The Human Circadian Timing System in Aging, Alzheimer's Disease And

The human circadian timing system in aging, Alzheimer’s disease and Depression

THE HUMAN CIRCADIAN TIMING SYSTEM IN AGING, ALZHEIMER’S DISEASE AND DEPRESSION

ACADEMISCH PROEFSCHRIFT ter verkrijging van de graad van doctor aan de Universiteit van Amsterdam op gezag van de Rector Magniﬁcus prof. mr. P.F. van der Heijden ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit op dinsdag 30 januari 2007, te 12:00 uur door

Yinghui Wu

geboren te HuangShan, Volksrepubliek China PROMOTIECOMMISSIE

Promotores: Prof. dr. D.F. Swaab Prof. dr. J.N. Zhou

Overige leden: Prof. dr. R.M. Buijs Prof. dr. G.A. Kerkhof Prof. dr. W.A. van Gool Dr. M.P. Gerkema Dr. G.T. J. van der Horst

Faculteit der Geneeskunde

This investigation was carried out at the Netherlands Institute for Neuroscience with ﬁnancial support from Royal Netherlands Academy of Arts and Sciences (01CDP019, 02CDP014, 04CDP026) National Key project for Basic Science of China (G1999054007) Hersenstichting Nederland (11F03.07) Netherlands Organisation for Scientiﬁc Research (NWO)

Publication of this thesis was ﬁnancially supported by Netherlands Insitute for Neuroscience Van Leersumfonds KNAW University of Amsterdam Internationale Stichting Alzheimer Onderzoek (ISAO) Stichting Onderzoek Licht & Gezondheid (SOLG) J. E. Jurriaanse Stichting

Book design: Henk Stoﬀels, Amsterdam

Print: Gildeprint bv, Enschede Contents

Chapter 1 General introduction 9

Chapter 2 Molecular changes underlying reduced pineal melatonin levels 47 in Alzheimer’s disease: alterations in preclinical and clinical stages

Chapter 3 A promoter polymorphism in the monoamine oxidase A gene 65 is associated with the increased MAOA activity in Alzheimer’s disease

Chapter 4 Pineal clock gene oscillation is disturbed in Alzheimer’s disease, 77 due to functional disconnection from the “master clock”

Chapter 5 Distribution of melatonin receptor MT1 immunoreactivity in 95 the human hypothalamus and pituitary gland: colocalization of MT1 with vasopressin and corticotropin-releasing hormone

Chapter 6 Decreased MT1 melatonin receptor expression in the 117 suprachiasmatic nucleus in aging and Alzheimer’s disease

Chapter 7 Increased number of neurons expressing melatonin receptor 133 MT1 in the suprachiasmatic nucleus in depression, and its relation to age at onset and disease duration

Chapter 8 General discussion 147

References 175

Summary 210

Nederlandse samenvatting 215

List of publications 220

Acknowledgements 221

List of abbreviations 222

Curriculum vitae 224

Chapter 1

General introduction

Content

A. The circadian timing system and its molecular basis: changes in aging, Alzheimer’s disease, 9 depression and other circadian rhythm disorders

B. The human pineal gland and melatonin in aging and Alzheimer’s disease 28

C. Melatonin receptors: localization, molecular biology and physiological signiﬁcance 36

D. Scope of the thesis 41

A. The circadian timing system and its molecular basis: changes in aging, Alzheimer’s disease, depression and other circadian rhythm disorders

Partly based upon the paper: Y-H Wu, D.F. Swaab, Sleep Medicine, accepted

The past decade has been an extremely exciting and fast paced time for circadian biology, in which the basic molecular mechanisms underlying the mammalian circadian clock were deﬁned. This paper reviews the circadian timing system and its molecular components that provide the mechanisms for the regulation of circadian rhythms in mammals including humans, and relates them to the circadian abnormalities in aging, Alzheimer’s disease, depression and other circadian rhythm-related diseases.

The circadian timing system Temporal variations in endocrine, metabolism, autonomic, pharmacokinetics, and in some aspects of disease show the pervasive inﬂuence of the circadian clock on human physiology and pathophysiology. These circadian rhythms, as they are called, are the external expression of an internal timing mechanism that measures daily time. Cir- cadian clocks are normally set or entrained by periodic environmental cues, with the daily light-dark cycle as the most pervasive and potent entraining stimulus in mammals. An entrained circadian clock ensures that rhythms in physiology and behavior are in tune with the 24-h day. Three major components are present in all circadian systems identiﬁed to date: (a) a light input pathway to a self-sustained master circadian pacemaker, (b) the circadian pacemaker, and (c) output pathways by which the circadian pacemaker regulates overt rhythms in biochemistry, physiology, and behavior throughout the organism 1.

9 chapter 1

Suprachiasmtic nucleus: the “central biological clock” In mammals, the endogenous biological circadian clock is located in the suprachiasmatic nucleus (SCN) of the anterior hypothalamus 2, 3. The SCN is situated just above the optic chiasm, on each side of the third ventricle 4, 5. SCN neurons are not homog- enous, but are classiﬁed according to their neuropeptide content 4, 6. A simpliﬁed view of the rodent SCN has a core of vasoactive intestinal peptide (VIP) expressing neurons and a surrounding shell of arginine vasopressin peptide (AVP) expressing neurons. Both parts of the SCN each contain approximately 10,000 tightly packed neurons in the mouse 4, and 16,000 neurons in the rat 5. In the ventrolateral part of the rodent SCN, neurons synthesize VIP, gastrin-releasing peptide (GRP) and/or peptide histidine isoleucine (PHI). In between these two cell populations, a distinct population of somatostatin (SOM)-positive neurons is found. Gamma-amino-butyric acid (GABA) is present throughout the SCN and colocalizes with the above-mentioned neurotransmitters. In addition, many more neurotransmitters have been reported to be present, such as galanin and antiotensin II (for review see 7). Animal experiments have shown that lesions of the SCN abolish the locomotor activity rhythm as well as other rhythms, while transplantation of a fetal SCN may restore circadian activity rhythms in such lesioned animals 7-13. Interestingly, a restored rhythm adapts to the circadian rhythm of the donor, not to that of the recipience 14. In both night-active (nocturnal) and day-active (diurnal) mammals, most SCN neurons are active during the daytime and inactive at night time 7, 15-17. The circadian rhythm in metabolic activity of the SCN is also similar in diurnal and nocturnal mammals, with a peak at midday 18. Humans and other mammals share a very similar anatomy and physiology of the circadian system, which strongly supports the critical role of the SCN in humans. The human SCN is also located in the anterior hypothalamus, bilaterally next to the third ventricle and on top of the optic chiasm, just as in other mammals (Fig. 1). The human SCN has similar structures and contains similar neurotransmitters AVP, VIP and GABA to the rodent 6, 19. The AVP and VIP neuropeptide show a clear circadian pattern in the human SCN 20, 21 (Fig. 2). The bilateral human SCN is about 1 mm3 in volume and contains about 100.000 neurons 22. The vasopressin subnucleus of the SCN has a volume of 0.25 mm3 and contains some 10.000 vasopressin neurons on each side 23. The human SCN additionally contains neurotensin as an important neurotransmitter 19. The projections from the retina to the human SCN and from the SCN to other brain regions of the hypothalamus in human are very similar to those in rat 24, 25. The circadian system is functionally similar in human and other mammals, and can be phase-shifted by light and melatonin, the most important day and night signals for the SCN, respectively 26. Finally, the importance of the SCN for circadian rhythms in humans is revealed by clinical observations. Pa- tients who had a lesion in the suprachiasmatic region in the hypothalamus, e.g. as the

10 general introduction

SCN

Fig. 1 The location of the suprachiasmatic nucleus (SCN) in the human brain.A. Midsagittal view of the human brain. B. a thionin-stained coronal section of the human brain.

11 chapter 1

A. 0.2 mm

III

Fig. 2 A. Distribution of arginine vasopressin (AVP)-expressing and vasoactive intestinal polypeptide (VIP)-expressing neurons in the human suprachiasmatic nucleus (SCN) in consecutive frontal sections. Drawing showing the distribution of the AVP neurons (dots) and VIP neurons (circles) through the mid-portion of the SCN. B. Circadian ﬂuctuations of the AVP and VIP-expressing cell numbers in the human SCN. The data is represented by mean ± S.E.M. (gray area indicates S.E.M.) III: third ventricle. Top is caudal, bottom is rostral. Modiﬁed from Hofman (2000). result of metastasis, indeed showed a decreased expression of vasopressin in the SCN and disturbed circadian rhythms 19, 27-29. In addition, in a patient with a hypothalamic astrocytoma that destroyed the SCN bilaterally, disruptions of the day/night rhythm of the wake/sleep pattern was observed 30.

Input to the SCN The endogenous biological clock, the SCN, can function autonomously, independent of any external time cues, but its intrinsic period is not exactly 24 h. The endogenous circadian clock is entrained to a 24-h environmental cycle by environmental cues, in particular light/dark cycles. Some blind people who lack the synchronizing input

12 general introduction

B. ) 12 3 10 x (

4 Vasopressin cell number 0 4 8 12 16 20 24 Time of day ( h ) 4 ) 3

10 3 x (

1 VIP cell number

0 4 8 12 16 20 24 Time of day ( h ) from light to the SCN frequently show an endogenous circadian rhythm (free-running rhythm) with a period longer than 24 hours. They often report sleeping problems and a periodical inability to adapt to the social day-night rhythm. Although lacking light input, which is the most important day signal for the SCN, their circadian rhythm could be entrained by melatonin, which is the night signal and input for the SCN, if it is taken every day at the preferred time of sleep 31, 32. Moreover, nonphotic social stimuli are also of utmost importance in the entrainment of the circadian rhythmicity in blind people 33. For general input pathways to the SCN (see Fig. 3).

