by short- and long-term treat- ments with sigma ligands. Br. J. Pharmacol. 134, 691-699. Bernardi F., Hartmann B., Casarosa E., Luisi S., Stomati M., Fadalti M., Florio P., Santuz M., Luisi M., Petraglia F., Genazzani A.R., 1998a. High levels of serum allopregnanolone in women with premature ovarian failure. Gynecol. Endocrinol. 12, 339-345.

157 Chapter 12 | References

Bernardi F., Salvestroni C., Casarosa E., Nappi R.E., Lanzone A., Luisi S., Purdy R.H., Petraglia F., Genazzani A.R., 1998b. Aging is associated with changes in allopregnanolone concentrations in brain, endocrine glands and serum in male rats. Eur. J. Endocrinol. 138, 316-321. Berr C., Lafont S., Debuire B., Dartigues J.F., Baulieu E.E., 1996. Relationships of dehydroepiandrosterone sulfate in the elderly with functional, psychological, and mental status, and short-term mortality: a French com- munity-based study. Proc. Natl. Acad. Sci. U. S. A. 93, 13410-13415. Bicikova M., Dibbelt L., Hill M., Hampl R., Starka L., 1998. Allopregnanolone in women with premenstrual syn- drome. Horm. Metab. Res. 30, 227-230. Bicikova M., Tallova J., Hill M., Krausova Z., Hampl R., 2000. Serum concentrations of some neuroactive steroids in women suffering from mixed anxiety-depressive disorder. Neurochem. Res. 25, 1623-1627. Biggio G., Barbaccia M.L., Follesa P., Serra M., Purdy R., Concas A., 2000. Neurosteroids and GABA(A) receptor plasticity In: Martin D.L., Olsen R.W. (Eds.), GABA in the nervous system: the view at fifty years. Williams and Wilkins, Philadelphia, pp. 207-232. Biggio G., Littera M., Giliberto G., Purdy R., Serra M., 1998. Carbamazepine elevates plasma and brain content of neuroactive steroids (abstract). Soc Neurosci Abstr 24, 998. Birzniece V., Turkmen S., Lindblad C., Zhu D., Johansson I.M., Backstrom T., Wahlstrom G., 2006. GABA(A) recep- tor changes in acute allopregnanolone tolerance. Eur. J. Pharmacol. 535, 125-134. Bitran D., Shiekh M., Mcleod M., 1995. Anxiolytic Effect of Progesterone Is Mediated by the Neurosteroid Allo- pregnanolone at Brain Gaba(a) Receptors. Journal of Neuroendocrinology 7, 171-177. Bixo M., Andersson A., Winblad B., Purdy R.H., Backstrom T., 1997. Progesterone, 5a-pregnane-3,20-dione and 3a-hydroxy-5a-pregnane-20-one in specific regions of the human female brain in different endocrine states. Brain Res. 764, 173-178. Björn I., Sundström-Poromaa I., Bixo M., Nyberg S., Bäckström G., Bäckström T., 2003. Increase of estrogen dose deteriorates mood during progestin phase in sequential hormonal therapy. J Clin Endocrinol Metab 88, 2026-2030. Bloch M., Schmidt P.J., Danaceau M.A., Adams L.F., Rubinow D.R., 1999. Dehydroepiandrosterone treatment of midlife dysthymia. Biol. Psychiatry 45, 1533-1541. Bonanno G., Raiteri M., 1987a. A carrier for GABA uptake exists on noradrenaline nerve endings in selective rat brain areas but not on serotonin terminals. J Neural Transm 69, 59-70. Bonanno G., Raiteri M., 1987b. Release-regulating GABAA receptors are present on noradrenergic nerve termi- nals in selective areas of the rat brain. Synapse 1, 254-257. Bond A., Lader M., 1974a. The use of analogue scales in rating subjective feelings. Br. J. Psychol. 47, 211-218. Bond A., Lader M., 1974b. The use of analogue scales in rating subjective feelings. British Journal of Medical Psychology 47, 211-218. Brambilla F., Biggio G., Pisu M.G., Bellodi L., Perna G., Bogdanovich-Djukic V., Purdy R.H., Serra M., 2003a. Neu- rosteroid secretion in panic disorder. Psychiatry Res. 118, 107-116.

158 Chapter 12 | References

Brambilla F., Mellado C., Alciati A., Pisu M.G., Purdy R.H., Zanone S., Perini G., Serra M., Biggio G., 2005. Plasma concentrations of anxiolytic neuroactive steroids in men with panic disorder. Psychiatry Res. 135, 185- 190. Brambilla P., Perez J., Barale F., Schettini G., Soares J.C., 2003b. GABAergic dysfunction in mood disorders. Mol. Psychiatry 8, 721-737, 715. Brett M., Anton J., Valbregue R., Poline J. (2002) Region of interest analysis using an SPM toolbox [abstract] 8th International Conference on Functional Mapping of the Human Brain, Sendai, Japan, pp 497 Brodie B.B., Pletscher A., Shore P.A., 1955. Evidence that serotonin has a role in brain function. Science 122, 968. Browne E.S., Wright B.E., Porter J.R., Svec F., 1992. Dehydroepiandrosterone: antiglucocorticoid action in mice. Am J Med Sci 303, 366-371. Carl P., Hogskilde S., Lang-Jensen T., Bach V., Jacobsen J., Sorensen M.B., Gralls M., Widlund L., 1994. Pharma- cokinetics and pharmacodynamics of eltanolone (pregnanolone), a new steroid intravenous anaesthetic, in humans. Acta Anaesthesiol. Scand. 38, 734-741. Carl P., Hogskilde S., Nielsen J.W., Sorensen M.B., Lindholm M., Karlen B., Backstrom T., 1990. Pregnanolone emul- sion. A preliminary pharmacokinetic and pharmacodynamic study of a new intravenous anaesthetic agent. Anaesthesia 45, 189-197. Cashin M.F., Moravek V., 1927. The physiological action of cholesterol. Am. J. Physiol. 82, 294-298. Celotti F., Negri-Cesi P., Poletti A., 1997. Steroid metabolism in the mammalian brain: 5alpha-reduction and aro- matization. Brain Res. Bull. 44, 365-375. Chantilis S., Dombroski R., Shackleton C.H., Casey M.L., MacDonald P.C., 1996. Metabolism of 5 alpha-dihydro- progesterone in women and men: 3 beta- and 3 alpha-,6 alpha-dihydroxy-5 alpha-pregnan-20-ones are major urinary metabolites. J. Clin. Endocrinol. Metab. 81, 3644-3649. Chaudron L.H., Klein M.H., Remington P., Palta M., Allen C., Essex M.J., 2001. Predictors, prodromes and incidence of postpartum depression. J. Psychosom. Obstet. Gynaecol. 22, 103-112. Chen S.W., Rodriguez L., Davies M.F., Loew G.H., 1996. The hyperphagic effect of 3 alpha-hydroxylated pregnane steroids in male rats. Pharmacol. Biochem. Behav. 53, 777-782. Coenen A.M., van Luijtelaar E.L., 1989. Effects of diazepam and two beta-carbolines on epileptic activity and on EEG and behavior in rats with absence seizures. Pharmacol. Biochem. Behav. 32, 27-35. Commissie Farmaceutische Hulp van het College voor Zorgverzekeringen, 2006. Farmacotherapeutisch Kom- pas. College voor Zorgverzekeringen, Diemen. Corpechot C., Robel P., Axelson M., Sjovall J., Baulieu E.E., 1981. Characterization and measurement of dehydroe- piandrosterone sulfate in rat brain. Proc. Natl. Acad. Sci. U. S. A. 78, 4704-4707. Corpechot C., Synguelakis M., Talha S., Axelson M., Sjovall J., Vihko R., Baulieu E.E., Robel P., 1983. Pregnenolone and its sulfate ester in the rat brain. Brain Res. 270, 119-125.

159 Chapter 12 | References

Crawley J.N., Glowa J.R., Majewska M.D., Paul S.M., 1986. Anxiolytic activity of an endogenous . Brain Res. 398, 382-385. Davis M., Whalen P.J., 2001a. The amygdala: vigilance and emotion. Molecular Psychiatry 6, 13-34. Davis M., Whalen P.J., 2001b. The amygdala: vigilance and emotion. Mol. Psychiatry 6, 13-34. de Lignieres B., Dennerstein L., Backstrom T., 1995. Influence of route of administration on progesterone me- tabolism. Maturitas 21, 251-257. de Wit H., Rukstalis M., 1997. Acute effects of triazolam in women: relationships with progesterone, estradiol and allopregnanolone. Psychopharmacology (Berl) 130, 69-78. Delgado P.L., 2000. Depression: the case for a monoamine deficiency. J. Clin. Psychiatry 61 Suppl 6, 7-11. Delgado P.L., Moreno F.A., 2000. Role of norepinephrine in depression. J. Clin. Psychiatry 61 Suppl 1, 5-12. Deshpande L.S., Khisti R.T., Chopde C.T., 2001. Cataleptic effect of neurosteroid 3alpha-hydroxy-5alpha-pregnan- 20-one in mice: modulation by serotonergic agents. Brain Res. 898, 13-26. Dimmock P.W., Wyatt K.M., Jones P.W., O’Brien P.M., 2000. Efficacy of selective serotonin-reuptake inhibitors in premenstrual syndrome: a systematic review. Lancet 356, 1131-1136. Dong E., Matsumoto K., Uzunova V., Sugaya I., Takahata H., Nomura H., Watanabe H., Costa E., Guidotti A., 2001. Brain 5alpha-dihydroprogesterone and allopregnanolone synthesis in a mouse model of protracted so- cial isolation. Proc. Natl. Acad. Sci. U. S. A. 98, 2849-2854. Drevets W.C., 2003. Neuroimaging abnormalities in the amygdala in mood disorders. Ann N Y Acad Sci 985, 420-444. Droogleever Fortuyn H.A., van Broekhoven F., Span P.N., Backstrom T., Zitman F.G., Verkes R.J., 2004. Effects of PhD examination stress on allopregnanolone and cortisol plasma levels and peripheral benzodiazepine receptor density. Psychoneuroendocrinology 29, 1341-1344. Drugan R.C., 1996. Peripheral benzodiazepine receptors: molecular pharmacology to possible physiological sig- nificance in stress-induced hypertension. Clin. Neuropharmacol. 19, 475-496. Eriksson E., 1999. Serotonin reuptake inhibitors for the treatment of premenstrual dysphoria. Int Clin Psychop- harmacol 14 Suppl 2, S27-33. Eser D., Schule C., Baghai T.C., Romeo E., Rupprecht R., 2006. Neuroactive steroids in depression and anxiety disorders: clinical studies. Neuroendocrinology 84, 244-254. Fabian T.J., Dew M.A., Pollock B.G., Reynolds C.F., 3rd, Mulsant B.H., Butters M.A., Zmuda M.D., Linares A.M., Trot- tini M., Kroboth P.D., 2001. Endogenous concentrations of DHEA and DHEA-S decrease with remission of depression in older adults. Biol. Psychiatry 50, 767-774. Fadalti M., Petraglia F., Luisi S., Bernardi F., Casarosa E., Ferrari E., Luisi M., Saggese G., Genazzani A.R., Bernas- coni S., 1999. Changes of serum allopregnanolone levels in the first 2 years of life and during pubertal development. Pediatr. Res. 46, 323-327. Favale E., Rubino V., Mainardi P., Lunardi G., Albano C., 1995. Anticonvulsant effect of fluoxetine in humans. Neurology 45, 1926-1927.