Light In humans and other mammals, light entrainment of the circadian system relies on retinal photoreceptors. There are two modes of photoreception: 1. visual photoreception, which provides a spatio-temporal image of the environment; 2. nonvisual photoreception, which aﬀects the circadian system. The visual photoreceptors three-

13 chapter 1 cone system and the rods are not responsible for transmitting the light signal to the circadian system in human and rodents 34-36. A recently discovered distinct subset of intrinsically photosensitive retinal ganglion cells containing the novel photopigment melanopsin is sensitive to light and projects to the SCN 37-40. Furthermore, blue-green light (450-500 nm) is the most effective for phase-shifting the circadian rhythm 34, 35 and matches exactly the sensitive peak of melanopsin. Photic information received by the retina is projected to the hypothalamic SCN via the retinohypothalamic tract (RHT). The RHT is the principal pathway mediating the entraining effects of light on the circadian pacemaker, the SCN. In rat, the RHT was found to originate from a distinct subset of retinal neurons 41 that contain pituitary adenylate cyclase-activating polypeptide (PACAP) and co-store glutamate, which code chemically for “darkness” and “light” information, respectively 42. Light activates these cells that directly innervate the SCN via the photopigment melanopsin in the PACAP cells. PACAP interacts with glutamate signaling during the light-induced phase shift. The sensitivity, spectral tuning, and slow kinetics of this light response matched those of the photic entrainment mechanism 43-46. The human primary visual pathway is indeed also regulated according to longterm (15–20 min) light exposure through the action of non-rod, non-cone photoreceptors via a photopigment with the characteristics of opsin: vitamin A 47. Recently the human RHT was studied by postmortem tracing procedure using neurobiotin as a tracer 48. The RHT appeared to leave the optic chiasm and enter the hypothalamus both medially and laterally of the SCN. The density of the RHT fibers decreases from rostral to caudal48 . The RHT terminates predominantly in a zone of the SCN that contains VIP neurons 19, but do not only contact VIP but also neurotensin cells in the SCN. In addition some vasopressin cells are innervated by the RHT in the ventral part of the SCN. Only few projections to the dorsal part of the SCN and the ventral part of the anterior hypothalamus were found 48. Light has two effects on the SCN: it leads to 1) mediation of the excitation of SCN- neuronal activity and 2) a phase shift in the circadian rhythm of the SCN. Light stimulation of the retina results in direct secretion of the excitatory neurotransmitter glutamate from the RHT into the ventral VIP subnucleus of the SCN 49-51 and immediately stimulates SCN-neuronal activity in the majority of SCN neurons 52, leading to an immediate influence on the SCN output.

Melatonin feeds back on the SCN Melatonin is mainly produced in the pineal gland in a circadian manner, controlled by the SCN. Melatonin can feedback on the SCN and is involved in the regulation of the circadian rhythm system 1, 53. Melatonin appears to act on the SCN by two distinct mechanisms. Firstly, melatonin acutely inhibits SCN neuronal ﬁring 54, 55, which may be important for deﬁning the sensitivity of the SCN to phase-shifting stimuli, and in

14 general introduction

Temperature

Pineal Physical activity melatonin IGL

GHT

RGT GABA, NPY RGC RHT Bright light SCN Glu

ACH Forebrain 5-HT 5-HT NA brain stem

Septum DR/MR LC Somatosensory input

Fig. 3 Schematic and simplified overview of the inputs to the suprachiasmatic nucleus, and their interactions. Inputs are in outlined font, structures in bold, tracts in normal font and neurotransmitters and hormones in italics. Abbreviations: 5-HT, 5-hydroxytryptamine (serotonin); ACH: acetylcholine; DR, dorsal raphe nucleus; GABA, y-aminobutyric acid; GHT, geniculohypothalamic tract; Glu, glutamate; IGL, intergeniculate leaflet; LC, locus coeruleus; MR, median raphe nucleus; NA, noradrenaline; NBM, nucleus basalis of Meynert; NPY, neuropptide Y, RGC, retinal ganglion cells; RGT, retinogeniculate tract; RHT, retinohy- potbalamic tract; SCN, suprachiasmatic nucleus; SHT, spinohypothalamic tract (From Van Someren, 1997). diurnal species may contribute to the regulation of sleep. Secondly, it entrains circadian rhythms in mammals, including human 56-58. The circadian effects of melatonin have led to substantial therapeutic applications for jet lag, shift work, blindness, and some circadian-based sleep disorders 6, 32, 59, 60. Melatonin appears to exert circadian functions via specific, high affinity G-protein- coupled MT1 and MT2 melatonin subtype receptors in the SCN (reviewed in 61, 62). Physiologically, the MT1 receptor mediates the acute inhibitory action of melatonin on the SCN 63, 64, which may contribute to the regulation of sleep, while the MT2 receptor may mediate the phase shift effect of melatonin 65-67. Melatonin receptors and their regulatory mechanisms are reviewed in the third section of this Introduction.

15 chapter 1

Other nonphotic inputs to the SCN Besides melatonin, there are other non-photic signals to the SCN, such as locomotor activity and arousal. Locomotor activity can have both phase-shifting and immediate effects on the SCN. The immediate effect of nonphotic stimuli is the opposite to that of light, with behavioral activity inhibiting SCN neuronal activity 16. Moreover, nonphotic stimuli can counteract the effect of light on the SCN, inhibiting light-induced phase shifts, SCN neuron activity and c-fos expression68-71 . The nonphotic effects are partially through neuropeptide Y (NPY) projections from the intergeniculate leaflet (IGL) in rodents, since lesioning this pathway prevents phase shifting to arousal, which however does not affect phase shifting to light72 . However, since the human SCN has a rather sparce plexus of very fine NPY axons, and the SCN itself contains a large number of NPY neurons, it is not clear whether the intergeniculate leaflet neurons indeed project to the human SCN or whether this projection is very much reduced or even absent in human beings 73, 74. It is possible that the geniculohypothalamic tract is present only in nocturnal species like the rat, and absent in diurnal species, like humans 75. The rodent SCN also receives non-photic input from serotonergic projections from the median raphe nucleus (76, 77, 78; reviewed in 72, 79). Nonphotic stimuli are also important in human circadian rhythm entrainment, which has been demonstrated in blind people who lack light input to the SCN 33.

Output from the SCN Studies in rodents show that there are extensive neuronal projections from the SCN to other brain areas within and outside of the hypothalamus. Postmortem human studies indicate that the SCN projections are similar to those in rodents 24, 25, 80. In mammals, including human, so far the densest SCN projections are to the subparaventricular zone between the SCN and the paraventricular nucleus (PVN) 25, 81, 82. Multi-synaptic pathways from the SCN to many effectors are proposed to amplify and distribute the circadian signals to effector systems 83. For instance, multi-synaptic pathways from the SCN carry circadian massages to the adrenocorticotropic axis and to autonomic ganglia, which innervates the viscera 53. Moreover, the multi-synaptic SCN projects to the pineal gland via the dorsomedial hypothalamic nucleus, the upper thoracic interomediolateral cell columns of the spinal cord (IML) and the superior cervical ganglia (SCG), controls the secretion of melatonin in a rhythmic manner 84 (see Fig. 4). Melatonin enters the bloodstream through passive diffusion, which can impart circadian information throughout the body (see review 85, 86). Rodent studies have revealed a strong projection from the SCN to the supraoptic nucleus (SON) with both inhibitory (GABAergic) and excitatory (glutaminergic) components 87, 88 that may also be responsible for the circadian rhythmicity in the SON. Such connections have, however, not been shown in the human brain, although

16 general introduction

Pineal PVN SCN

Eye

SCG

Spinal cord

Fig. 4 The circadian biological clock of the brain, suprachiasmatic nucleus (SCN) in the hypothalamus, is synchronized to environmental light information through the retina-hypothalamic track, and innervates the pineal gland via a multi-synaptic pathway in human. Abbreviations: PVN: paraventricular nucleus, SCG: superior cervical ganglion.

SCN fibers come very close to the SON 24 and possibly even contact SON dendrites or interneurons. In addition, the lateral retinohypothalamic tract projections that innervate the ventral part of the SON 25 may impose a diurnal rhythm on vasopressin release. In rodent, innervation of the anterior and posterior hypothalamus contributes to circadian control of the orexin/hypocretin system, which is important in the regulation of arousal and temperature 89; (see review 90). The rodent SCN also projects to the intergeniculate leaflet and raphe nucleus, which both have afferents to the SCN 91, and could form a feedback loop.

17 chapter 1

The neuronal transmitters of the projections from the SCN include AVP, VIP, GABA and glutamate. The nature of the signals employed by the SCN to convey rhythmicity to other regions of the brain is still unclear. Recent studies suggest that the clock-controlled secretory peptide AVP, GABA, glutamate, protein prokineticin 2, and TGFα from the SCN contribute to extracellular signaling, which controls behavioral and neuroendocrine cycles (see review 1). AVP shows a clear circadian rhythm and circannual rhythm in the SCN of mammals, including humans 20, 21(Fig. 2). The AVP neuropeptide pattern in the SCN is driven by the AVP mRNA rhythm 92, which is directly controlled by the clock genes in the SCN 93. Based on studies in rats, it is known that AVP does not only modulate the rhythm of activity within the SCN, but also acts as the main SCN message to induce the rhythmicity in other brain regions 94-96. Although AVP may not be required for the expression of the circadian rhythm of the animal (as circadian rhythms are present in the AVP-deﬁcient Brattleboro rat), we should keep in mind that in natural knockout animals biological adaptation and ﬂexibility during development could mask the importance of a gene and its product (i.e. AVP). GABA release from the SCN projections in the PVN area is essential for the circadian rhythm in melatonin production and for the light-induced suppression of melatonin. The GABA release in the PVN area inhibits the sympathetic IML, thereby suppressing noradrenaline release in the pineal gland via the superior cervical ganglion. The circadian rhythm of melatonin is the result of GABA release from the SCN to the PVN during the (subjective) day, without GABA releasing at night. Light stimulation at night causes GABA release from the SCN to the PVN and results in the immediate suppression of melatonin secretion 84, 97. Glutamate release from the SCN into the PVN seems to be crucial for melatonin to reach nighttime peak levels 98 and the SCN thus regulates melatonin secretion both by an inhibitory and excitatory signal. Via these neuronal and endocrine mechanisms (such as melatonin), the SCN coordinates circadian rhythmicity throughout the body 6, 53.

Molecualr components of the circadian oscillators Circadian clockwork of the master oscillator The circadian clock mechanism in mammals involves two interlocking transcriptional feedback loops 99, 100. Initially only a single intracellular transcriptional-translational feedback loops was envisioned, include a series of clock genes, i.e. Clock, Bmal1, Cryp- tochrome genes (Cry1-2), Period genes 1-3 (Per1-3) gene 101-103. The transcriptional factors CLOCK and BMAL1 heterodimerize and activate the expression of three Per genes and two Cry genes by binding to E-box elements in their promoters 104-108. CRY and PER proteins multimerize and translocate to the nucleus, where CRY proteins repress the transcriptional activity of the CLOCK-BMAL1 dimer 109, 110, possibly via

18 Mammals general introduction

REV-ERBá Rev-erbá Per1.2 E-box E-box

Rorá PER/CRY Cry1.2 CLK/BMAL1 E-box E-box RORá

Bmal1 RRE Fig. 5 The mamalian circadian clockworks is comprised of two interlocked feedback loops. The core loop determines the period and the amplitude of circadian oscillations, while the second (stabilizing loop) is important to ﬁnd-tune and stabilize these oscillations. Abbreva- tions: RRE. Rev-erb/Ror element. Modiﬁed from Emery & Reppert, 2006.

inhibition of p300 histone acetyl transferase activity 104. Recently, a second feedback loop has been revealed in mammals and involves the circadian regulation of Bmal1 (Fig. 5). Rev-erbα was identified as a negative regulator for Bmal1 expression, as well as for Clock and Cry1 gene 104, 111, 112, while Ror was identified as an activator of Bmal1 transcription 113. Indeed, Rev-erbα and Ror are able to bind to Rev-erb/Ror elements in the Bmal1 promoter, in a functionally competitive way 111, 113. Interestingly, the Rev- erbα mRNA rhythm is probably directly regulated by CLOCK/BMAL1 heterodimers, through E box elements in the Rev-erbα 111. Sato et al., 113 also provided evidence that CLOCK regulates Rora expression, probably through E-box elements in its promoter (Fig. 5). In the SCN circadian loop, components such as protein accumulation, degradation, and nuclear translocation of the proteins, are drawn out to produce a 24-h cycle and able to sustain oscillation indefinitely 114-116. Factors outside the loop that affect stability of the proteins also alter clock oscillation. In particular, casein kinase I (CKI) δ and ε were shown to phosphorylate PER1 and PER 2, CRYs and BMAL1. CKIε/δ appear to be part of multimeric complexes containing CRY and PER proteins 117, 118, 103. These phosphorylation events target the proteins for degradation 117, 119. Identification of the key features of the clock mechanism remains an important goal. Recent studies suggest the involvement of new factors in the core clock mechanism. In addition to REV-ERBα, other factors could bind to the same elements in the Bmal1, Clock and Cry1 promoters, such as REV-ERBβ, RORβ or RORγ, and may also have a role in the molecular clock mechanism 111-113.

Peripheral oscillators Although the SCN is the master oscillator in mammals, intriguingly, there are also circadian oscillators scattered throughout the body, as the genes involved in the in-

19 chapter 1 tracellular SCN clock mechanism have been found to be rhythmically expressed in other brain areas, in peripheral organs, and even in immortalized cell lines in culture 120-122. Indeed, genetic studies have shown that the molecular composition of the timing mechanisms in the SCN clock and slave oscillators is very similar 123, 124. However, the extra-SCN ‘slave’ oscillators expressed in vivo can only sustain 24-h oscillations for a few days without input from the master clock, and the amplitude decreases very rapidly 121, 122, 125. The actual mechanism that distinguishes the self-sustaining oscillatory function of the master clock from the damped oscillation of slave oscillators is unknown, but the mechanism may be due to the global differences in clock protein levels and/or kinetics rather than to the existence of a specific element (gene/protein) expressed only in the SCN. The master oscillator SCN originally synchronizes the peripheral slave oscillator via neuronal and hormonal cues. There are neural outputs from the SCN to the peripheral organs by means of the autonomic nervous system, supporting direct (multisynaptic) neural control 126. Hormonal signals are capable of entraining peripheral oscillators, as glucocorticoid agonists can effectively shift peripheral oscillators in mice 127. Recent findings indicate a prominent role of feeding rhythms in coupling peripheral oscillators (e.g. liver) to the central clock SCN 128, 129. The complex interaction of neural, hormonal and behavioural outputs may regulate the timing of slave oscillators by converging on Per1 and Per2 gene expression in peripheral tissues (see review 1). Once the slave oscillator is coordinated by the SCN, the synchronized molecular oscillation may be transduced into local rhythms via local first-order clock controlled genes (CCGs) 93, 130. D-element binding protein (DBP) is a well described CCG gene that depends on CLOCK-BMAL1, since it regulates the rhythmic transcription of key enzymes involved in hepatic metabolic processes and it itself regulates a number of genes 131. In fact most CCGs seem to be regulated indirectly by the clock, via transcription factors themselves oscillating in a circadian fashion. Recent evidence indicates that DBP works together with the related basic leucine zipper transcription factor E4BP4, whose rhythm phase is the opposite of that of DBP, to affect differentially the same cis-acting element and thereby drive rhythmicity in the responsive genes 132. Synchronized slave oscillators can regulate local rhythms in physiology and behavior. A hierarchical multioscillatory system seems to confer precise phase control and stability on the widely distributed physiological systems it regulates 133.

Clock networks in human Humans and rodents appear to have a similar set of clock genes and similar mechanisms of molecular clockwork (see review 134). Studies of clock gene oscillation in humans are rather limited, due to the diﬃculties to obtain samples. Human PER1 and CLOCK are expressed widely in brain regions,

20 general introduction non-neuronal tissues and cell lines 135-141. In situ hybridization shows that Clock gene expression is present in the human brain, which is similar to the observations in the mouse 137. Expression of clock genes is also present in the oral mucosa, skin and peripheral blood cells in human 140-143. As in mice, in the human oral mucosa and skin, Per1, Bmal1 and Cry1 gene expression are rhythmic, while Clock mRNA levels are constant 142. Again, as in peripheral tissues of rodents 120, 144, Per1 and Bmal1 rhythms are antiphase, with Per1 peaking in the early light phase in human oral mucosa, skin and blood mononuclear cells 142. Molecular clock mechanisms are probably very similar to those in rodents. This is supported by the ﬁndings that E-box elements of human and mouse Per1 are all conserved, and that PER proteins are substrates for CKI, both in rodents and in humans 115, 135.

Circadian timing system changes in aging and Alzheimer’s Disrupted circadian rhythms in aging and Alzheimer’s disease There is aboudant evidence that aging is characterized by a progressive deterioration of circadian timekeeping (reviewed in 145). Habitual bedtime and wake time are earlier in people in the 40- to 50-year age range than in young subjects. In addition, the middle-aged have a greater orientation toward morningness and have an advanced phase of circadian rhythm of temperature 146. 80% of the subjects in the 50- to 80-year age range show a spontaneous internal desynchronization of rhythms that may affect sleep patterns and other aspects of biological aging 147, 148. Weakened and fragmented circadian sleep-active rhythm was found during aging, with a strong decrease in “actual sleep time” and “sleep efficiency”, and increased “sleep latency” in the old and oldest volunteers under natural environmental conditions 149. Circadian rhythm disruptions are frequently associated with poor sleep, reduced daytime alertness, and decreased cognitive performance 150. Age-related changes have been found in endocrine circadian rhythms, e.g., rhythmic levels of cortisol, melatonin, vasopressin, pulsatile luteinizing hormone (LH), testosterone secretion, endorphine levels and many others hormones 151-156, but the entrained body temperature rhythms appear to be only slightly affected 157, 158. An earlier timing of multiple endogenous rhythms, including core body temperatures and plasma melatonin are also found during aging, which is in accordance with the earlier bed and wake times in elderly 146, 159. The age-related circadian rhythm disturbances are even more pronounced in Alzhe- imer’s disease 148, 160-162. AD patients often suffer from circadian system-related behavioral disturbances—daytime agitation, nightly restlessness and sundowning, which is characterized by an exacerbation of symptoms indicating increased arousal and agitation in the late afternoon, evening or night 163. Disruptions of the circadian rhythms are often so severe that they are even thought to contribute to mental decline in AD

21 chapter 1 patients 164. Moreover, nocturnal insomnia and wandering in AD patients often pose unbearable problems for caregivers, and are the principal causes of institutionalization 165. The endocrine circadian rhythms, such as circadian ﬂuctuations of peripheral cortisol and melatonin, are even more impaired in AD patients than in controls 166. Melatonin secretion impairment is related to age, disturbance of sleep–wake rhythms, and the severity of mental impairment 155, 167. Nocturnal growth hormone secretion as well as the mean levels and nadir values of plasma cortisol are also related to mental impairments 155.

Possible mechanisms underlying the circadian rhythm changes in aging and AD The circadian rhythm disturbances during aging and Alzheimer’s disease168 are probably due to alterations in the circadian timing system: 1. disruptions in the SCN; 2. a decreased input to the SCN; 3. output of the SCN.