160 Chapter 12 | References

Ferguson H.C., Bartram A.C., Fowlie H.C., Cathro D.M., Birchall K., Mitchell F.L., 1964. A Preliminary Investigation of Steroid Excretion in Depressed Patients before and after Electrol-Convulsive Therapy. Acta Endocri- nol (Copenh) 47, 58-68. Fernández G., Weis S., Stoffel-Wagner B., Tendolkar I., Reuber M., Beyenburg S., Klaver P., Fell J., de Greiff A., Ruhlmann J., Reul J., Elger C.E., 2003. Menstrual cycle-dependent neural plasticity in the adult human brain is hormone, task, and region specific. J Neurosci 23, 3790-3795. Ferrara G., Serra M., Zammaretti F., Pisu M.G., Panzica G.C., Biggio G., Eva C., 2001. Increased expression of the neuropeptide Y receptor Y(1) gene in the medial amygdala of transgenic mice induced by long-term treatment with progesterone or allopregnanolone. J. Neurochem. 79, 417-425. Finn D.A., Gee K.W., 1994. The estrus cycle, sensitivity to convulsants and the anticonvulsant effect of a neuroac- tive steroid. J. Pharmacol. Exp. Ther. 271, 164-170. Finn D.A., Phillips T.J., Okorn D.M., Chester J.A., Cunningham C.L., 1997. Rewarding effect of the neuroactive steroid 3 alpha-hydroxy-5 alpha-pregnan-20-one in mice. Pharmacol. Biochem. Behav. 56, 261-264. Fish E.W., Faccidomo S., DeBold J.F., Miczek K.A., 2001. Alcohol, allopregnanolone and aggression in mice. Psy- chopharmacology (Berl) 153, 473-483. Follesa P., Biggio F., Caria S., Gorini G., Biggio G., 2004. Modulation of GABA(A) receptor gene expression by al- lopregnanolone and ethanol. Eur. J. Pharmacol. 500, 413-425. Follesa P., Porcu P., Sogliano C., Cinus M., Biggio F., Mancuso L., Mostallino M.C., Paoletti A.M., Purdy R.H., Biggio G., Concas A., 2002. Changes in GABAA receptor gamma 2 subunit gene expression induced by long- term administration of oral contraceptives in rats. Neuropharmacology 42, 325-336. Freeman E.W., 2004. Luteal phase administration of agents for the treatment of premenstrual dysphoric disor- der. CNS Drugs 18, 453-468. Freeman E.W., Frye C.A., Rickels K., Martin P.A., Smith S.S., 2002. Allopregnanolone levels and symptom improve- ment in severe premenstrual syndrome. J Clin Psychopharmacol 22, 516-520. Freeman E.W., Purdy R.H., Coutifaris C., Rickels K., Paul S.M., 1993. Anxiolytic metabolites of progesterone: cor- relation with mood and performance measures following oral progesterone administration to healthy female volunteers. Neuroendocrinology 58, 478-484. Friston K.J., Buechel C., Fink G.R., Morris J., Rolls E., Dolan R.J., 1997. Psychophysiological and modulatory inter- actions in neuroimaging. Neuroimage 6, 218-229. Friston K.J., Holmes A.P., Worsley K.J., Poline J.B., Frith C.D., Frackowiak R.S.J., 1995. Statistical parametric maps in functional imaging: A general linear approach. Human Brain Mapping 2, 189-210. Friston K.J., Penny W., Phillips C., Kiebel S., Hinton G., Ashburner J., 2002. Classical and Bayesian inference in neuroimaging: theory. Neuroimage 16, 465-483. Gallagher P., Young A., 2002. Cortisol/DHEA ratios in depression. Neuropsychopharmacology 26, 410. Ganong W., 1991, 15th edition. Review of medical physiology. Appleton and Lange, London.

161 Chapter 12 | References

Gavish M., Weizman A., Karp L., Tyano S., Tanne Z., 1986. Decreased peripheral benzodiazepine binding sites in platelets of neuroleptic-treated schizophrenics. Eur. J. Pharmacol. 121, 275-279. Gehlert S., Hartlage S., 1997. A design for studying the DSM-IV research criteria of premenstrual dysphoric dis- order. J. Psychosom. Obstet. Gynaecol. 18, 36-44. Genazzani A.R., Petraglia F., Bernardi F., Casarosa E., Salvestroni C., Tonetti A., Nappi R.E., Luisi S., Palumbo M., Purdy R.H., Luisi M., 1998. Circulating levels of allopregnanolone in humans: gender, age, and endocrine influences. J. Clin. Endocrinol. Metab. 83, 2099-2103. Gigli G.L., Diomedi M., Troisi A., Marciani M.G., Pasini A., 1996. Fluoxetine and seizures. Neurology 47, 303. Girdler S.S., Straneva P.A., Light K.C., Pedersen C.A., Morrow A.L., 2001. Allopregnanolone levels and reactivity to mental stress in premenstrual dysphoric disorder. Biol. Psychiatry 49, 788-797. Givalois L., Grinevich V., Li S., Garcia-De-Yebenes E., Pelletier G., 1998. The octadecaneuropeptide-induced re- sponse of corticotropin-releasing hormone messenger RNA levels is mediated by GABA(A) receptors and modulated by endogenous steroids. Neuroscience 85, 557-567. Glue P., Bailey J., Wilson S., Hudson A., Nutt D.J., 1992. Thyrotropin-releasing hormone selectively reverses lo- razepam-induced sedation but not slowing of saccadic eye movements. Life Sci. 50, PL25-30. Goldstein J.M., Jerram M., Poldrack R., Ahern T., Kennedy D.N., Seidman L.J., Makris N., 2005. Hormonal cycle modulates arousal circuitry in women using functional magnetic resonance imaging. J Neurosci 25, 9309-9316. Goodyer I.M., Herbert J., Altham P.M., Pearson J., Secher S.M., Shiers H.M., 1996. Adrenal secretion during major depression in 8- to 16-year-olds, I. Altered diurnal rhythms in salivary cortisol and dehydroepiandroster- one (DHEA) at presentation. Psychol. Med. 26, 245-256. Goodyer I.M., Park R.J., Herbert J., 2001. Psychosocial and endocrine features of chronic first-episode major depression in 8-16 year olds. Biol. Psychiatry 50, 351-357. Gray H.S., Holt B.L., Whitaker D.K., Eadsforth P., 1992. Preliminary study of a pregnanolone emulsion (Kabi 2213) for i.v. induction of general anaesthesia. Br. J. Anaesth. 68, 272-276. Green K.L., Azarov A.V., Szeliga K.T., Purdy R.H., Grant K.A., 1999. The influence of menstrual cycle phase on sensitivity to ethanol-like discriminative stimulus effects of GABA(A)-positive modulators. Pharmacol. Biochem. Behav. 64, 379-383. Griffin L.D., Mellon S.H., 1999. Selective serotonin reuptake inhibitors directly alter activity of neurosteroidog- enic enzymes. Proc. Natl. Acad. Sci. U. S. A. 96, 13512-13517. Grobin A.C., Morrow A.L., 2000. 3alpha-hydroxy-5alpha-pregnan-20-one exposure reduces GABA(A) receptor alpha4 subunit mRNA levels. Eur. J. Pharmacol. 409, R1-2. Gron G., Friess E., Herpers M., Rupprecht R., 1997. Assessment of cognitive performance after progesterone administration in healthy male volunteers. Neuropsychobiology 35, 147-151.

162 Chapter 12 | References

Guidotti A., Costa E., 1998. Can the antidysphoric and anxiolytic profiles of selective serotonin reuptake inhibi- tors be related to their ability to increase brain 3 alpha, 5 alpha-tetrahydroprogesterone (allopregna- nolone) availability? Biol. Psychiatry 44, 865-873. Gulyas A.I., Acsady L., Freund T.F., 1999. Structural basis of the cholinergic and serotonergic modulation of GABAergic neurons in the hippocampus. Neurochem. Int. 34, 359-372. Hammarbäck S., Damber J.E., Bäckström T., 1989. Relationship between symptom severity and hormone chang- es in women with premenstrual syndrome. J Clin Endocrinol Metab 68, 125-130. Hansen C.R., Jr., Kroll J., Mackenzie T.B., 1982. Dehydroepiandrosterone and affective disorders. Am. J. Psychia- try 139, 386-387. Hariri A.R., Mattay V.S., Tessitore A., Fera F., Smith W.G., Weinberger D.R., 2002. Dextroamphetamine modulates the response of the human amygdala. Neuropsychopharmacology 27, 1036-1040. Hariri A.R., Mattay V.S., Tessitore A., Fera F., Weinberger D.R., 2003. Neocortical modulation of the amygdala response to fearful stimuli. Biol Psychiatry 53, 494-501. Havlikova H., Hill M., Kancheva L., Vrbikova J., Pouzar V., Cerny I., Kancheva R., Starka L., 2006. Serum profiles of free and conjugated neuroactive pregnanolone isomers in nonpregnant women of fertile age. J. Clin. Endocrinol. Metab. 91, 3092-3099. Hechter O., Grossman A., Chatterton R.T., Jr., 1997. Relationship of dehydroepiandrosterone and cortisol in dis- ease. Med Hypotheses 49, 85-91. Herbert J., 1998. Neurosteroids, brain damage, and mental illness. Exp Gerontol 33, 713-727. Herings R.M.C., 1994. Geneesmiddelen als determinant van ongevallen. Universiteit van Utrecht, Faculteit Far- macie, Utrecht. Herman J.P., Ostrander M.M., Mueller N.K., Figueiredo H., 2005. Limbic system mechanisms of stress regulation: hypothalamo-pituitary-adrenocortical axis. Prog Neuropsychopharmacol Biol Psychiatry 29, 1201-1213. Herzog A.G., Frye C.A., 2003. Seizure exacerbation associated with inhibition of progesterone metabolism. Ann. Neurol. 53, 390-391. Heuser I., Deuschle M., Luppa P., Schweiger U., Standhardt H., Weber B., 1998. Increased diurnal plasma concen- trations of dehydroepiandrosterone in depressed patients. J. Clin. Endocrinol. Metab. 83, 3130-3133. Heydari B., Le Melledo J.M., 2002. Low pregnenolone sulphate plasma concentrations in patients with general- ized social phobia. Psychol. Med. 32, 929-933. Hindmarch I., 1980. Psychomotor function and psychoactive drugs. Br. J. Clin. Pharmacol. 10, 189-209. Holdstock L., Penland S.N., Morrow A.L., de Wit H., 2006. Moderate doses of ethanol fail to increase plasma lev- els of neurosteroid 3alpha-hydroxy-5alpha-pregnan-20-one-like immunoreactivity in healthy men and women. Psychopharmacology (Berl) 186, 442-450. Holsboer D.F., 1995. Neuroendocrinology of mood disorders. In: Bloom F.E., Kupfer D.J. (Eds.), Psychopharmacol- ogy: the fourth generation of progress. Raven Press, New York, pp. 957-969.

163 Chapter 12 | References

Holsboer F., 2000. The corticosteroid receptor hypothesis of depression. Neuropsychopharmacology 23, 477- 501. Holsboer F., 2001. Stress, hypercortisolism and corticosteroid receptors in depression: implications for therapy. J. Affect. Disord. 62, 77-91. Holzbauer M., 1976. Physiological aspects of steroids with anaesthetic properties. Med Biol 54, 227-242. Hommer D.W., Matsuo V., Wolkowitz O., Chrousos G., Greenblatt D.J., Weingartner H., Paul S.M., 1986. Benzodi- azepine sensitivity in normal human subjects. Arch. Gen. Psychiatry 43, 542-551. Hopfenbeck J.R., Cowley D.S., Radant A., Greenblatt D.J., Roy-Byrne P.P., 1995. Effects of diphenhydramine on human eye movements. Psychopharmacology (Berl) 118, 280-286. Hosie A.M., Wilkins M.E., da Silva H.M.A., Smart T.G., 2006. Endogenous neurosteroids regulate GABAA receptors through two discrete transmembrane sites. Nature 444, 486-489. Hyyppa M.T., Kronholm E., Virtanen A., Leino A., Jula A., 2003. Does simvastatin affect mood and steroid hor- mone levels in hypercholesterolemic men? A randomized double-blind trial. Psychoneuroendocrinology 28, 181-194. in ‘t Veld B.A., Launer L.J., Breteler M.M., Hofman A., Stricker B.H., 2002. Pharmacologic agents associated with a preventive effect on Alzheimer’s disease: a review of the epidemiologic evidence. Epidemiol. Rev 24, 248-268. Jaworska-Feil L., Budziszewska B., Leskiewicz M., Lason W., 2000. Effects of some centrally active drugs on the allopregnanolone synthesis in rat brain. Pol J Pharmacol 52, 359-365. Jones B.C., Little A.C., Boothroyd L., Debruine L.M., Feinberg D.R., Smith M.J., Cornwell R.E., Moore F.R., Perrett D.I., 2005. Commitment to relationships and preferences for femininity and apparent health in faces are strongest on days of the menstrual cycle when progesterone level is high. Horm Behav 48, 283-290. Kalimi M., Shafagoj Y., Loria R., Padgett D., Regelson W., 1994. Anti-glucocorticoid effects of dehydroepiandros- terone (DHEA). Mol Cell Biochem 131, 99-104. Kanwisher N., McDermott J., Chun M.M., 1997. The fusiform face area: a module in human extrastriate cortex specialized for face perception. J Neurosci 17, 4302-4311. Karp L., Weizman A., Tyano S., Gavish M., 1989. Examination stress, platelet peripheral benzodiazepine binding sites, and plasma hormone levels. Life Sci. 44, 1077-1082. Khisti R.T., Chopde C.T., 2000. Serotonergic agents modulate antidepressant-like effect of the neurosteroid 3al- pha-hydroxy-5alpha-pregnan-20-one in mice. Brain Res. 865, 291-300. Khisti R.T., Chopde C.T., Jain S.P., 2000. Antidepressant-like effect of the neurosteroid 3alpha-hydroxy-5alpha- pregnan-20-one in mice forced swim test. Pharmacol. Biochem. Behav. 67, 137-143. Khisti R.T., Penland S.N., VanDoren M.J., Grobin A.C., Morrow A.L., 2002. GABAergic neurosteroid modulation of ethanol actions. World J. Biol. Psychiatry 3, 87-95.