Disturbances in the SCN and its output during aging and AD The circadian and circannual fluctuations in vasopressin-expressing neuron numbers in the SCN decrease during aging 169, 170. A marked decrease was found in the number of vasopressin-expressing neurons in the SCN only in subjects of 80 to 100 years of age, while in Alzheimer’s disease these changes occurred earlier and were more dramatic 23, 171, 172. This is also confirmed by in situ hybridization studies on AVP mRNA expressing-neurons in the SCN 173. In the SCN of AD patients the density of vasopressin and neurotensin neurons are decreased, while the glial fibrillary acidic protein (GFAP) stained astrocytes are increased, which indicates a decreased activity of the SCN in AD 174. Moreover, AD-characterized neuropathological changes have been found in the SCN of Alzheimer patients, including the presence of pretangles in the SCN of AD patients 175-177 and tangles 174, as well as few diffuse amyloid plaques 174, 177. AVP is a major peptide output of the SCN neurons, which does not only modulate the rhythm of activity within the SCN, but also acts as the main SCN message to induce the rhythmicity in other brain regions 94-96. The total amount of vasopressin-mRNA seemed to decrease in aged controls over 80 years old compared to those in 60- to 80-year-old. Furthermore, it was 3 times lower in AD patients than in age and sex- matched controls, and no clear diurnal rhythm of vasopressin mRNA was observed in AD patients 173. GABA is one of the main neurotransmitters of the SCN. Preliminary animal studies suggest that GABA network in the SCN might be disrupted during aging, which, however, needs to be confirmed 178. The data mentioned above support the idea that alterations in the SCN and its output is the underlying anatomical substrate for the clinically often-observed disturbances in circadian rhythmicity in aging and Alzheimer’s disease. Age-related changes in the molecular machinery that generates rhythmicity within

22 general introduction

SCN cells were proposed. The recent discovery of clock genes in mammals now makes it possible to examine how the clock mechanism itself may be changing in old age. It has been reported that aging is associated with a reduction in the expression of Bmal1 and Clock in the SCN, while the circadian profiles of Per1 and Per2 in the SCN remain normal during aging in hamster and rat 179, 180, 181. In mice, aging has been found to diminish the amplitude of Per2 but not Per1 expression 182. Moreover, light stimulation-induced Per1 and Per2 gene expression in the SCN of hamster and rat is decreased in aging 179, 180. Interestingly, Oster et al 183 reported a decay of the clock with age in mPer1−/− mCry2−/− mice at the behavioral and the molecular levels, which provided evidence that a clock defect can make the circadian oscillator fall apart more quickly, resembling accelerated aging. Whereas heterozygous Clock mutant mice showed similar circadian rhythm of locomotor activity as wild-type mice, suggesting that the Clock mutation does not render mice more susceptible to the effects of age on the circadian pacemaker 184. So far only few studies are available on the alterations of the molecular clockwork in the aging process of human. There is recent evidence that circadian expression of clock genes was markedly impaired in senescent human aortic vascular smooth muscle cells (HSMC) as compared with young HSMC cells 185. Serum-induced activation of the cAMP signaling pathway was significantly weaker in senescent cells compared to young cells, while treatment with activators of this pathway effectively restored the impaired clock gene expression of senescent cells 185. It is suggested that regulation of clock gene expression may be a novel strategy for the treatment of age-associated impairment of circadian rhythmicity. So far, no data are available on the alterations of the circadian molecular clockwork and its underlying mechanisms during Alzheimer’s disease. Since the SCN is characterized by degeneration during Alzheimer’s disease, we propose that the molecular clock oscillations may also be disrupted in the SCN of AD patients. Since many peripheral clocks (i.e. the pineal gland) are controlled by the master clock (the SCN), disturbances in the SCN may de-synchronize the oscillations of peripheral clocks during the process of Alzheimer’s disease. We investigated the circadian profile of clock genes in a peripheral oscillator, i.e. the pineal gland, in AD patients also because this may provide information about the effectiveness of the synchronization by the SCN during AD. It is hoped that a deeper understanding of the mechanisms underlying age- and AD-related changes in circadian rhythms, by studies of changes in circadian clock gene expression, protein levels and activity, may lead to novel approaches for attenuating changes in the circadian pacemaker during aging and AD.

Reduced Zeitgebers and decreased input to the SCN during aging and AD Due to the occurrence of degeneration of the SCN during aging and AD, the strength

23 chapter 1 of Zeitgebers may be reduced, and the input neural pathways involved in entrainment (synchronization) of the central clock may be dysfunctional or sub-sensitive during aging and even more so in AD. These changes may contribute to the circadian rhythm disturbances in these conditions (see review 6). Light is considered to be most crucial for the entrainment of the SCN, acting via the photic input pathway (retina- RHT-SCN). Several factors have shown to attenuate this projection to the circadian timing system during aging and even more pronounced in AD. Aged people and in particularly AD patients are exposed to reduced illumination levels in their daily lives compared to young people 162, 186, which is related to more night-time awakenings and more daytime sleepiness 187-189. Moreover, the capacity of the lens to transmit light is progressively decreasing during aging. Cataract and maculopathy are more common in the elderly 190, 191, and age-related maculopathy is associated with AD 192. Further- more, retina and the optic nerve, which provides direct and indirect light input to the SCN, shows degenerative changes in AD 193-197. At last, visual field defects and/or optic disc cupping compatible with the diagnosis “glaucoma” were found 5 times more frequently in Alzheimer patients than in controls, which may also have implications on the input to the circadian system 198, 199. Melatonin can feedback on the SCN via melatonin receptors and is involved in the regulation of the circadian rhythm system 63, 200, 201. Melatonin levels are strongly decreased during aging and even more pronounced in AD, accompanied by an impaired day/night rhythm 202, 203. We propose that the circadian effect of melatonin on the SCN may be disrupted during aging and AD. Whether the melatonin receptors in the SCN is affected in theses conditions and contribute to the aging- and AD related circadian disorders is a notable issue. Taken together, the decreased input to the SCN may contribute to the de-activation of the SCN during aging and AD.

Circadian rhythm changes in depression Abundant evidence has shown that circadian timing system alterations are present in depression. 90% of the patients with major depression suffer from sleep disturbances 204. Depressive patients often show a diurnal variation in depressive feelings. They feel worst in the morning and suffer from early morning awakening 205, 206. Depres- sion scores are highest in winter and lowest in summer and ratings of hostility, anger, irritability and anxiety also show strong seasonal effects 207. Patients with seasonal affective disorders often suffer from recurrent depressive episodes during either the fall-winter months (winter form) or the spring-summer months (summer form), while recovering spontaneously outside of these seasons. Seasonal variations in mood and behavior have been documented in epidemiological studies in the general population 208. Moreover, decreased amplitude of circadian rhythms of body temperature, plasma

24 general introduction cortisol, norepinephrine, and thyroid stimulating hormone are reported in depressed patients 205, 209. Recently, our group observed that the number of SCN neurons expressing vasopressin is increased and the mRNA of vasopressin is decreased, indicating a decreased synthetic and transport activity of vasopressin neurons in the SCN 210. The number of neurons in the SCN staining for nitric oxide synthesis or neurophysin is also decreased in depression 211. In addition, melatonin production is disturbed in depressed patients 212, 213. Furthermore, supplementary melatonin to depressed patients may alleviate the depression state and improve sleep 214. These data support an important role of the circadian rhythm system in the pathogenesis of depression. For seasonal affective disorder, it has been proposed that a genetic basis may play a role. A genetic defect at the level of the clock regulation by afferent pathways involving NPY and serotonin is also possible 215. So far, no mutation screening of the circadian clock genes has yet been carried out in seasonal affective disorders. Polymorphisms in the CKIepsilon binding region of hPER2 gene which is associated with advanced sleep phase syndrome (ASPS) (see the section below), are unlikely to play an important role in the development of bipolar disorder 216. A single nucleotidepolymorphism (SNP), T3111C, in the 3’ flanking region of the human Clock gene was not associate with major depression 217, but it may be related with recurrence rate in depressed patients 218. This preliminary observation leads to hypothesize a role for the Clock gene polymorphism in the regulation of long-term illness recurrence in bipolar disorder. Another study in 159 seasonal affective disorder patients and 159 control subjects demonstrated that a polymorphism in NAPS2 471 Leu/Ser was significantly associated with seasonal affective disorder 219. Most recently, Nievergelt et al have assessed evidence for linkage and association involving polymorphisms in 10 circadian clock genes (ARNTL, CLOCK, CRY2, CSNK1epsilon, DBP, GSK3epsilon, NPAS2, PER1, PER2, and PER3) to bipolar affective disorder in 52 affected families 220. They reported evidence suggestive of an association of the circadian genes PER3 and ARNTL with bipolar disorder, which warrants further study with larger samples.

Other human circadian rhythm-related diseases Since many studies report that clock gene mutations have profound eﬀects on circadian rhythmicity in rodents, it is likely that mutations in human clock genes may result in rhythm-related syndromes. Some intrinsic sleep disorders are associated with circadian system function: advanced sleep phase syndrome (ASPS), delayed sleep phase syndrome (DSPS), non-24-hour sleep-wake syndrome and irregular sleep-wake pattern 221, as well as seasonal aﬀective disorders. Recent studies searching for clock gene mutations or polymorphisms that may give rise to circadian sleep disorders have revealed a genetic origin for these circadian rhythm diseases (for review see 124).

25 chapter 1

Advanced sleep phase syndrome (ASPS) ASPS is a rare disorder which is characterized by very early sleep onset (<19:00 h) and offset (<04:00 h) 221. Recently, a familial form of ASPS was described in 29 members of three families who have a profound phase advance (4 hours) of circadian rhythms of sleep-wake, melatonin and temperature, which are associated with a very short circadian period 222. Sleep recordings, when done at the patient’s desired sleep time, are generally normal. However, when performed at later times, early morning awak- ing and reduced total sleep time are observed and evening sleepiness is reported. Transmission of the ASPS is consistent with an autosomal dominant mode of inheritance 222. Morningness scores of first-degree relatives revealed a greater tendency for morningness, which indicates the presence of family ASPS carriers 222. This familial form of ASPS was recently associated with a missense mutation in a phosphorylation site within the casein kinase I (CKI)-binding domain of the human PER2 gene 223. This mutant form of PER2 is hypophosphorylated 223 and in mouse this Per2 mutation produces a short circadian phenotype resembling of ASPS 224. In addition, the hamster tau mutation, a missense mutation in CKIepsilon, also shortens the endogenous circadian period and results in behavior that resembles ASPS 225. Similarly, a missense mutation (T44A) in the human CKIdelta gene may also cause ASPS. This mutant kinase has decreased enzymatic activity in vitro. Transgenic mice carrying this mutation have a shorter circadian period, a phenotype mimicking human ASPS 226. Besides, a recent study reported that a single-nucleotide polymorphism in the 5’-untranslated region of the hPER2 gene (C111G) was associated with morning preference and is a potential candidate allele for ASPS 227. Therefore mutations that alter the phosphorylation state of PER2 can disrupt sleep-wake patterns by altering the circadian timing of sleep. How- ever, another study which also identified a Japanese family in which ASPS segregates with an autosomal dominant mode of inheritance did not reveal the missense mutation (bp2106 A/G) in hPer2 or any linkages between affected individuals and hPer2 228. The latter study supports genetic heterogeneity of familial ASPS in humans.

Delayed sleep phase syndrome (DSPS) DSPS is one of the most common circadian rhythm-related sleep disorders, and occurs more frequently in young adults. It results in a phenotype which is opposite to that of ASPS: individuals sleep (03:00–06:00 h) and wake (10:00–14:00 h) later than desired 221. Sleep quality, sleep stage distribution, and sleep duration are normal when they sleep at their desired times, although their sleep latency is frequently longer than 30 min. When sleep is planned earlier, a significant increase in sleep latency and wake time during the first part of the night is observed 221. One human study of 48 DSPS patients reported a significant association between amino acid polymorphisms in hPER3 and DSPS 229. This polymorphism might alter the CKI epsilon-dependent

26 general introduction phosphorylation of hPER3 and cause DSPS. Johansson et al reported that the same Per3 polymorphism was associated with self-reported diurnal preference in a group of both seasonal depressive patients and controls 219. Recently Archer et al. has identified a length polymorphism (four or five repeating units) in a region of the human PER3 that contains several putative CKIepsilon phosphorylation sites 230. Moreover, they found that this length polymorphism in Per3 correlated significantly with extreme diurnal preference and DSPS 230. The longer allele was associated with morningness whereas the shorter allele was associated with eveningness. The shorter allele was strongly associated with DSPS subjects, 75% of whom were homozygous for the shorter allele 230. Polymorphisms in the entire coding region of the hClock gene were not associated with DSPS 231. A T3111C polymorphism in the 3’-untranslated region of hClock, which had been reportedly associated with morning or evening preference for activity 217, seemed to be associated with DSPS 231.

Non-24-h sleep-wake syndrome Non-24-h sleep-wake syndrome is extremely rare in sighted subjects living in normal environmental 24-hour light/dark conditions. This syndrome is observed in blind patients, astronauts and sub-mariners living in a non-natural light/dark cycle 232. It can also occur in psychiatric conditions such as schizoid or avoidant personality disorders 233, 234. Patients with non-24h sleep-wake syndrome often have irregular sleep-wake cycles and periodic insomnia with daytime sleepiness 232. So far, mutation screening analyses of human Per3 and Clock gene did not reveal an association with Non-24-h sleep-wake syndrome 231. However, the frequency of R54W and the A157V variants of the human melatonin 1a receptor (MT1) gene tends to occur more frequently in non-24-h sleep-wake syndrome patients than in controls 235, 236. Further conﬁrmation is needed and its possible role on the etiology of the disorder must be investigated. In addition to a mutation in core clock genes, other genes important for input and output pathways of the clock mechanism may also be involved in human rhythm disorders. Mouse knock-outs for other genes display defects in circadian rhythmicity, such as PAC1 receptor, whose ligand is present in the retina-hypothalamic tract 237, Crx gene expressed in the retina 238, and malanopsin, a novel opsin involved in circadian photoreception 239, 240. The neuropeptides pituitary adenylate cyclase-activating polypeptide (PACAP) and vasoactive intestinal peptide (VIP) are implicated in the photic entrainment of circadian rhythms in the SCN. Mice lacking of the VPAC(2) receptor for VIP and PACAP (Vipr2(-/-)) are incapable of sustaining normal circadian rhythms of rest/activity behavior. Moreover, they show both disrupted light response and clock functions 241. This study highlights the role of intercellular neuropeptider- gic signaling in the coupling of the SCN neurons and sustaining a robust circadian

27 chapter 1 rhythmicity within the SCN. Mouse mutants for the NCAM cell adhesion molecule 242, tyrosine kinase fyn 243, retinoid-related orphan receptor (RORalpha) gene 244, TrkB tyrosine kinase gene 245 and ras-associated binding protein Rab3a, which is involved in synaptic vesicle traﬃcking 246, also present circadian rhythm defects.

Conclusion Circadian rhythms are an important aspect of human biology 247. Understanding the molecular mechanism of circadian rhythms in human, in addition to further extending our understanding of circadian and sleep disorders, can provide an unprecedented opportunity for pharmacological manipulation of the SCN clockwork, which could revo- lutionize treatment of jet lag, sleep disorders, and some neuropsychiatric illnesses.

B. The human pineal gland and melatonin in aging and Alzheimer’s disease

Ying-Hui Wu, Dick F. Swaab. Journal of Pineal Research 38 (3): 145-152, 2005

Abstract The pineal gland is a important structure in the circadian system which produces melatonin under the control of the central clock, the suprachiasmatic nucleus (SCN). The SCN and the output of the pineal gland, i.e. melatonin, are synchronized to the 24-hour day by environmental light, received by the retina and transmitted to the SCN via the retinohypothalamic tract. Melatonin not only plays an important role in the regulation of circadian rhythms, but also acts as antioxidant and neuroprotector that may be of importance in aging and Alzheimer’s disease (AD). Circadian disorders, such as sleep-wake cycle disturbances, are associated with aging, and even more pronounced in AD. Many studies have reported disrupted melatonin production and rhythms in aging and in AD that, as we showed, are taking place as early as in the very ﬁrst preclinical AD stages (neuropathological Braak stage I-II). Degeneration of the retina-SCN-pineal axis may underlie these changes. Our recent studies indicate that a dysfunction of the sympathetic regulation of pineal melatonin synthesis by the SCN is responsible for melatonin changes during the early AD stages. Reactivation of the circadian system (retina-SCN-pineal pathway) by means of light therapy and melatonin supplementation, to restore the circadian rhythm and to relieve the clinical circadian disturbances, has shown promising positive results.

Pineal gland and melatonin In humans, the pineal gland is 5 mm long, 1-4 mm thick and weighs about 100 mg, both in men and in women 248. The pineal gland contains two major cell types: neu- roglial cells and the predominant pinealocytes that produce melatonin.

28 general introduction

The pineal gland is an important structure in the circadian system that is innervated by a neural multi-synaptic pathway originating in the suprachiasmatic nucleus (SCN) that is located in the anterior hypothalamus. The SCN is the major circadian pacemaker of the mammalian brain and plays a central role in the generation and regulation of biological rhythms 53, 249. The pineal gland produces melatonin in a marked circadian fashion 250, reflecting signals originating in the SCN. The human SCN innervates only a small number of hypothalamic nuclei directly 24, 25. However, it may impose circadian fluctuations indirectly on the organism by means of melatonin released from the pineal gland 85. The biosynthetic pathway of pineal melatonin has been studied thoroughly. L- Tryptophan is taken up from the circulation and converted to serotonin (5-HT) by tryptophan hydroxylase. 5-HT is metabolized by the rate-limiting enzyme arylalkylamine N-acetyltransferase (AA-NAT) to N-acetyl-5-hydroxytryptamine, and in turn by hydroxyindole-o-methyltransferase to melatonin. 5-HT can also be oxidized by monoamine oxidase A (MAOA) to 5-hydroxyindoleacetic acid 250, 251. In all vertebrates, the activity of the rhythm-generating enzyme AA-NAT increases at night by a factor 7 to 150, depending on the species. The molecular mechanisms regulating AA-NAT are also remarkably different among species. For instance, in the rat, pineal AA-NAT is regulated at both mRNA level and protein level; however, in sheep and rhesus macaque, pineal AA-NAT mRNA levels show relatively little change over a 24- hour period and changes in AA-NAT activity are primarily regulated at the protein level 252, 253. In the human pineal gland, significant daily fluctuations in AA-NAT mRNA levels were not detected either 254, which suggests that pineal AA-NAT activity may be mainly regulated on the posttranscriptional level in human. The main environmental control of the pineal melatonin synthesis is light intensity. Light perceived by the retina reaches the SCN through the retinohypothalamic tract (RHT), which has been revealed by an in vitro postmortem tracing procedure, also in the human hypothalamus 48. The SCN innervates the pineal gland via the dorsomedial hypothalamic nucleus, the upper thoracic intermediolateral cell columns of the spinal cord and the superior cervical ganglia (SCG), resulting in the rhythmic secretion of melatonin 84, 255. The importance of ocular light as a temporal cue has been clearly demonstrated in circadian studies of blind people, who were bilaterally enucleated, showing desynchronized melatonin and cortisol rhythm 256, 257. Abundant evidence indicates that in humans the sympathetic stimulus is crucial for melatonin secretion. A very large goiter may compress the SCG, thus altering melatonin synthesis in patients 258. After bilateral T1-T2 ganglionectomy in a patient with hyperhidrosis, melatonin levels in the CSF and plasma were markedly reduced and the diurnal rhythm was abolished 259. A circadian rhythm of β1-adrenergic receptors has been found in human pinealocytes 260. Propanolol, a β−adrenergic receptor antagonist, causes a dose-dependent

29 chapter 1 decrease in melatonin levels, or even totally abolishes the night-time surge 261, 262. In turn, melatonin elicits two distinct, separable eﬀects on the SCN, i.e., acute neuronal inhibition and phase shifting, through melatonin receptors in the SCN64 . The ability of melatonin to phase-shift the circadian system has been extensively investigated in humans 32, 60, 263, 264. Moreover, melatonin acts as an eﬀective free radical scavenger 265 and indirect antioxidant, i.e., it stimulates antioxidative enzymes 266, and functions as neuroprotector that may be of importance in aging and in Alzheimer disease (AD) (reviewed in 267, 268).