164 Chapter 12 | References

Khisti R.T., VanDoren M.J., Matthews D.B., Morrow A.L., 2004. Ethanol-induced elevation of 3 alpha-hydroxy-5 alpha-pregnan-20-one does not modulate motor incoordination in rats. Alcohol Clin. Exp. Res. 28, 1249- 1256. Khisti R.T., VanDoren M.J., O’Buckley T., Morrow A.L., 2003. Neuroactive steroid 3 alpha-hydroxy-5 alpha-preg- nan-20-one modulates ethanol-induced loss of righting reflex in rats. Brain Res. 980, 255-265. Kimura D., 1999. Sex and cognition. The MIT Press, Cambridge. Kinouchi K., Maeda S., Saito K., Inoki R., Fukumitsu K., Yoshiya I., 1989. Effects of d- and l-pentazocine on the release and uptake of norepinephrine in rat brain cortex. Res Commun Chem Pathol Pharmacol 63, 201- 213. Kinsora J.J., Corbin A.E., Snyder B.J., Wiley J.N., Meltzer L.T., Heffner T.G., 1998. Effects of igmesine in preclinical antidepressant tests (abstract). Soc Neurosci Abstr 24, 744. Kirschbaum C., Kudielka B.M., Gaab J., Schommer N.C., Hellhammer D.H., 1999. Impact of gender, menstrual cycle phase, and oral contraceptives on the activity of the hypothalamus-pituitary-adrenal axis. Psychosom Med 61, 154-162. Klink R., Robichaud M., Debonnel G., 2002. Gender and gonadal status modulation of dorsal raphe nucleus sero- tonergic neurons. Part I: effects of gender and pregnancy. Neuropharmacology 43, 1119-1128. Kokate T.G., Svensson B.E., Rogawski M.A., 1994. Anticonvulsant activity of neurosteroids: correlation with gam- ma-aminobutyric acid-evoked chloride current potentiation. J. Pharmacol. Exp. Ther. 270, 1223-1229. Kroboth P.D., Salek F.S., Pittenger A.L., Fabian T.J., Frye R.F., 1999. DHEA and DHEA-S: a review. J Clin Pharmacol 39, 327-348. Krueger K.E., Papadopoulos V., 1990. Peripheral-type benzodiazepine receptors mediate translocation of cho- lesterol from outer to inner mitochondrial membranes in adrenocortical cells. J Biol Chem 265, 15015- 15022. Krueger K.E., Papadopoulos V., 1992. Mitochondrial benzodiazepine receptors and the regulation of steroid bio- synthesis. Annu Rev Pharmacol Toxicol 32, 211-237. Kurshan N., Neill Epperson C., 2006. Oral contraceptives and mood in women with and without premenstrual dysphoria: a theoretical model. Arch. Womens Ment. Health 9, 1-14. Laconi M.R., Casteller G., Gargiulo P.A., Bregonzio C., Cabrera R.J., 2001. The anxiolytic effect of allopregna- nolone is associated with gonadal hormonal status in female rats. Eur. J. Pharmacol. 417, 111-116. Lambert J.J., Belelli D., Harney S.C., Peters J.A., Frenguelli B.G., 2001. Modulation of native and recombinant GABA(A) receptors by endogenous and synthetic neuroactive steroids. Brain Res. Brain Res. Rev. 37, 68-80. Lan N.C., Gee K.W., 1994. Neuroactive steroid actions at the GABAA receptor. Horm. Behav. 28, 537-544. Lancel M., 1999. Role of GABAA receptors in the regulation of sleep: initial sleep responses to peripherally ad- ministered modulators and agonists. Sleep 22, 33-42.

165 Chapter 12 | References

Lancel M., Faulhaber J., Schiffelholz T., Mathias S., Deisz R.A., 1997a. Muscimol and midazolam do not potentiate each other’s effects on sleep EEG in the rat. J. Neurophysiol. 77, 1624-1629. Lancel M., Faulhaber J., Schiffelholz T., Romeo E., Di Michele F., Holsboer F., Rupprecht R., 1997b. Allopregna- nolone affects sleep in a benzodiazepine-like fashion. J. Pharmacol. Exp. Ther. 282, 1213-1218. Le Melledo J.M., Van Driel M., Coupland N.J., Lott P., Jhangri G.S., 2000. Response to flumazenil in women with premenstrual dysphoric disorder. Am. J. Psychiatry 157, 821-823. Legrain S., Berr C., Frenoy N., Gourlet V., Debuire B., Baulieu E.E., 1995. Dehydroepiandrosterone sulfate in a long-term care aged population. Gerontology 41, 343-351. Lenzinger E., Diamant K., Vytiska-Binstorfer E., Kasper S., 1997. [Premenstrual dysphoric disorder. An overview of diagnosis, epidemiology and therapeutic approaches]. Nervenarzt 68, 708-718. Locchi F., Dall’olio R., Gandolfi O., Rimondini R., 2008. Olanzapine counteracts stress-induced anxiety-like behav- ior in rats. Neurosci. Lett. 438, 146-149. Luisi S., Petraglia F., Benedetto C., Nappi R.E., Bernardi F., Fadalti M., Reis F.M., Luisi M., Genazzani A.R., 2000. Serum allopregnanolone levels in pregnant women: changes during pregnancy, at delivery, and in hyper- tensive patients. J. Clin. Endocrinol. Metab. 85, 2429-2433. Maayan R., Yagorowski Y., Grupper D., Weiss M., Shtaif B., Kaoud M.A., Weizman A., 2000. Basal plasma de- hydroepiandrosterone sulfate level: a possible predictor for response to electroconvulsive therapy in depressed psychotic inpatients. Biol. Psychiatry 48, 693-701. Maes M., Meltzer H., 1995. The serotonin hypothesis of major depression. In: Bloom F.E., Kupfer D.J. (Eds.), Psy- chopharmacology: the fourth generation of progress. Raven Press, New York, pp. 933-944. Majewska M.D., 1992. Neurosteroids: endogenous bimodal modulators of the GABAA receptor. Mechanism of ac- tion and physiological significance. Prog Neurobiol 38, 379-395. Majewska M.D., Harrison N.L., Schwartz R.D., Barker J.L., Paul S.M., 1986. Steroid hormone metabolites are bar- biturate-like modulators of the GABA receptor. Science 232, 1004-1007. Maldjian J.A., Laurienti P.J., Kraft R.A., Burdette J.H., 2003. An automated method for neuroanatomic and cy- toarchitectonic atlas-based interrogation of fMRI data sets. Neuroimage 19, 1233-1239. Manfredini R., Caracciolo S., Salmi R., Boari B., Tomelli A., Gallerani M., 2000. The association of low serum cho- lesterol with depression and suicidal behaviours: new hypotheses for the missing link. J. Int. Med. Res. 28, 247-257. Marowsky A., Yanagawa Y., Obata K., Vogt K.E., 2005. A specialized subclass of interneurons mediates dopamin- ergic facilitation of amygdala function. Neuron 48, 1025-1037. Marx C.E., Duncan G.E., Gilmore J.H., Lieberman J.A., Morrow A.L., 2000. Olanzapine increases allopregnanolone in the rat cerebral cortex. Biol. Psychiatry 47, 1000-1004. Marx C.E., Stevens R.D., Shampine L.J., Uzunova V., Trost W.T., Butterfield M.I., Massing M.W., Hamer R.M., Mor- row A.L., Lieberman J.A., 2006. Neuroactive steroids are altered in schizophrenia and bipolar disorder: relevance to pathophysiology and therapeutics. Neuropsychopharmacology 31, 1249-1263.

166 Chapter 12 | References

Matsumoto K., Uzunova V., Pinna G., Taki K., Uzunov D.P., Watanabe H., Mienville J.M., Guidotti A., Costa E., 1999. Permissive role of brain allopregnanolone content in the regulation of -induced righting reflex loss. Neuropharmacology 38, 955-963. Matsuno K., Kobayashi T., Tanaka M.K., Mita S., 1996. Sigma 1 receptor subtype is involved in the relief of behav- ioral despair in the mouse forced swimming test. Eur. J. Pharmacol. 312, 267-271. Matsuura K., Shiraishi H., Hara A., Sato K., Deyashiki Y., Ninomiya M., Sakai S., 1998. Identification of a principal mRNA species for human 3alpha-hydroxysteroid dehydrogenase isoform (AKR1C3) that exhibits high prostaglandin D2 11-ketoreductase activity. J. Biochem. 124, 940-946. Maurice T., Phan V.L., Urani A., Kamei H., Noda Y., Nabeshima T., 1999. Neuroactive neurosteroids as endogenous effectors for the sigma1 (sigma1) receptor: pharmacological evidence and therapeutic opportunities. Jpn J Pharmacol 81, 125-155. Maurice T., Roman F.J., Privat A., 1996. Modulation by neurosteroids of the in vivo (+)-[3H]SKF-10,047 binding to sigma 1 receptors in the mouse forebrain. J Neurosci Res 46, 734-743. Maurice T., Urani A., Phan V.L., Romieu P., 2001. The interaction between neuroactive steroids and the sigma1 receptor function: behavioral consequences and therapeutic opportunities. Brain Res. Brain Res. Rev. 37, 116-132. McAuley J.W., Friedman C.I., 1999. Influence of endogenous progesterone on alprazolam pharmacodynamics. J. Clin. Psychopharmacol. 19, 233-239. McAuley J.W., Reynolds I.J., Kroboth F.J., Smith R.B., Kroboth P.D., 1995. Orally administered progesterone en- hances sensitivity to triazolam in postmenopausal women. J. Clin. Psychopharmacol. 15, 3-11. McEwen B.S., 1991. Non-genomic and genomic effects of steroids on neural activity. Trends Pharmacol. Sci. 12, 141-147. Meikle A.W., Dorchuck R.W., Araneo B.A., Stringham J.D., Evans T.G., Spruance S.L., Daynes R.A., 1992. The pres- ence of a dehydroepiandrosterone-specific receptor binding complex in murine T cells. J Steroid Bio- chem Mol Biol 42, 293-304. Melchior C.L., Ritzmann R.F., 1994. Dehydroepiandrosterone is an anxiolytic in mice on the plus maze. Pharma- col. Biochem. Behav. 47, 437-441. Mellon S.H., 1994. Neurosteroids: biochemistry, modes of action, and clinical relevance. J. Clin. Endocrinol. Me- tab. 78, 1003-1008. Meltzer H., 1989. Serotonergic dysfunction in depression. Br J Psychiatry Suppl 25-31. Menozzi R., Florio P., Bondi M., Luisi S., Cobellis L., Genazzani A.R., Del Rio G., Petraglia F., 2002. Increased response of plasma allopregnanolone to corticotropin-releasing hormone in obese patients. Neuroen- docrinology 75, 124-129. Merryman W., Boiman R., Barnes L., Rothchild I., 1954. Progesterone anesthesia in human subjects. J. Clin. En- docrinol. Metab. 14, 1567-1569.