Aging Circadian rhythms disruptions in aging There is a great deal of evidence indicating that aging is characterized by a progressive deterioration of circadian timekeeping (reviewed in 145). Changes in the intrinsic free-runing period of the clock during aging have been reported but remain equivocal 147, 269, 270. In addition, aged people frequently show a spontaneous internal desynchronization of rhythms 147. For instance, the onset of the activity cycle is earlier than that of the morning rise in body temperature, which may aﬀect sleep-wake patterns 271, 272. Changes in circadian rhythms are frequently associated with a reduction in night- time sleep quality, a decrease in daytime alertness, and an attenuation in cognitive performance (reviewed in 145).

Melatonin production during aging Reduced melatonin concentrations during aging, especially nocturnal levels, have been extensively reported in the pineal gland, plasma, cerebrospinal fluid (CSF) and in urine as 6-hydroxymelatonin (reviewed in 202, 249), although in some studies the age-related melatonin difference was not statistically significant 273-275. Recently, our group measured circadian salivary free melatonin levels in 52 healthy young (21-25 years of age), middle-aged (41-53 years of age), old (60-72 years of age) and very old volunteers (80-93 years of age). A clear circadian rhythm of salivary melatonin was present in all age groups. We found, however, that a step-wise decrease in the circadian amplitude of salivary melatonin occurred early in life, around 40 years of age. The amplitude in the middle-aged subjects was only 60% of that of the young subjects. Both the old and very old subjects showed increased daytime (baseline) melatonin levels 151. Studies of the melatonin metabolite 6-sulfatoxymelatonin show that age-related decrease in melatonin production occurs even as early as 20-30 years of age 276. Zhao et al. 277 found that a decline of the nocturnal serum melatonin peak was significant at the age of 60 and further declined from 70 years of age onwards. The discrepancies among different studies may be due to the differences in methods, criteria for select- ing subjects (health conditions, medical history, etc), and experimental conditions.

30 general introduction

Moreover, the inconsistency may depend on the large (20-fold) genetically determined inter-individual variability in human nightly melatonin secretion 278 and on the size of pineal gland 279. In view of the large individual variability, longitudinal studies assessing melatonin rhythmicity within subjects across time need to be performed to resolve the question when exactly melatonin begins to decline with age in an individual. Besides the age-related decline of melatonin production, age-related changes in the timing of the melatonin rhythm have also been reported 159. Moreover, older subjects enter sleep and awake earlier relative to their nightly melatonin secretory episode 159, which indicates that aging is also associated with a change in the internal phase relationship between the sleep-wake cycle and the output of the circadian pacemaker.

Possible mechanisms for age-related melatonin changes There are some observations that give insight into the possible mechanisms of the decreased melatonin levels during aging. The pineal gland shows clear age-related changes. Human pineal gland calcification increases with age 280. Some studies have related calcification of the pineal gland to a disturbed circadian rhythmicity in the sleep-wake cycle 281 and a decline in melatonin production with age 279, 282. However, Bojkowski and Arendt 283 have reported that there was no relationship between plasma melatonin or the metabolite of melatonin in urine, 6-sulphatoxymelatonin, and pineal calcification. Indeed, there is no direct evidence that pineal calcification affects pineal metabolism. Moreover, the pineal gland shows no obvious signs of degeneration 284-286; even in very old subjects the pineal parenchyma is histologically still apparently active 287. However, the only histological study available on this topic suggests that the noradrenergic innervation of the pineal gland originating from the SCN may be affected during aging 288. The central clock SCN shows age-related degenerative alterations. The circadian rhythm of melatonin levels is regulated by the SCN, the clock of the brain. Age-related alterations have been noted in the human SCN. Circadian and circannual rhythmicity of neuropeptide synthesizing neurons of the human SCN, such as vasopressin, are reduced with aging 169, 170, 289. It has been shown that the number of vasopressin-immunoreactive neurons in the human SCN exhibits a marked diurnal oscillation in young subjects, with low vasopressin neuron numbers during the night and peak values during the early morning. However, this rhythm disappeared in subjects over the age of 50 169. The marked annual oscillation of vasopressin-expressing neurons in the SCN of young subjects, i.e., low vasopressin neuron numbers during the summer and peak values in autumn, was also disrupted in subjects over 50 years 170. A significant decrease in the number of vasopressin-expressing neurons in the SCN was found only in subjects of 80 to 100 years of age 23. How these changes in the SCN translate into the observed changes in circadian rhythmicity is a topic of current research. In addition

31 chapter 1 to the changes in SCN vasopressin and in melatonin rhythms, age-related changes in the amplitude of other circadian rhythms have also been reported in human, e.g., core body temperature, cortisol, vasopressin, blood pressure, pulsatile LH, testosterone secretion, β-endorphine levels, etc (reviewed in 249). These findings suggest that the changes observed in the melatonin rhythm may be part of a general effect of aging on the central clock SCN and/or its regulation. Both Zeitgebers and synchronization are disturbed during aging. Light has significant effects on melatonin synthesis in the pineal gland via the SCN and a multisynaptic pathway thereafter. Reports suggest that aged people are exposed to reduced illumination levels in their daily lives (reviewed in 145), and there are studies showing a negative correlation between environmental light intensity and sleep disturbances in the aged and in AD patients 187-189. In addition, the capacity of the lens to transmit light progressively decreases during aging, which may also contribute to disrupted melatonin production and circadian disturbances in the elderly 290. Nevertheless, elderly people have a maintained responsiveness of the circadian pacemaker to light, which implies that scheduled bright light exposure can be used to treat circadian phase disturbances and related sleep disorders in older people 291.

Alzheimer’s disease Circadian rhythm disruptions in AD The fragmented sleep-wake pattern which occurs in elderly, is even more pronounced in AD patients 148, 162. Continuous measurement of the circadian rest-activity cycle for 589 days in a demented patient with probable AD revealed slowly progressive changes in temporal organization until death 292. Many AD patients also often suffer from circadian system related behavioral disturbances, such as daytime agitation and nightly restlessness 189, 293. In AD, circadian rhythm disturbances are often so severe that they are even thought to contribute to mental decline 164. Moreover, nocturnal insomnia and wandering in AD patients often pose unbearable problems for caregivers, and are the principal causes of institutionalization 165. However, until now there is no successful pharmaceutical treatment for the circadian disturbances in AD. Hypnotic or antip- sychotic medication is only slightly effective for relieving circadian disturbances 294. Benzodiazepines have insignificant effects on sundowning 295, while sleep-wake cycle disturbances may even be aggravated by a classic neuroleptic like haloperidol 296.

Melatonin changes in preclinical AD and clinical AD Impairment of melatonin secretion is not only related to age but also to the severity of mental impairment 155. Numerous studies demonstrate that nocturnal melatonin levels are selectively decreased in AD 254, 297, and that day time melatonin levels are increased in AD patients 298, indicating that the neurodegenerative process aﬀects the circadian-

32 general introduction pineal system (reviewed in 249). AD patients with disturbed sleep-wake cycle possess melatonin secretion rhythm disorders 167 and the disappearance of daily melatonin rhythm in AD patients is consistent with clinical circadian rhythm disorders, such as delirium, agitation and sleep-wake disturbance 299, 300. We have observed a strong reduction in postmortem CSF melatonin levels in Alzhe- imer patients. CSF melatonin levels of AD patients were only one fifth those in control subjects. The melatonin levels of patients with apolipoprotein (APOE) ε4/4 type were even significantly lower than those expressing APOE ε3/4 301. Interestingly, the melatonin levels in CSF decrease with the progression of AD neuropathology as determined by the Braak stages 302. More strikingly, CSF melatonin levels are already reduced in preclinical “AD” subjects, who are cognitively still intact “control” subjects that show only the earliest signs of AD neuropathology (Braak stages I-II) 302. Also, in the postmortem human pineal gland, we found reduced nocturnal melatonin production and a disappearance of day/night variations of melatonin content already from the earliest, preclinical AD stages (Braak stages I-II) onwards 254. A significant high correlation exists between pineal melatonin content and CSF melatonin levels 254 and between CSF and plasma melatonin levels 303, suggesting that reduced melatonin levels may serve as an early marker for the very first stages of AD that could so far not be monitored in any other way. Melatonin deficiency is possibly not only a consequence of the AD process, it may contribute to the pathogenesis of AD, since melatonin was found both in in vitro and in vivo experiments to act as an antioxidant and neuroprotector (reviewed in 267, 268). A recent study reported that melatonin increases survival and inhibits oxidative and amyloid pathology in a transgenic mouse model of AD 304.