167 Chapter 12 | References

Michael A., Jenaway A., Paykel E.S., Herbert J., 2000. Altered salivary dehydroepiandrosterone levels in major depression in adults. Biol. Psychiatry 48, 989-995. Miczek K.A., Fish E.W., De Bold J.F., 2003. Neurosteroids, GABAA receptors, and escalated aggressive behavior. Horm Behav 44, 242-257. Mohan P.F., Cleary M.P., 1992. Studies on nuclear binding of dehydroepiandrosterone in hepatocytes. Steroids 57, 244-247. Monaghan E.P., Navalta L.A., Shum L., Ashbrook D.W., Lee D.A., 1997. Initial human experience with ganaxolone, a neuroactive steroid with antiepileptic activity. Epilepsia 38, 1026-1031. Monnet F.P., Mahe V., Robel P., Baulieu E.E., 1995. Neurosteroids, via sigma receptors, modulate the [3H]norepinephrine release evoked by N-methyl-D-aspartate in the rat hippocampus. Proc. Natl. Acad. Sci. U. S. A. 92, 3774-3778. Monteleone P., Fabrazzo M., Serra M., Tortorella A., Pisu M.G., Biggio G., Maj M., 2004. Long-term treatment with clozapine does not affect morning circulating levels of allopregnanolone and THDOC in patients with schizophrenia: a preliminary study. J. Clin. Psychopharmacol. 24, 437-440. Monteleone P., Luisi M., Colurcio B., Casarosa E., Ioime R., Genazzani A.R., Maj M., 2001. Plasma levels of neuroac- tive steroids are increased in untreated women with anorexia nervosa or bulimia nervosa. Psychosom. Med. 63, 62-68. Monteleone P., Luisi S., Tonetti A., Bernardi F., Genazzani A.D., Luisi M., Petraglia F., Genazzani A.R., 2000. Allo- pregnanolone concentrations and premenstrual syndrome. Eur J Endocrinol 142, 269-273. Morris K.D., Moorefield C.N., Amin J., 1999. Differential modulation of the gamma-aminobutyric acid type C re- ceptor by neuroactive steroids. Mol. Pharmacol. 56, 752-759. Morrison M.F., Ten Have T., Freeman E.W., Sammel M.D., Grisso J.A., 2001. DHEA-S levels and depressive symp- toms in a cohort of African American and Caucasian women in the late reproductive years. Biol. Psy- chiatry 50, 705-711. Morrow A.L., Pace J.R., Purdy R.H., Paul S.M., 1990. Characterization of steroid interactions with gamma-ami- nobutyric acid receptor-gated chloride ion channels: evidence for multiple steroid recognition sites. Mol. Pharmacol. 37, 263-270. N-Wihlbäck A.C., Sundström-Poromaa I., Bäckström T., 2006. Action by and sensitivity to neuroactive steroids in menstrual cycle related CNS disorders. Psychopharmacology 186, 388-401. Nadeson R., Goodchild C.S., 2001. Antinociceptive properties of neurosteroids III: experiments with alphadolone given intravenously, intraperitoneally, and intragastrically. Br. J. Anaesth. 86, 704-708. Nappi R.E., Petraglia F., Luisi S., Polatti F., Farina C., Genazzani A.R., 2001. Serum allopregnanolone in women with postpartum “blues”. Obstet. Gynecol. 97, 77-80. Nohria V., Giller E., 2007. Ganaxolone. Neurotherapeutics 4, 102-105. Norberg L., Backstrom T., Wahlstrom G., 1999. Anaesthetic effects of pregnanolone in combination with allopreg- nanolone, thiopental, and : an EEG study in the rat. Br. J. Anaesth. 82, 731-737.

168 Chapter 12 | References

Norberg L., Wahlstrom G., Backstrom T., 1987. The anaesthetic potency of 3 alpha-hydroxy-5 alpha-pregnan-20- one and 3 alpha-hydroxy-5 beta-pregnan-20-one determined with an intravenous EEG-threshold method in male rats. Pharmacol Toxicol 61, 42-47. Ochsner K.N., Gross J.J., 2005. The cognitive control of emotion. Trends Cogn Sci 9, 242-249. Orentreich N., Brind J.L., Rizer R.L., Vogelman J.H., 1984. Age changes and sex differences in serum dehydroepi- concentrations throughout adulthood. J. Clin. Endocrinol. Metab. 59, 551-555. Orentreich N., Brind J.L., Vogelman J.H., Andres R., Baldwin H., 1992. Long-term longitudinal measurements of plasma dehydroepiandrosterone sulfate in normal men. J. Clin. Endocrinol. Metab. 75, 1002-1004. Ormiston T., Wolkowitz O.M., Reus V.I., Johnson R., Manfredi F., 2004. Hormonal changes with cholesterol reduc- tion: a double-blind pilot study. J. Clin. Pharm. Ther. 29, 71-73. Orser B.A., Canning K.J., Macdonald J.F., 2002. Mechanisms of general anesthesia. Curr Opin Anaesthesiol 15, 427-433. Osran H., Reist C., Chen C.C., Lifrak E.T., Chicz-DeMet A., Parker L.N., 1993. Adrenal and cortisol in major depression. Am. J. Psychiatry 150, 806-809. Owens M.J., Nemeroff C.B., 1991. Physiology and pharmacology of corticotropin-releasing factor. Pharmacol. Rev. 43, 425-473. P’An S.Y., Gardocki J.F., Hutcheon D.E., Rudel H., Kodet M.J., Laubach G.D., 1955. General anesthetic and other pharmacological properties of a soluble steroid, 21-hydroxypregnanedione sodium succianate. J. Phar- macol. Exp. Ther. 115, 432-441. Padberg F., di Michele F., Zwanzger P., Romeo E., Bernardi G., Schule C., Baghai T.C., Ella R., Pasini A., Rupprecht R., 2002. Plasma concentrations of neuroactive steroids before and after repetitive transcranial mag- netic stimulation (rTMS) in major depression. Neuropsychopharmacology 27, 874-878. Paoletti A.M., Lello S., Fratta S., Orru M., Ranuzzi F., Sogliano C., Concas A., Biggio G., Melis G.B., 2004. Psycho- logical effect of the oral contraceptive formulation containing 3 mg of drospirenone plus 30 microg of ethinyl estradiol. Fertil. Steril. 81, 645-651. Paoletti A.M., Romagnino S., Contu R., Orru M.M., Marotto M.F., Zedda P., Lello S., Biggio G., Concas A., Melis G.B., 2006. Observational study on the stability of the psychological status during normal pregnancy and increased blood levels of neuroactive steroids with GABA-A receptor agonist activity. Psychoneuroen- docrinology 31, 485-492. Papadopoulos V., Brown A.S., 1995. Role of the peripheral-type benzodiazepine receptor and the polypeptide diazepam binding inhibitor in steroidogenesis. J Steroid Biochem Mol Biol 53, 103-110. Papadopoulos V., Lecanu L., Brown R.C., Han Z., Yao Z.X., 2006. Peripheral-type benzodiazepine receptor in neu- rosteroid biosynthesis, neuropathology and neurological disorders. Neuroscience 138, 749-756. Parivar K., Wessen A., Widman M., Nilsson A., 1996. Pharmacokinetics of eltanolone following bolus injection and constant rate infusion. J. Pharmacokinet. Biopharm. 24, 535-549.

169 Chapter 12 | References

Parizek A., Hill M., Kancheva R., Havlikova H., Kancheva L., Cindr J., Paskova A., Pouzar V., Cerny I., Drbohlav P., Hajek Z., Starka L., 2005. Neuroactive pregnanolone isomers during pregnancy. J Clin Endocrinol Metab 90, 395-403. Pasini A., Tortorella A., Gale K., 1996. The anticonvulsant action of fluoxetine in substantia nigra is dependent upon endogenous serotonin. Brain Res. 724, 84-88. Patchev V.K., Hassan A.H., Holsboer D.F., Almeida O.F., 1996. The neurosteroid tetrahydroprogesterone attenu- ates the endocrine response to stress and exerts glucocorticoid-like effects on vasopressin gene tran- scription in the rat hypothalamus. Neuropsychopharmacology 15, 533-540. Patchev V.K., Montkowski A., Rouskova D., Koranyi L., Holsboer F., Almeida O.F., 1997. Neonatal treatment of rats with the neuroactive steroid tetrahydrodeoxycorticosterone (THDOC) abolishes the behavioral and neu- roendocrine consequences of adverse early life events. J Clin Invest 99, 962-966. Patchev V.K., Shoaib M., Holsboer F., Almeida O.F., 1994. The neurosteroid tetrahydroprogesterone counteracts corticotropin-releasing hormone-induced anxiety and alters the release and gene expression of cortico- tropin-releasing hormone in the rat hypothalamus. Neuroscience 62, 265-271. Paul S.M., Purdy R.H., 1992. Neuroactive steroids. Faseb J 6, 2311-2322. Penton-Voak I.S., Perrett D.I., Castles D.L., Kobayashi T., Burt D.M., Murray L.K., Minamisawa R., 1999. Menstrual cycle alters face preference. Nature 399, 741-742. Peoples R.W., Giridhar J., Isom G.E., 1991. Gamma-aminobutyric acid enhancement of potassium-stimulated re- lease of [3H]norepinephrine by multiple mechanisms in rat cortical slices. Biochem Pharmacol 41, 119- 123. Pezawas L., Meyer-Lindenberg A., Drabant E.M., Verchinski B.A., Munoz K.E., Kolachana B.S., Egan M.F., Mattay V.S., Hariri A.R., Weinberger D.R., 2005. 5-HTTLPR polymorphism impacts human cingulate-amygdala interactions: a genetic susceptibility mechanism for depression. Nat Neurosci 8, 828-834. Phelps E.A., LeDoux J.E., 2005. Contributions of the amygdala to emotion processing: from animal models to human behavior. Neuron 48, 175-187. Phillips M.L., Drevets W.C., Rauch S.L., Lane R., 2003a. Neurobiology of emotion perception I: The neural basis of normal emotion perception. Biol. Psychiatry 54, 504-514. Phillips M.L., Drevets W.C., Rauch S.L., Lane R., 2003b. Neurobiology of emotion perception II: Implications for major psychiatric disorders. Biological Psychiatry 54, 515-528. Piccinelli M., Pini S., Bellantuono C., Wilkinson G., 1995. Efficacy of drug treatment in obsessive-compulsive dis- order. A meta-analytic review. Br. J. Psychiatry 166, 424-443. Pierucci-Lagha A., Covault J., Feinn R., Khisti R.T., Morrow A.L., Marx C.E., Shampine L.J., Kranzler H.R., 2006. Subjective effects and changes in steroid hormone concentrations in humans following acute consump- tion of alcohol. Psychopharmacology (Berl) 186, 451-461. Prasad A., Imamura M., Prasad C., 1997. Dehydroepiandrosterone decreases behavioral despair in high- but not low-anxiety rats. Physiol Behav 62, 1053-1057.

170 Chapter 12 | References

Price J.L., 2003. Comparative aspects of amygdala connectivity. Ann N Y Acad Sci 985, 50-58. Prince R.J., Simmonds M.A., 1993. Differential antagonism by of alphaxalone and pregnanolone potentiation of [3H] binding suggests more than one class of binding site for steroids at GABAA receptors. Neuropharmacology 32, 59-63. Protopopescu X., Pan H., Altemus M., Tuescher O., Polanecsky M., McEwen B., Silbersweig D., Stern E., 2005. Orbitofrontal cortex activity related to emotional processing changes across the menstrual cycle. Proc Natl Acad Sci U S A 102, 16060-16065. Purdy R.H., Moore P.H., Jr., Rao P.N., Hagino N., Yamaguchi T., Schmidt P., Rubinow D.R., Morrow A.L., Paul S.M., 1990. Radioimmunoassay of 3 alpha-hydroxy-5 alpha-pregnan-20-one in rat and human plasma. Steroids 55, 290-296. Purdy R.H., Morrow A.L., Moore P.H., Jr., Paul S.M., 1991. Stress-induced elevations of gamma-aminobutyric acid type A receptor-active steroids in the rat brain. Proc. Natl. Acad. Sci. U. S. A. 88, 4553-4557. Rahman M., Lindblad C., Johansson I.M., Backstrom T., Wang M.D., 2006. Neurosteroid modulation of recom- binant rat alpha5beta2gamma2L and alpha1beta2gamma2L GABA(A) receptors in Xenopus oocyte. Eur. J. Pharmacol. 547, 37-44. Rajkowski K.M., Robel P., Baulieu E.E., 1997. Hydroxysteroid sulfotransferase activity in the rat brain and liver as a function of age and sex. Steroids 62, 427-436. Rang H., Dale M., Ritter J., Moore P., 2003, fifth. Pharmacology. Churchill Livingstone, London. Rapkin A.J., Biggio G., Concas A., 2006. Oral contraceptives and neuroactive steroids. Pharmacol. Biochem. Behav. 84, 628-634. Rapkin A.J., Morgan M., Goldman L., Brann D.W., Simone D., Mahesh V.B., 1997. Progesterone metabolite allopreg- nanolone in women with premenstrual syndrome. Obstet. Gynecol. 90, 709-714. Rasgon N., McGuire M., Tanavoli S., Fairbanks L., Rapkin A., 2000. Neuroendocrine response to an intravenous L-tryptophan challenge in women with premenstrual syndrome. Fertil. Steril. 73, 144-149. Rauch S.L., Shin L.M., Wright C.I., 2003. Neuroimaging studies of amygdala function in anxiety disorders. Ann N Y Acad Sci 985, 389-410. Reddy D.S., Kaur G., Kulkarni S.K., 1998. Sigma (sigma1) receptor mediated anti-depressant-like effects of neu- rosteroids in the Porsolt forced swim test. Neuroreport 9, 3069-3073. Reddy D.S., O’Malley B.W., Rogawski M.A., 2005. Anxiolytic activity of progesterone in progesterone receptor knockout mice. Neuropharmacology 48, 14-24. Reus V.I., Wolkowitz O., Roberts E., Chan T., Turetsky N., Manfredi F., Weingartner H., 1993. Dehydroepiandroster- one (DHEA) and memory in depressed patients (abstract). Neuropsychopharmacology 9, 66S. Robichaud M., Debonnel G., 2006. Allopregnanolone and ganaxolone increase the firing activity of dorsal raphe nucleus serotonergic neurons in female rats. Int. J. Neuropsychopharmacol. 9, 191-200.

171 Chapter 12 | References

Roca C.A., Schmidt P.J., Altemus M., Deuster P., Danaceau M.A., Putnam K., Rubinow D.R., 2003. Differential men- strual cycle regulation of hypothalamic-pituitary-adrenal axis in women with premenstrual syndrome and controls. J Clin Endocrinol Metab 88, 3057-3063. Rogers C., Lemaire S., 1991. Role of the sigma receptor in the inhibition of [3H]-noradrenaline uptake in brain synaptosomes and adrenal chromaffin cells. Br. J. Pharmacol. 103, 1917-1922. Romano-Torres M., Borja-Lascurain E., Chao-Rebolledo C., del-Rio-Portilla Y., Corsi-Cabrera M., 2002. Effect of diazepam on EEG power and coherent activity: sex differences. Psychoneuroendocrinology 27, 821-833. Romeo E., Brancati A., De Lorenzo A., Fucci P., Furnari C., Pompili E., Sasso G.F., Spalletta G., Troisi A., Pasini A., 1996. Marked decrease of plasma neuroactive steroids during alcohol withdrawal. Clin. Neuropharmacol. 19, 366-369. Romeo E., Strohle A., Spalletta G., di Michele F., Hermann B., Holsboer F., Pasini A., Rupprecht R., 1998. Effects of antidepressant treatment on neuroactive steroids in major depression. Am. J. Psychiatry 155, 910-913. Rosenbaum J.F., Pollock R.A., Otto M.W., Pollack M.H., 1995. Integrated treatment of panic disorder. Bull. Men- ninger Clin. 59, A4-26. Rosenfeld R.S., Rosenberg B.J., Fukushima D.K., Hellman L., 1975. 24-Hour secretory pattern of dehydroisoan- drosterone and dehydroisoandrosterone sulfate. J. Clin. Endocrinol. Metab. 40, 850-855. Rostami K., Van Bergeijk L., Van Broekhoven F., Janssen B.P.W., Mulder C.J.J., 2001. Elevation of pancreatic en- zymes: a marker for intracranial hemorrhage. Romanian Journal of Gastroenterology 10, 61-63. Rubinow D.R., Schmidt P.J., 2006. Gonadal steroid regulation of mood: the lessons of premenstrual syndrome. Front Neuroendocrinol 27, 210-216. Rudolph U., Crestani F., Mohler H., 2001. GABA(A) receptor subtypes: dissecting their pharmacological functions. Trends Pharmacol. Sci. 22, 188-194. Rupprecht R., Holsboer F., 1999. Neuroactive steroids: mechanisms of action and neuropsychopharmacological perspectives. Trends Neurosci. 22, 410-416. Rupprecht R., Koch M., Montkowski A., Lancel M., Faulhaber J., Harting J., Spanagel R., 1999. Assessment of neuroleptic-like properties of progesterone. Psychopharmacology (Berl) 143, 29-38. Rupprecht R., Reul J.M., Trapp T., van Steensel B., Wetzel C., Damm K., Zieglgansberger W., Holsboer F., 1993. Progesterone receptor-mediated effects of neuroactive steroids. Neuron 11, 523-530. Salek F.S., Bigos K.L., Kroboth P.D., 2002. The influence of hormones and pharmaceutical agents on DHEA and DHEA-S concentrations: a review of clinical studies. J Clin Pharmacol 42, 247-266. Saletu B., Grunberger J., Cepko H., 1987. Pharmaco-EEG and psychometric studies with a novel selective benzo- diazepine agonist/antagonist Ro 23-0364. Int J Clin Pharmacol Ther Toxicol 25, 421-437. Sanders D., Warner P., Backstrom T., Bancroft J., 1983. Mood, sexuality, hormones and the menstrual cycle. I. Changes in mood and physical state: description of subjects and method. Psychosom. Med. 45, 487- 501.

172 Chapter 12 | References

Schiller P.H., Tehovnik E.J., 2003. Cortical inhibitory circuits in eye-movement generation. Eur. J. Neurosci. 18, 3127-3133. Schiller P.H., Tehovnik E.J., 2005. Neural mechanisms underlying target selection with saccadic eye movements. Prog. Brain Res. 149, 157-171. Schmidt P.J., Nieman L.K., Danaceau M.A., Adams L.F., Rubinow D.R., 1998. Differential behavioral effects of gonadal steroids in women with and in those without premenstrual syndrome. N Engl J Med 338, 209- 216. Schmidt P.J., Purdy R.H., Moore P.H., Jr., Paul S.M., Rubinow D.R., 1994. Circulating levels of anxiolytic steroids in the luteal phase in women with premenstrual syndrome and in control subjects. J. Clin. Endocrinol. Metab. 79, 1256-1260. Schule C., di Michele F., Baghai T., Romeo E., Bernardi G., Zwanzger P., Padberg F., Pasini A., Rupprecht R., 2003. Influence of sleep deprivation on neuroactive steroids in major depression. Neuropsychopharmacology 28, 577-581. Schule C., Romeo E., Uzunov D.P., Eser D., di Michele F., Baghai T.C., Pasini A., Schwarz M., Kempter H., Rupprecht R., 2006. Influence of mirtazapine on plasma concentrations of neuroactive steroids in major depression and on 3alpha-hydroxysteroid dehydrogenase activity. Mol. Psychiatry 11, 261-272. Schumacher M., Akwa Y., Guennoun R., Robert F., Labombarda F., Desarnaud F., Robel P., De Nicola A.F., Baulieu E.E., 2000. Steroid synthesis and metabolism in the nervous system: trophic and protective effects. J Neurocytol 29, 307-326. Scott L.V., Salahuddin F., Cooney J., Svec F., Dinan T.G., 1999. Differences in adrenal steroid profile in chronic fatigue syndrome, in depression and in health. J. Affect. Disord. 54, 129-137. Sear J.W., 1998. Eltanolone: 50 years on and still looking for steroid hypnotic agents! Eur. J. Anaesthesiol. 15, 129-132. Selye H., 1941a. Anesthetic effect of steroid hormones. Proc. Soc. Exp. Biol. Med. 46, 116-121. Selye H., 1941b. Studies concerning the anesthetic action of steroid hormones. J. Pharmacol. Exp. Ther. 73, 127- 141. Semeniuk T., Jhangri G.S., Le Melledo J.M., 2001. Neuroactive steroid levels in patients with generalized anxiety disorder. J. Neuropsychiatry Clin. Neurosci. 13, 396-398. Serra M., Littera M., Pisu M.G., Muggironi M., Purdy R.H., Biggio G., 2000. Steroidogenesis in rat brain induced by short- and long-term administration of carbamazepine. Neuropharmacology 39, 2448-2456. Serra M., Pisu M.G., Muggironi M., Parodo V., Papi G., Sari R., Dazzi L., Spiga F., Purdy R.H., Biggio G., 2001. Oppo- site effects of short- versus long-term administration of fluoxetine on the concentrations of neuroactive steroids in rat plasma and brain. Psychopharmacology (Berl) 158, 48-54. Serra M., Sanna E., Mostallino M.C., Biggio G., 2007. Social isolation stress and neuroactive steroids. Eur. Neu- ropsychopharmacol. 17, 1-11.

173 Chapter 12 | References

Shen H., Gong Q.H., Aoki C., Yuan M., Ruderman Y., Dattilo M., Williams K., Smith S.S., 2007. Reversal of neu- rosteroid effects at alpha4beta2delta GABA(A) receptors triggers anxiety at puberty. Nat Neurosci 10, 469-477. Shimada M., Yoshinari K., Tanabe E., Shimakawa E., Kobashi M., Nagata K., Yamazoe Y., 2001. Identification of ST2A1 as a rat brain neurosteroid sulfotransferase mRNA. Brain Res. 920, 222-225. Smith S.S., Ruderman Y., Frye C., Homanics G., Yuan M., 2006. Steroid withdrawal in the mouse results in anx- iogenic effects of 3alpha,5beta-THP: a possible model of premenstrual dysphoric disorder. Psychophar- macology (Berl) 186, 323-333. Soares C.N., Cohen L.S., Otto M.W., Harlow B.L., 2001. Characteristics of women with premenstrual dysphoric disorder (PMDD) who did or did not report history of depression: a preliminary report from the Harvard Study of Moods and Cycles. J. Womens Health Gend. Based Med. 10, 873-878. Soderpalm A.H., Lindsey S., Purdy R.H., Hauger R., Wit de H., 2004. Administration of progesterone produces mild sedative-like effects in men and women. Psychoneuroendocrinology 29, 339-354. Spielberger C.D., Gorsuch R.L., Lushene R.E., 1970. STAI manual for the State Trait Anxiety Inventory. Consulting Psychologists Press, Palo Alto. Spigelman I., Li Z., Liang J., Cagetti E., Samzadeh S., Mihalek R.M., Homanics G.E., Olsen R.W., 2003. Reduced inhibition and sensitivity to neurosteroids in hippocampus of mice lacking the GABA(A) receptor delta subunit. J. Neurophysiol. 90, 903-910. Spigset O., 1997. The selective serotonin reuptake inhibitor fluvoxamine: aspects on pharmacokinetics, pharma- codynamics and adverse effects in healthy volunteers. Umeå University, Umeå. Spivak B., Maayan R., Kotler M., Mester R., Gil-Ad I., Shtaif B., Weizman A., 2000. Elevated circulatory level of GABA(A)--antagonistic neurosteroids in patients with combat-related post-traumatic stress disorder. Psychol. Med. 30, 1227-1231. Steiner M., Dunn E., Born L., 2003. Hormones and mood: from menarche to menopause and beyond. J. Affect. Disord. 74, 67-83. Stewart R.A., Sharples K.J., North F.M., Menkes D.B., Baker J., Simes J., 2000. Long-term assessment of psycho- logical well-being in a randomized placebo-controlled trial of cholesterol reduction with pravastatin. The LIPID Study Investigators. Arch. Intern. Med. 160, 3144-3152. Stoffel-Wagner B., 2001. Neurosteroid metabolism in the human brain. Eur. J. Endocrinol. 145, 669-679. Strohle A., Pasini A., Romeo E., Hermann B., Spalletta G., di Michele F., Holsboer F., Rupprecht R., 2000. Fluox- etine decreases concentrations of 3 alpha, 5 alpha-tetrahydrodeoxycorticosterone (THDOC) in major depression. J. Psychiatr. Res. 34, 183-186. Strohle A., Romeo E., di Michele F., Pasini A., Hermann B., Gajewsky G., Holsboer F., Rupprecht R., 2003. Induced panic attacks shift gamma-aminobutyric acid type A receptor modulatory neuroactive steroid composi- tion in patients with panic disorder: preliminary results. Arch. Gen. Psychiatry 60, 161-168.