Mechanisms underlying the melatonin changes in the progress of AD The pineal gland shows molecular changes both in preclinical and clinical AD. How- ever, in the pineal gland of AD patients, cells or afferent fibers are clear from neuropathological hallmarks of AD, i. e., neurofibrillary tangles, the accumulation of neu- rofilaments, tau, hyperphosphorylated tau or β/A4 amyloid deposition 286. There was, moreover, no alternation in calcium deposition in the pineal or choroid plexus in AD 305. Neither pineal weight nor pineal total protein content show changes in AD 254. Recently we have taken a comprehensive approach to study the molecular changes in the melatonin synthesis pathway and its noradrenergic innervation in human postmortem pineal glands from age and sex matched controls, “preclinical AD” subjects (Braak stages I-II) and clinical AD patients (Braak stage VI). The circadian melatonin rhythm disappears due to decreased nocturnal melatonin levels in both “preclinical AD” and AD patients. Moreover the circadian rhythm of β1-adrenergic receptor mRNA disappears in both patient groups, which suggests a dysfunction of the SCN

33 chapter 1 innervation to the pineal. The precursor of melatonin, serotonin, was stepwise depleted during the course of AD, as indicated by the up-regulated monoamine oxidase A (MAO-A) activity and gene expression. We concluded from our study that a dysregulation of noradrenergic innervation and the depletion of serotonin by increased MAO-A resulted in the loss of melatonin rhythm and reduced melatonin levels from the earliest AD stages onwards 254. In fact, the dysfunction of the SCN innervation to the pineal in AD has also been suggested by previous morphological studies, which demonstrated that the superior cervical ganglia, and the noradrenergic fibers in the pineal gland show swollen axons in AD patients 286, 288. Actually the SCN of AD patients is severely affected 23, 169, which may well be responsible for a disrupted sympathetic control of melatonin synthesis. Increased MAOA activity and mRNA levels, as found in the AD pineal gland, seem to be a general phenomenon in AD, since it was also reported in the cortex, thalamus, hypothalamus and white matter of AD patients 306- 308. Interestingly, MAO-A gene polymorphisms are suggested to be associated with an increased susceptibility for AD 309. However, these polymorphisms have not been studied relative to the changes in pineal melatonin synthesis in AD. The biological clock SCN is severely affected in AD. It shows prominent degenerative changes in AD 23, 173, 310, and also the typical cytoskeletal AD alterations of pretangles 175, 177, 311 and tangles 174. Diffuse amyloid plaques, however, are infrequently noted in this nucleus of AD patients 174, 177. The vasopressin-expressing neuron numbers in the SCN decrease earlier and are more dramatic in AD than in aged controls 23. In addition, the total amount of vasopressin mRNA is 3 times lower in AD patients than in elderly controls, and the diurnal rhythm of vasopressin mRNA apparent in controls is no longer visible in AD patients 173. The density of vasopressin and neurotensin neuron is decreased, while the glial fibrillary acidic protein (GFAP) stained astrocytes are increased in the SCN of AD patients 174. We propose that these degenerative changes in the SCN most probably result in a disrupted melatonin synthesis, and may underlie the clinically common circadian rhythm disorders in AD. The question whether the SCN shows such alterations already in the earliest AD stages and thus causes the changes in the pineal gland in preclinical AD stages, is now under our investigation. The input of environmental light to the circadian timing system is also disrupted in AD. Besides the degenerative changes that are present in the SCN of AD, it is also important to note that several factors attenuate the input of environmental light to the circadian timing system during AD. AD patients are exposed to less environmental light than their age-matched controls 186. Furthermore, the retina and optic nerve show degenerative changes in AD, however, without neurofibrillary tangles, neuritic plaques or amyloid angiopathy 196, 197. Moreover, age-related maculopathy is associated with AD 192 and “glaucoma” was found 5 times more frequently in AD patients than in aged controls 198. In contrast to the observed degenerative changes mentioned above, some

34 general introduction studies did not find significant AD-related changes in the retina and optic nerve cells 312-314. Anyhow, the retiohypothalamic connection and SCN efferents are still present in AD 25, 48, and studies on the reactivation of the circadian system (retina-SCN-pineal pathway) by means of light therapy and melatonin supplementation to restore the circadian rhythm and relieve the symptoms are commonly carried out (see below).

Melatonin Supplementation In AD patients, melatonin has been suggested to improve circadian rhythmicity, decreasing agitated behavior, confusion and ‘sundowning’ in uncontrolled studies 315-317. Melatonin has also been suggested to have beneficial effects on memory in AD 163, 316-318, possibly through protection against oxidative stress and neuroprotective capabilities (reviewed in 267, 268). However, these suggestions need to be confirmed in well-controlled studies, and it should be noted that a few randomized placebo-controlled trials of melatonin administration to AD patients did not find improved sleep-wake pattern 319, 320. If there are beneficial effects of melatonin in AD, it may relate both to the indole’s ability to synchronize circadian rhythms and to the antioxidant action of melatonin.

Light Therapy and Other Non-Pharmaceutical Stimuli Supplementary exposure to morning bright light has shown beneficial effects on sleep quality and daytime vigilance of elderly 321. It also significantly increases melatonin secretion in aged people, to levels similar to those in young adults 188 and it improves body temperature rhythm 291, which implies that scheduled bright light exposure can be expected to facilitate the circadian rhythm entrainment and synchronization. In AD patients, bright light therapy improved both sleep-wake rhythm disorders and behavioral disorders, such as sundowning, wandering, agitation and delirium 319, 322, 323. Moreover, it appeared to improve the cognitive state of AD patients 324, especially in the early stages of the disease 323. In contrast to the literature mentioned above, Mishima et al. 325 found that bright daytime light treatment induced a significant reduction in nighttime activity, but only in patients with vascular dementia and not in patients with AD. It should be noted, however, that the proportion of vascular dementias has always been overestimated when no neuropathological confirmation is performed, since in most demented patients with vascular lesions Alzheimer changes are found as well. Other non-pharmacological means to improve the circadian rhythmicity in aging and in AD are also reported. An increased level of physical activity improves circadian rhythmicity in healthy elderly people, as was found following a 3-month fitness training period 326. Pacing also improves the synchronization of the rest activity cycle in AD patients 292. Improved circadian rest activity rhythmicity was also observed following transcutaneous electrical nerve stimulation (TENS) in AD patients 327, 328 In summary, pineal melatonin rhythms and production decrease in aging and in AD,

35 chapter 1 even as early as in the very ﬁrst preclinical AD stages (Braak stage I-II). Our recent studies indicate that it is the sympathetic regulation of pineal melatonin synthesis by the SCN that is dysfunctioning and is responsible for melatonin changes during the early AD stages. There is still plasticity in the circadian system (retina-SCN-pineal pathway) in aging and in AD, and stimulation of the circadian system by non-pharmacological means, such as light therapy, melatonin, or TENS, has shown important therapeutic consequences for elderly and AD patients. Whether, in addition to light, the administration of melatonin is more eﬀective in cases of circadian rhythm disorders and cognitive function decline, is currently under investigation by our group.

C. Melatonin receptors: localization, molecular biology and physiological significance

Introduction There is growing evidence of the complexity of the role of the pineal hormone melatonin in modulating a large number of physiological processes, including circadian rhythms, seasonal rhythms, sleep, oxidative stress, reproduction, stress, mood, body temperature, cognition, cardiovascular physiology, oncogenesis, and retina physiology 57, 86, 318, 329l, 330-334. Recognition of the important role of melatonin in diﬀerent branches of physiology makes the investigation of melatonin receptors of great importance.

Identification and localization of melatonin receptors The development of 2-[125]-iodomelatonin for use in a radioimmunoassay meant a breakthrough in the studies on putative melatonin receptors 335, 336. By means of this tool, melatonin receptors have been identiﬁed, characterized and localized 337-340. Numerous striking features of melatonin receptor distribution in the mammalian brain have been revealed. First, the distribution of 125I-Mel binding within the brain showed great variability among species 341. Second, 125I-Mel binding is constantly present in the SCN and in the pars tuberalis of the pituitary gland (PT) of all the mammals studied, including human 342, 343. Melatonin receptors in the SCN are thought to mediate circadian responses to melatonin 342, while melatonin receptors in the PT are believed to contribute to reproductive responses, such as modulating the release of prolactin (for review, see 344, 345). In the hypothalamus of several rodent species, speciﬁc 125I-Mel binding is found in the paraventribular, dorsomedial and ventromedial hypothalamic nuclei, anterior hypothalamus, medial preoptic area, and paraventricular thalamic nuclei, in addition to the SCN and PT. Moreover, many other brain regions, such as the hippocampus, cerebral cortex, area postrema, amygdala, and retina, were

36 general introduction observed to have speciﬁc Mel-binding (for review, see 343). In the human brain, the distribution of Mel-binding seemed to be quite similar to that in rodents 346, 347. However, due to practical problems in obtaining suitable human tissue, Mel-binding in human brain has not been the subject of profound investigations. In particular, the data on the exact melatonin receptor distribution in the human hypothalamus and pituitary, which are the main targets of melatonin, still very limited 200, 347. Peripheral tissues have also been investigated for the presence of melatonin binding sites. Tissues in which binding has been found include harderian gland 348, 349, spleen 350 adrenal gland 351, human kidney 352, arteries 353, human spermatozoa 354, and human T-lymphocytes 355, 356. Often, these binding sites require better characterization and clariﬁcation of their physiological importance. Taken together, the wide distribution of melatonin receptors in the periphery may indicate a physiological importance of melatonin that exceeds its central role.