174 Chapter 12 | References

Strohle A., Romeo E., di Michele F., Pasini A., Yassouridis A., Holsboer F., Rupprecht R., 2002. GABA(A) receptor- modulating neuroactive steroid composition in patients with panic disorder before and during paroxet- ine treatment. Am. J. Psychiatry 159, 145-147. Strohle A., Romeo E., Hermann B., Pasini A., Spalletta G., di Michele F., Holsboer F., Rupprecht R., 1999. Concen- trations of 3 alpha-reduced neuroactive steroids and their precursors in plasma of patients with major depression and after clinical recovery. Biol. Psychiatry 45, 274-277. Stromstedt M., Warner M., Banner C.D., MacDonald P.C., Gustafsson J.A., 1993. Role of brain cytochrome P450 in regulation of the level of anesthetic steroids in the brain. Mol Pharmacol 44, 1077-1083. Strous R.D., Spivak B., Yoran-Hegesh R., Maayan R., Averbuch E., Kotler M., Mester R., Weizman A., 2001. Analysis of neurosteroid levels in attention deficit hyperactivity disorder. Int. J. Neuropsychopharmacol. 4, 259- 264. Sundstrom I., Andersson A., Nyberg S., Ashbrook D., Purdy R.H., Backstrom T., 1998. Patients with premenstrual syndrome have a different sensitivity to a neuroactive steroid during the menstrual cycle compared to control subjects. Neuroendocrinology 67, 126-138. Sundstrom I., Andersson A., Nyberg S., Ashbrook D., Purdy R.H., Bäckström T., 1998. Patients with premenstrual syndrome have a different sensitivity to a neuroactive steroid during the menstrual cycle compared to control subjects. Neuroendocrinology 67, 126-138. Sundstrom I., Ashbrook D., Backstrom T., 1997a. Reduced benzodiazepine sensitivity in patients with premen- strual syndrome: a pilot study. Psychoneuroendocrinology 22, 25-38. Sundstrom I., Backstrom T., 1998. Citalopram increases pregnanolone sensitivity in patients with premenstrual syndrome: an open trial. Psychoneuroendocrinology 23, 73-88. Sundstrom I., Backstrom T., 1999. Endocrine response to pregnanolone. Gynecol. Endocrinol. 13, 352-360. Sundstrom I., Nyberg S., Backstrom T., 1997b. Patients with premenstrual syndrome have reduced sensitivity to midazolam compared to control subjects. Neuropsychopharmacology 17, 370-381. Sundstrom I., Spigset O., Andersson A., Appelblad P., Backstrom T., 1999. Lack of influence of menstrual cycle and premenstrual syndrome diagnosis on pregnanolone pharmacokinetics. Eur. J. Clin. Pharmacol. 55, 125-130. Svec F., Porter J., 1997. The effect of dehydroepiandrosterone (DHEA) on Zucker rat food selection and hypotha- lamic neurotransmitters. Psychoneuroendocrinology 22 Suppl 1, S57-62. Sveindottir H., Backstrom T., 2000. Prevalence of menstrual cycle symptom cyclicity and premenstrual dys- phoric disorder in a random sample of women using and not using oral contraceptives. Acta Obstet. Gynecol. Scand. 79, 405-413. Takebayashi M., Kagaya A., Uchitomi Y., Kugaya A., Muraoka M., Yokota N., Horiguchi J., Yamawaki S., 1998. Plasma dehydroepiandrosterone sulfate in unipolar major depression. Short communication. J Neural Transm 105, 537-542.

175 Chapter 12 | References

Thomas G., Frenoy N., Legrain S., Sebag-Lanoe R., Baulieu E.E., Debuire B., 1994. Serum dehydroepiandrosterone sulfate levels as an individual marker. J. Clin. Endocrinol. Metab. 79, 1273-1276. Timby E., Balgard M., Nyberg S., Spigset O., Andersson A., Porankiewicz-Asplund J., Purdy R.H., Zhu D., Back- strom T., Poromaa I.S., 2006. Pharmacokinetic and behavioral effects of allopregnanolone in healthy women. Psychopharmacology (Berl) 186, 414-424. Tollefson G.D., Haus E., Garvey M.J., Evans M., Tuason V.B., 1990. 24 Hour urinary dehydroepiandrosterone sulfate in unipolar depression treated with cognitive and/or pharmacotherapy. Ann Clin Psychiatry 2, 39-45. Tottori K., Miwa T., Uwahodo Y., Yamada S., Nakai M., Oshiro Y., Kikuchi T., Altar C.A., 2001. Antidepressant-like responses to the combined sigma and 5-HT1A receptor agonist OPC-14523. Neuropharmacology 41, 976- 988. Tsao D.Y., Freiwald W.A., Tootell R.B., Livingstone M.S., 2006. A cortical region consisting entirely of face-selec- tive cells. Science 311, 670-674. Turkmen S., Lundgren P., Birzniece V., Zingmark E., Backstrom T., Johansson I.M., 2004. 3beta-20beta-dihydroxy- 5alpha-pregnane (UC1011) antagonism of the GABA potentiation and the learning impairment induced in rats by allopregnanolone. Eur. J. Neurosci. 20, 1604-1612. Twyman R.E., Macdonald R.L., 1992. Neurosteroid regulation of GABAA receptor single-channel kinetic proper- ties of mouse spinal cord neurons in culture. J. Physiol. 456, 215-245. Ukai M., Maeda H., Nanya Y., Kameyama T., Matsuno K., 1998. Beneficial effects of acute and repeated adminis- trations of sigma receptor agonists on behavioral despair in mice exposed to tail suspension. Pharmacol. Biochem. Behav. 61, 247-252. Urani A., Privat A., Maurice T., 1998. The modulation by neurosteroids of the scopolamine-induced learning im- pairment in mice involves an interaction with sigma1 (sigma1) receptors. Brain Res. 799, 64-77. Urani A., Roman F.J., Phan V.L., Su T.P., Maurice T., 2001. The antidepressant-like effect induced by sigma(1)-re- ceptor agonists and neuroactive steroids in mice submitted to the forced swimming test. J. Pharmacol. Exp. Ther. 298, 1269-1279. Uzunov D.P., Cooper T.B., Costa E., Guidotti A., 1996. Fluoxetine-elicited changes in brain neurosteroid content measured by negative ion mass fragmentography. Proc. Natl. Acad. Sci. U. S. A. 93, 12599-12604. Uzunova V., Ceci M., Kohler C., Uzunov D.P., Wrynn A.S., 2003. Region-specific dysregulation of allopregnanolone brain content in the olfactory bulbectomized rat model of depression. Brain Res. 976, 1-8. Uzunova V., Sheline Y., Davis J.M., Rasmusson A., Uzunov D.P., Costa E., Guidotti A., 1998. Increase in the cer- ebrospinal fluid content of neurosteroids in patients with unipolar major depression who are receiving fluoxetine or fluvoxamine. Proc. Natl. Acad. Sci. U. S. A. 95, 3239-3244. Uzunova V., Wrynn A.S., Kinnunen A., Ceci M., Kohler C., Uzunov D.P., 2004. Chronic antidepressants reverse cerebrocortical allopregnanolone decline in the olfactory-bulbectomized rat. Eur. J. Pharmacol. 486, 31-34.

176 Chapter 12 | References

van Broekhoven F., Backstrom T., Verkes R.J., 2006. Oral progesterone decreases saccadic eye velocity and increases sedation in women. Psychoneuroendocrinology van Broekhoven F., Kan C.C., Zitman F.G., 2002. Dependence potential of antidepressants compared to benzodi- azepines. Prog. Neuropsychopharmacol. Biol. Psychiatry 26, 939-943. van Broekhoven F., Verkes R.J., 2003. Neurosteroids in depression: a review. Psychopharmacology (Berl) 165, 97-110. van Rood Y., Goulmy E., Blokland E., Pool J., van Rood J., van Houwelingen H., 1991. Month-related variability in immunological test results; implications for immunological follow-up studies. Clin Exp Immunol 86, 349-354. van Steveninck A.L., van Berckel B.N., Schoemaker R.C., Breimer D.D., van Gerven J.M., Cohen A.F., 1999. The sensitivity of pharmacodynamic tests for the central nervous system effects of drugs on the effects of sleep deprivation. J. Psychopharmacol. 13, 10-17. van Wingen G., van Broekhoven F., Verkes R.J., Petersson K.M., Backstrom T., Buitelaar J., Fernandez G., 2007. How progesterone impairs memory for biologically salient stimuli in healthy young women. J. Neurosci. 27, 11416-11423. Vaswani M., Linda F.K., Ramesh S., 2003. Role of selective serotonin reuptake inhibitors in psychiatric disorders: a comprehensive review. Prog. Neuropsychopharmacol. Biol. Psychiatry 27, 85-102. Vuilleumier P., Richardson M.P., Armony J.L., Driver J., Dolan R.J., 2004. Distant influences of amygdala lesion on visual cortical activation during emotional face processing. Nat Neurosci 7, 1271-1278. Walker J.M., Bowen W.D., Walker F.O., Matsumoto R.R., De Costa B., Rice K.C., 1990. Sigma receptors: biology and function. Pharmacol. Rev. 42, 355-402. Wang M., Seippel L., Purdy R.H., Backstrom T., 1996. Relationship between symptom severity and steroid vari- ation in women with premenstrual syndrome: study on serum pregnenolone, , 5 alpha-pregnane-3,20-dione and 3 alpha-hydroxy-5 alpha-pregnan-20-one. J Clin Endocrinol Metab 81, 1076-1082. Weber B., Lewicka S., Deuschle M., Colla M., Heuser I., 2000. Testosterone, and dihydrotesto- sterone concentrations are elevated in female patients with major depression. Psychoneuroendocrinol- ogy 25, 765-771. Wen S., Dong K., Onolfo J.P., Vincens M., 2001. Treatment with dehydroepiandrosterone sulfate increases NMDA receptors in hippocampus and cortex. Eur. J. Pharmacol. 430, 373-374. West L.A., Horvat R.D., Roess D.A., Barisas B.G., Juengel J.L., Niswender G.D., 2001. Steroidogenic acute regula- tory protein and peripheral-type benzodiazepine receptor associate at the mitochondrial membrane. Endocrinology 142, 502-505. Wetzel C.H., Hermann B., Behl C., Pestel E., Rammes G., Zieglgansberger W., Holsboer F., Rupprecht R., 1998. Functional antagonism of gonadal steroids at the 5-hydroxytryptamine type 3 receptor. Mol Endocrinol 12, 1441-1451.

177 Chapter 12 | References

Wible J.H., Jr., Zerbe R.L., DiMicco J.A., 1985. Benzodiazepine receptors modulate circulating plasma vasopressin concentration. Brain Res. 359, 368-370. Wihlback A.C., Sundstrom-Poromaa I., Backstrom T., 2006. Action by and sensitivity to neuroactive steroids in menstrual cycle related CNS disorders. Psychopharmacology (Berl) 186, 388-401. Wikander I., Sundblad C., Andersch B., Dagnell I., Zylberstein D., Bengtsson F., Eriksson E., 1998. Citalopram in premenstrual dysphoria: is intermittent treatment during luteal phases more effective than continuous medication throughout the menstrual cycle? J. Clin. Psychopharmacol. 18, 390-398. Wolf O.T., Kirschbaum C., 1999. Actions of dehydroepiandrosterone and its sulfate in the central nervous system: effects on cognition and emotion in animals and humans. Brain Res. Brain Res. Rev. 30, 264-288. Wolf O.T., Koster B., Kirschbaum C., Pietrowsky R., Kern W., Hellhammer D.H., Born J., Fehm H.L., 1997. A single administration of dehydroepiandrosterone does not enhance memory performance in young healthy adults, but immediately reduces cortisol levels. Biol. Psychiatry 42, 845-848. Wolkowitz O., Reus V.I., Manfredi F., Roberts E., 1992. Antiglucocorticoid effects of DHEA-S in Alzheimer’s disease (reply). Am. J. Psychiatry 149, 1126. Wolkowitz O.M., 1994. Prospective controlled studies of the behavioral and biological effects of exogenous cor- ticosteroids. Psychoneuroendocrinology 19, 233-255. Wolkowitz O.M., Reus V.I., 1999. Treatment of depression with antiglucocorticoid drugs. Psychosom. Med. 61, 698-711. Wolkowitz O.M., Reus V.I., Keebler A., Nelson N., Friedland M., Brizendine L., Roberts E., 1999. Double-blind treat- ment of major depression with dehydroepiandrosterone. Am. J. Psychiatry 156, 646-649. Wolkowitz O.M., Reus V.I., Roberts E., Manfredi F., Chan T., Ormiston S., Johnson R., Canick J., Brizendine L., We- ingartner H., 1995. Antidepressant and cognition-enhancing effects of DHEA in major depression. Ann. N. Y. Acad. Sci. 774, 337-339. Wolkowitz O.M., Reus V.I., Roberts E., Manfredi F., Chan T., Raum W.J., Ormiston S., Johnson R., Canick J., Brizen- dine L., Weingartner H., 1997. Dehydroepiandrosterone (DHEA) treatment of depression. Biol. Psychiatry 41, 311-318. World Health Organization, 2004, 2nd edition. ICD-10: international statistical classification of diseases and related health problems: tenth revision. WHO, Geneva. Worsley K.J., Marrett S., Neelin P., Vandal A.C., Friston K.J., Evans A.C., 1996. A unified statistical approach for determining significant signals in images of cerebral activation. Human Brain Mapping 4, 58-73. Wurtman R.J., 1983. Behavioural effects of nutrients. Lancet 1, 1145-1147. Xiong G.L., Benson A., Doraiswamy P.M., 2005. Statins and cognition: what can we learn from existing rand- omized trials? CNS Spectr. 10, 867-874. Yang C.C., Jick S.S., Jick H., 2003. Lipid-lowering drugs and the risk of depression and suicidal behavior. Arch. Intern. Med. 163, 1926-1932.

178 Chapter 12 | References

Young-Xu Y., Chan K.A., Liao J.K., Ravid S., Blatt C.M., 2003. Long-term statin use and psychological well-being. J. Am. Coll. Cardiol. 42, 690-697. Young J., Corpechot C., Haug M., Gobaille S., Baulieu E.E., Robel P., 1991. Suppressive effects of dehydroepi- androsterone and 3 beta-methyl-androst-5-en-17-one on attack towards lactating female intruders by castrated male mice. II. Brain neurosteroids. Biochem Biophys Res Commun 174, 892-897. Yu R., Ticku M.K., 1995. Chronic neurosteroid treatment decreases the efficacy of benzodiazepine ligands and neurosteroids at the gamma-aminobutyric acidA receptor complex in mammalian cortical neurons. J. Pharmacol. Exp. Ther. 275, 784-789. Zhu D., Birzniece V., Backstrom T., Wahlstrom G., 2004. Dynamic aspects of acute tolerance to allopregnanolone evaluated using anaesthesia threshold in male rats. Br. J. Anaesth. 93, 560-567. Zou L.B., Yamada K., Sasa M., Nakata Y., Nabeshima T., 2000. Effects of sigma(1) receptor agonist SA4503 and neuroactive steroids on performance in a radial arm maze task in rats. Neuropharmacology 39, 1617- 1627.

179 Chapter 12 | References

samenvatting

180 samenvatting samenvatting

In dit proefschrift is het effect van het neuroactieve steroid allopregnanolon onderzocht. Allopregna- nolon is een endogene metaboliet van progesteron en een positieve allosterische modulator van de GABA(A) receptor. Het wordt een neuroactieve steroid genoemd omdat het de exciteerbaarheid moduleert van mem- braangebonden neurotransmitterreceptoren. De beschikbare literatuur refereert voornamelijk naar dierex- perimenteel onderzoek, maar in dit proefschrift zijn mensen onderzocht. In dit hoofdstuk worden de resulta- ten samengevat.

Hoofdstukken 1 - 3

In deze hoofdstukken wordt een algemene introductie op het concept neuroactieve steroiden gegeven, naast farmacologische gegevens over progesteron en allopregnanolon en een beschrijving van het serotonine- systeem.

Hoofdstuk 4

In dit hoofdstuk wordt de beschikbare literatuur over de rol van neuroactieve steroiden bij depressie beschouwd.

Allopregnanolone is een positieve allosterische modulator van de GABA(A) receptor en heeft anxiolytische, sedatieve en anticonvulsieve eigenschappen. Hoewel de concentraties allopregnanolone in plasma en liquor verlaagd zijn bij depressieve patienten en normaliseren na behandeling met psychofarmaca, is het niet duidelijk of deze normalisatie slechts een epifenomeen van de psychofarmacologische behandeling is of dat het een noodzakelijke voorwaarde is voor genezing. Onderzoek bij patienten die werden behandeld met niet-psychofarmacologische methoden, zoals slaapdeprivatie, repetitieve transcraniele magnetische stimulatie of electroconvulsieve therapie, is daarom nodig.

Aan allopregnanolon wordt een antidepressief effect toegedicht onder andere omdat het de immobilisatietijd ver- kort in de gedwongen zwemtest, een diermodel van depressie. Als verklaring voor het antidepressieve effect wordt een negatieve feedback op de HPA-as en een stimulerend effect op de serotonine en noradrenalinehuishouding in de synapsen aangevoerd. Selectieve serotonineheropnameremmers verhogen de allopregnanolonconcetratie in de hersenen van dieren en dit wordt als verklaring gezien van het snelle positieve effect van deze klasse medicijnen bij het prementrueel dysfoor syndroom. Het neuroactieve steroid dehydroepiandrosteron (DHEA) heeft een positief ef- fect op de cognitieve functies in dierenstudies maar dit heeft men bij mensen niet kunnen aantonen. DHEA heeft ook antidepressieve effecten en klinische trials bij mensen laten positieve resultaten zien.

183 Samenvatting

Hoofdstuk 5

Hoewel de toename van allopregnanolon na acute stress bij dieren herhaaldelijk is aangetoond, ontbreken gegevens hieromtrent bij mensen. Daarom hebben we een nieuw stressmodel voor mensen toegepast, de ver- dediging van iemands proefschrift tegenover een college van hoogleraren en gepromoveerd deskundigen, en daarbij de spiegels allopregnanolon in bloed gemeten. Het bleek dat ook bij mensen (n = 15) acute stress leidt tot een verhoging in allopregnanolonconcentratie. Daarnaast werd ook de perifere benzodiazepinereceptor(PBR)d ichtheid gemeten. De perifere benzodiazepinereceptor is belangrijk in de biosynthese van allopregnanolon. Dit werd gemeten omdat het aangetoond is dat PBR dichtheid toeneemt na acute stress maar het niet bekend is of er een correlatie bestaat tussen toegenomen allopregnanolonplasma spiegels en PBR dichtheid. Wij vonden in onze studie een correlatie, zij het een kleine. Omdat de vitale functies bloeddruk en hartfrequentie toenamen sa- men met de algemeen erkende endocrinologische stressparameter cortisol, kunnen we aannemen dat het door ons toegepaste model van acute stress bij mensen valide is. De resultaten geven ondersteuning aan het idee om een specifieke PBR ligand te ontwikkelen dat de synthese van anxiolytische neuroactieve steroiden reguleert.

Hoofdstuk 6

De huidige gegevens afkomstig van mensgebonden onderzoek zijn beperkt tot afname van bloedmonsters en de toediening van de precursor van allopregnanolon, progesteron. Wij dienden een nieuwe farmaceutische formulering voor intraveneuze toediening toe. Negen gezonde mannen en negen gezonde vrouwen kregen al- lopregnanolon toegediend in drie, in sterkte toenemende, intraveneuze doses. Saccadische oogsnelheid (SEV), subjectieve scores (visueel analoge schalen), en electroencefalografie (EEG) werden gebruikt om het effect van allopregnanolon te meten. Er werden over de dag herhaaldelijk bloedmonsters afgenomen om allopregnano- lonconcentraties te kunnen bepalen. Allopregnanolon zorgde voor een grotere afname in SEV bij vrouwen dan bij mannen, en verhoogde de subjectieve sedatiescore. Subjectieve scores van tevredenheid verminderden bij mannen maar namen toe bij vrouwen. De concentraties allopregnanolon in bloed waren bij mannen meer toege- nomen dan bij vrouwen. Overige farmacokinetische parameters waren niet verschillend tussen beide sexen. Bij mannen nam de beta power op het EEG toe. De conclusie is dat mannen en vrouwen verschillend reageren op allopregnanolon.

Hoofdstuk 7

Deze studie werd uitgevoerd om het effect van oraal progesteron en de metaboliet allopregnanolon bij vrouwen te onderzoeken. Vrouwen (n = 15) namen in hun folliculaire fase progesteron (400 mg) of placebo

184 Samenvatting

oraal in. Afhankelijke maten waren de concentraties in bloed van progesteron en allopregnanolon, saccadische oogsnelheid (SEV), subjectieve scores (visueel analoge schalen), en reactietijd. Progesteron verminderde SEV en vergrootte sedatie. Dit effect is waarschijnlijk te wijten aan een verhoogde GABA activiteit.

Hoofdstuk 8

Kennis van de mate van sensitiviteit voor progesteron en allopregnanolon bij vrouwen gedurende de cy- clus is nodig om de pathofysiologie van het premenstrueel dysfoor syndroom te kunnen begrijpen. Daarom heb- ben we onderzocht of effect van een enkele dosis progesteron op de GABA(A) receptor sensitiviteit en stemming verschilt tussen de folliculaire en luteale fase van de normale cyclus. Vrouwen (n = 9) ontvingen oraal proges- teron (400 mg) of placebo tijdens de folliculaire en luteale fase in een dubbelblind, gerandomiseerd, crossover design. Afhankelijke maten waren saccadische oogsnelheid (SEV), subjectieve scores van sedatie, tevredenheid en kalmte (visueel analoge schalen), en reactietijden op een eenvoudige en complexe aandachtstaak. Toediening van progesteron verlaagde SEV en vergrootte sedatie meer in de luteale dan in de folliculaire fase. Complexe reactietijd was meer toegenomen in de luteale dan in de folliculaire fase na progesteron maar dit verschil was niet statistisch significant. De effecten zijn waarschijnlijk toe te schrijven aan een verhoogde GABA activiteit veroorzaakt door de progesteronmetaboliet allopregnanolon. De conclusie is dat het effect van orale progeste- ron op de GABA(A) receptor sensitiviteit zoals gemeten met SEV afhankelijk is van de fase van de menstruele cyclus.

Hoofdstuk 9

Wij onderzochten placebogecontroleerd, met functionele magnetische resonantiebeeldvorming, of een enkele dosis progesteron bij jonge, gezonde vrouwen (n = 18) in de folliculaire fase het effect van biologisch relevante stimuli op de amygdala moduleert. De orale progesterondosis (400 mg) deed de plasmaconcentraties van progesteron en allopregnanolon toenemen tot het niveau dat bereikt wordt tijdens de luteale fase en vroege zwangerschap. De resultaten uit de beeldvorming laten zien dat progesteron selectief de amygdala activiteit deed toenemen. Het wordt verondersteld dat de toegenomen activiteit te wijten is aan een disinhibitie van in- hiberende interneuronen. Verder wijzen functionele connectiviteitsanalyses er op dat progesteron functionele koppeling van de amygdala met verder weg gelegen hersenregio’s moduleerde. Deze resultaten laten een neu- raal mechanisme zien waarmee progesteron mogelijk negatieve effecten op angst en stemming medieert.

185 Samenvatting

Publications

186 Publications publications

JOURNAL ARTICLES AND BOOK CHAPTERS

Rostami K, Van Bergeijk L, Van Broekhoven F, Janssen BPW, Mulder CJJ. Elevation of pancreatic enzymes: a marker for intracranial hemorrhage. Romanian Journal of Gastroenterology (2001) 10:61-63.

Van Broekhoven F, Kan CC, Zitman FG. Dependence potential of antidepressants compared to benzodiazepines. Progress in Neuro-Psychopharmacology & Biological Psychiatry (2002) 26:939–943.

Van Broekhoven F, Verkes RJ, Janzing JGE. Psychiatric symptoms during isotretinoin therapy. Ned. Tijdschr. Geneeskd. (2003) 147:2341-2343.

Van Broekhoven F and Verkes RJ. Neurosteroids in depression: a review. Psychopharmacology (2003) 165:97–110.

Droogleever Fortuyn HA, Van Broekhoven F, Span PN, Bäckström T, Zitman FG, Verkes RJ. Effects of PhD examination stress on allopregnanolone and cortisol plasma levels and peripheral benzodiazepine receptor density. Psychoneuroen- docrinology (2004) 29:1341–1344.

Van Broekhoven F, Bäckström T, Verkes RJ. Oral progesterone decreases saccadic eye velocity and increases sedation in women. Psychoneuroendocrinology (2006) 31:1190-1199.

Van Broekhoven F, Bäckström T, Van Luijtelaar G, Buitelaar JK, Smits P, Verkes RJ. Effects of allopregnanolone on seda- tion in men, and in women on oral contraceptives. Psychoneuroendocrinology (2007) 32:555-564.

Van Wingen GA, van Broekhoven F, Verkes RJ, Petersson KM, Bäckström T, Buitelaar JK, Fernández G. Progesterone selectively increases amygdala reactivity in women. Mol. Psychiatry (2008) 13:325-333.

Van Wingen GA, van Broekhoven F, Verkes RJ, Petersson KM, Bäckström T, Buitelaar JK, Fernández G. How progesterone impairs memory for biologically salient stimuli in healthy young women. J. Neurosci. (2007) 27:11416-11423.

Van Broekhoven F, Bäckström T, Buitelaar JK, Verkes RJ. Increased sensitivity to oral progesterone in the luteal phase in healthy women (submitted).

189 Publications

Bäckström T, et al. Neuroactive steroids: effects on cognitive functions. In: Michael S. Ritsner and Abraham Weizman (Eds.) Neuroactive Steroids in Brain Function, Behavior and Neuropsychiatric Disorders. Novel Strategies for Research and Treatment. Springer, New York, 2008, pp.103-121.

Bäckström T, et al. Neuroactive steroids in brain and relevance to mood. In: Michael S. Ritsner and Abraham Weizman (Eds.) Neuroactive Steroids in Brain Function, Behavior and Neuropsychiatric Disorders. Novel Strategies for Research and Treatment. Springer, New York, 2008, pp.423-433.

190 Publications

ABSTRACTS

Van Broekhoven F, Kan CC, Zitman FG. Afhankelijkheid van antidepressiva. Oral presentation at the FADO congres, November 13, 2001, Utrecht, the Netherlands.

Van Broekhoven F, Kan CC, Zitman FG. Antidepressiva versus benzodiazepine dependence. Oral presentation at the Voorjaarscongres NVvP, April 3, 2002, Groningen, the Netherlands.

Droogleever Fortuyn HA, Van Broekhoven F, Span PN, Bäckström T, Zitman FG, Verkes RJ. Effects of PhD examination stress on peripheral benzodiazepine receptors. European Neuropsychopharmacology (2002) 12(suppl 3):S355. Poster presentation at the 15th Congress of the European College of Neuropsychopharmacology, Barcelona, Spain, October 5-9, 2002.

Droogleever Fortuyn HA, Van Broekhoven F, Span PN, Bäckström T, Zitman FG, Verkes RJ. Effects of PhD examination stress on allopregnanolone and cortisol plasma levels and peripheral benzodiazepine recep- tor density. Oral presentation at the 2nd International Meeting on Steroids and Nervous System, February 23, 2003, Torino, Italy.

Droogleever Fortuyn HA, Van Broekhoven F, Span PN, Bäckström T, Zitman FG, Verkes RJ. Effects of PhD examination stress on allopregnanolone and cortisol plasma levels and peripheral benzodiazepine receptor density. Poster presenta- tion at the 2nd Dutch Endo-Neuro-Psycho Meeting, Doorwerth, The Netherlands, June 3, 2003.

Van Broekhoven F and Verkes RJ. Neurosteroids in mood disorders. Oral presentation at the Onderzoekersdagen Psychiatrie, June 13, 2003, Zandvoort, the Netherlands.

Van Broekhoven F, Bäckström T, Sundström I, Van Luijtelaar EL, Smits P, Verkes RJ. Intravenous infusion of allopregna- nolone induces sedation in humans. European Neuropsychopharmacology (2004) 14(suppl 3):S368. Poster presentation at the 17th Congress of the European College of Neuropsychopharmacology, Stockholm, Sweden, October 9-13, 2004.

Van Broekhoven F and Verkes RJ. Neuroactive steroids. Oral presentation at the WetenSAPdag NVvP, November 2, 2004, Utrecht, The Netherlands.

Van Broekhoven F, Bäckström T, Sundström I, van Luijtelaar EL, Smits P, Verkes RJ. Neuroactive steroids in mood and stress regulation. Oral presentation at the 4th Dutch Endo-Neuro-Psycho Meeting, June 3, 2005, Doorwerth, The Netherlands.

191 Publications

dankwoord

192 dankwoord dankwoord

Dit proefschrift zou nooit tot stand zijn gekomen zonder de medewerking van anderen. Op deze plaats wil ik hen daarvoor bedanken.

Prof.dr. J.K. Buitelaar, beste Jan, dank voor je heldere en scherpe blik, beslisvaardigheid en voortstuwende kracht. Het hielp mij door te gaan en niet alles “op de lange baan” te schuiven. Als opleider heb je mij tijdens mijn opleiding een aantal keren in de gelegenheid gesteld om in Umeå in alle rust, weg uit de drukte van de kliniek, aan mijn onderzoek te werken. Dat heb ik zeer gewaardeerd. Prof.dr. T. Bäckström, dear Torbjörn, thank you very much for spending so much time with me at your lab and leading me the way into the field of neuroactive steroids. I have appreciated your hospitality during my stays in Umeå and the sti- pendium that you obtained for me. Your patience, enthusiasm for science, and clear help with the processing of the data and the writing of the papers, have helped me a lot. Dr. R.J. Verkes, beste Robbert Jan, dank voor het mij er toe aanzetten (buitenlandse) congressen te bezoeken en daar posters te presenteren of orale presentatie’s te verzorgen. Zonder jouw hulp bij de statistische verwerking van de data, was dit proefschrift nooit tot stand gekomen. Dr. C.C. Kan, beste Cees, op meerdere momenten heb je een cruciale rol gespeeld in mijn medische carriere. Allereerst was er onze samenwerking tijdens jouw promotie-onderzoek. Daarna was je mijn begeleider van de wetenschappelijke stage dat uitmondde in een gezamenlijke publicatie. Tijdens jouw promotieplechtigheid was ik paranimf. En later was je één van mijn supervisoren tijdens de opleiding tot psychiater. Prof.dr. R.J. van der Gaag, beste Rutger-Jan, hartelijk dank voor het faciliteren van de afronding van mijn proefschrift aan het eind van de rit.

Ik ben ook de proefpersonen dankbaar voor het feit dat zij bereid waren zich te onderwerpen aan de soms lange en vermoeiende onderzoeksdagen en voor hun flexibiliteit ten aanzien van de planning van deze dagen. Arif Wahaj en Marieke Meeuwis, dank voor de samenwerking tijdens jullie wetenschappelijke stages bij mij. Miriam Kos, dank je wel voor het bewerken van de EEG-data. Guido van Wingen, dank voor de samenwerking tijdens onze fMRI-metingen. Dank zij jou werd ik een beetje wegwijs in deze voor mij tot dan toe onbekende wereld van de neuro-imaging. Collega’s van de opleiding, dank voor jullie belangstelling en waarnemingen tijdens mijn perioden van afwezigheid.

195 Dankwoord

Lieve ouders, ik dank jullie voor alle bagage die ik van huis uit mee kreeg en jullie onvoorwaardelijke belangstelling. curriculum vitae Dankzij de kansen die jullie mij gaven heb ik kunnen doen wat ik wilde. Ik prijs me gelukkig met het feit dat ik mijn ervaringen steeds met jullie heb kunnen delen. Lieve Ellen, aan jou ben ik de meeste dank verschuldigd. Want wat heb jij veel alleen moeten doen in de opvoeding van de kinderen. Denk hierbij vooral aan de keren dat ik in Zweden verbleef om aan mijn onderzoek te werken. Het was een zeer drukke en soms erg vermoeiende tijd voor ons allebei, niet alleen vanwege mijn onderzoek, maar juist ook vanwege het feit dat ik dat combineerde met mijn opleiding tot psychiater en we in deze tijd ook onze kinderen mochten krijgen. Maar het was ook een afwisselende tijd waarin we samen veel genoten hebben. Vincent, Heleen en Maartje, ik hou heel veel van jullie en hoewel ik heb geprobeerd het werken aan mijn proefschrift zo veel mogelijk te doen wanneer jullie in bed lagen, is dat, zoals jullie weten, lang niet altijd gelukt. Soms zat ik dus ook overdag achter mijn laptop en kon ik niet met jullie spelen. Dank jullie wel voor het zo goed mogelijk proberen mij op die momenten met rust te laten.

196 curriculum vitae

Curriculum vitae

curriculum vitae

Frank van Broekhoven wordt geboren op 8 december 1969 in Groenlo. Hij groeit op in Eibergen. In 1988 behaalt hij zijn VWO diploma aan scholengemeenschap Marianum in Groenlo en begint vervolgens aan de studie fysiotherapie in Arnhem. In 1993 behaalt hij daarvan het diploma. Daarna start hij met de studie geneeskunde aan de Katholieke Universiteit Nijmegen (thans: Radboud Universiteit Nijmegen). De wetenschappelijke stage volgt hij op de afdeling Psychiatrie van het UMC St Radboud onder begeleiding van dr. C.C. Kan. Het betreft onderzoek naar de mogelijke afhankelijkheid van antidepressiva. In 1999 behaalt hij zijn arts-exa- men. Aansluitend is hij als arts-assistent-niet-in-opleiding gedurende ruim een jaar werkzaam in het Rijnstate ziekenhuis in Arnhem op de afdelingen Inwendige Geneeskunde, Longziekten, Cardiologie, Spoedeisende Hulp en Psychiatrie. Vanaf november 2000 verricht hij aan de afdeling Psychiatrie van het UMC St Radboud het hier beschreven wetenschappelijk onderzoek gericht op neuroactieve steroiden, in het bijzonder allopregnanolon (promotores: prof.dr. J.K. Buitelaar en prof. dr. T. Bäckström; co-promotor dr. R.J. Verkes). Vanaf september 2003 tot maart 2008 volgt hij tevens de opleiding tot psy- chiater, waarvan hij het basisgedeelte doorloopt aan de afdeling Psychiatrie van het UMC St Radboud (opleider: prof.dr. J.K. Buitelaar) en de stage Sociale Psychiatrie aan de GGz Nijmegen (opleider: dr. H.J. Gijsman). Thans is hij als psychiater verbonden aan Karakter Universitair Centrum voor Kinder- en Jeugdpsychiatrie in Nijme- gen, waar hij de tweejarige vervolgopleiding volgt voor de aantekening Kinder- en Jeugdpsychiatrie (opleider: prof.dr. R.J. van der Gaag). Frank van Broekhoven is sinds 1998 getrouwd met Ellen van Broekhoven-Grutters en heeft drie kinderen: Vincent (1999), Heleen (2001) en Maartje (2003).

199 Effects of progesterone and allopregnanolone on stress, attention, cognition and mood | Frank van Broekhoven

200