Molecular identification of melatonin receptor subtypes Originally, binding sites for 125I-Mel have been classified, based on kinetic properties and ligand specificity into two subtypes, the ML-1 and the ML-2 melatonin binding protein. ML-1, a high affinity binding protein, with affinity in the picomolar range (10-300 pM), was identified with binding being rapid, stable, saturable and reversible. The relative order of potency of several melatonin-related compounds is 2-iodomelatonin > melatonin > 6-hydroxymelatonin > N-acetylserotonin > prazosin > serotonin. ML-2 is a low affinity binding protein (0.3-5 nM), and has a distinct pharmacological character (for review, see 357). Three receptor subtypes with high affinity for melatonin have been cloned in avian, 63, 358, 359 characterized and termed Mel1a, Mel1b and Mel1c . So far, only Mel1a and Mel1b are identified in mammals. The nomenclature committee of the International Union of Basic and Clinical Pharmacology (IUPHAR) has designated them as MT1 and MT2 360. They are both high affinity G protein-coupled seven transmembrane receptors 62.

Mel1c was not addressed by the IUPHAR. The low-affinity melatonin binding protein “ML-2” was designated by the IUPHAR committee as MT3 melatonin receptor subtype 360. MT3 is not a G-protein coupled receptor; it shares 95% if its homology with the human quinon reductase 2, an enzyme involved in detoxification 361. The functional significance of this receptor subtype still needs further study. Nuclear melatonin binding were recently reported in rat liver 362, human B-lymphocytes 363 and rat brain 364. In the latter two studies, the binding proteins were identified as the Retinoid Z Receptor a (RZRa) and RZRb respectively. These receptors belong to a superfamily of nuclear orphan receptors, involved in ligand-induced transcriptional control.

37 chapter 1

A mammalian melatonin receptor-related receptor, structurally related to the melatonin receptors but incapable of binding melatonin has also been isolated 365. The natural ligand of this receptor is not clear yet, but an interesting distribution of this receptor in neuroendocrine tissues has been reported in rodents 365, 366.

Melatonin receptor subtypes, their distribution and role in physiological processes MT1 MT1 is considered to be the most widely expressed melatonin receptor subtype in mammals. All melatonin binding detected in mouse brain and pituitary disappears in MT1 receptor-deficient mice 64. Moreover, Mel-binding in rodent SCN is resistant to inhibition by the selective MT2 receptor antagonist 4P-PDOT 65. These data support a primary role of the MT1 receptor in the mammalian brain. MT1 mRNA is located in many regions of the mammalian brain, including that of rodents and humans, which correlates with the distribution of receptor binding 64, 347, 367-369. MT1 receptor, which is expressed in the SCN of the hypothalamus and in the PT of the pituitary, is involved in modulating circadian rhythm and reproduction 63, 64. Me- latonin appears to feedback to the SCN by two distinct mechanisms. Firstly, melatonin acutely inhibits SCN neuronal firing 54, 55, which may be important for defining the sensitivity of the SCN to environmental stimuli, and in diurnal species may contribute to the regulation of sleep. Secondly, it entrains circadian rhythms in mammals, including human 56-58. In transgenic mice without MT1 receptor, the acute inhibitory effect of melatonin on SCN multiunit activity is completely absent, while the phase-shifting response of melatonin appears to be normal 64. These data demonstrates the role of the MT1 receptor in the SCN. MT1 receptor, expressed in the pars tuberalis, mediates the action of melatonin to influence the prolactin secretion from the pars distalis, and is involved in the regulation of reproduction 370, 371. Apart from a central localization, the MT1 mRNA is also present in many peripheral tissues of rodents 367, 368, 372. In humans, the distribution of MT1 mRNA and MT1-immunoreactivity in numerous peripheral tissues is well-documented, such as in the human breast 373, the eye and the retina 374, 375, in hair follicles 376, and in the coronary artery 377. Elucidating the physiological role of MT1 in many of these tissues is currently under investigation and may yield interesting results.

The MT1 receptor can couple to numbers of G-proteins, including Giα2, Giα3 and 378 379, 380 Gαq , Giαs, Giαz and Giα16 , which may also explain at least part of its variety of responses within the body. Melatonin acting through the MT1 receptor can induce distinct cellular responses. Activation of MT1 has shown to produce inhibitory responses on the cAMP signal transduction cascade 63, 378, 381-383, resulting in a decrease

38 general introduction in PKA activity 383, 384 and in CREB phosphorylation 383, 385. In addition to playing a role in the cAMP-dependent cascade, MT1 can couple to a stimulation of PLC-dependent 378, 380 386, 387 signal transduction cascades directly or indirectly via Gβγ subunits, and is able to activate PKC 383. Stimulation of recombinant human MT1 receptors can also potentiate the prostaglandin F2α-induced release of arachidonate and hydrolysis of phosphoinositide 386. Furthermore, activation of the MT1 receptor induces a transient elevation in cytosolic calcium ion concentration and in inositol phosphate accumulation 378, 387. The MT1 receptor can stimulate c-jun N-terminal kinase (JNK) activity379 and also modulate MAP kinases 379, 388.

MT2 The MT2 receptor has a much lower expression level in the mammalian brain compared to the MT1 receptor. Moreover, its localizations are also more limited than the MT1 receptor. MT2 is present in the rodent SCN 389, hippocampus, the retina, kidney, ovary, cardiac vessels 368, 372, as well as in the human retina 390, hippocampus 391, coronary artery 392 and hair follicles 376. MT2 receptor mediates the phase-shift effect of melatonin in the SCN, and is involved in the regulation of circadian rhythms 64, 67. Moreover, MT2 receptor is also involved in retinal physiology 393, dilating cardiac vessels 394 and inflammatory response in the microcirculation 395. Most recently, it was demonstrated that the MT2 receptor in the hippocampus mediates the inhibitory effect of melatonin on the long-term potentiation of hippocampal neurons, via the adenylyl cyclase-protein kinase A pathway. This indicates an involvement of MT2-receptor in the regulation of hippocampal synaptic plasticity 396.

MT3 The MT3 subtype, as revealed through radioligand binding and enzymatic assays, is expressed in the liver, kidney, brain, heart, brown adipose tissue, skeletal muscle, lung, etc 397. Recently it was reported that the MT3 receptor may be involved in the regulation of intraocular pressure in rabbits 398 and in the inﬂammatory response in the microvasculature 395.

Recent studies of the pharmacological specificity of recombinant melatonin receptors have identified compounds that distinguish between the subtypes. For instance, the selective MT2 receptor antagonist 4P-PDOT has been used for experimental purposes 399. Selective pharmacological compounds provide special approaches to distinguish the effects mediated by the receptor subtypes, and may also have potential clinical use.

39 chapter 1

Regulation of melatonin receptors Melatonin receptors can be regulated in a homologous manner, i. e. by melatonin itself, or in a heterologous manner, that is by other stimuli such as the photoperiod or estradiol.

By melatonin Melatonin receptor binding density, primarily MT1 receptor subtype, was strongly increased in the SCN and PT in rodents by decreasing melatonin levels after constant light exposure or pinealectomy, and was decreased after a single melatonin injection, indicating melatonin involvement in the daily regulation of the receptor protein 400, 401. This was conﬁrmed by a strong inverse correlation between melatonin binding capacity and plasma melatonin concentration 400, 402, 403. In GT1-7 gonadotropin-releasing hormone (GnRH)-secreting neurons, gene expression of endogenous MT1 and MT2 melatonin receptors was decreased after short-term exposure to melatonin404 . Mela- tonin is also involved in the regulation of MT1 mRNA in the SCN and PT of rodents 403. The changes in MT1 mRNA and melatonin binding may be regulated, in part, by a cAMP-dependent signaling pathway as shown in primary culture of the ovine pars tuberalis 405. Melatonin also internalizes and desensitizes endogenous MT2 melatonin receptors in the rat suprachiasmatic nucleus, which may be of relevance for deﬁning the periods of sensitivity of the mammalian circadian clock to melatonin.

Heterologous regulation of melatonin receptor In addition to melatonin, clock activity and photoperiod may play an important role in the regulation of the melatonin receptors. Diurnal rhythms of melatonin receptor binding and mRNA levels in the SCN have been reported in previous animal studies 402, 403, 406, 407. Moreover, this diurnal ﬂuctua- tion occurs even in pinealectomized rats, while one hour light exposure was shown to reverse the nocturnal decrease in the SCN 401. Apart from by photoperiod, the MT1 gene expression may also be regulated by the circadian clock, possibly by clock gene proteins acting on a putative enhancer element located on the promoter region of the MT1 gene 403, 408, 409. However, whether the MT1 is a clock-controlled gene needs further study. Circulating estradiol levels are able to down-regulate MT1 receptor in the ovary in immature female rats 410. Moreover, 17beta-estradiol exposure diﬀerentially affect the density and function of hMT1 and hMT2 melatonin receptors expressed in Chinese hamster ovary (CHO) cells (CHO-MT1/CHO-MT2 cells) 411. Furthermore, MT1 receptor expression in the cerebral and caudal arteries is negatively correlated with circulating estradiol levels 412. A correlation has been reported between melatonin receptor expression and its

40 general introduction function. In highly photoresponsive animals such as F344 rats, melatonin has an enhanced effect on photoresponsiveness when measuring endpoints like inhibition of reproduction and weight gain 413. F344 rats have higher levels of melatonin binding in those brain regions that are thought to be involved in these photoresponses compared to less photoresponsive rats 413. Moreover, studies in rodent show that those animals that are more sensitive to the circadian actions of melatonin have higher level of melatonin binding in the SCN 72. In vitro studies also show that melatonin receptor expression levels can dramatically alter the efficacy of melatonin in specific physiological responses 388. Therefore, changes of melatonin receptor expression by multiple factors or due to disease states 414 may serve to prompt certain physiological events in response to specific cues.

Melatonin has substantial therapeutic potentials 415. The most encouraging area for therapeutic uses exploits its circadian and hypnotic effects. The availability of recombinant melatonin receptors will now permit the development of specific agonists and antagonists which may prove beneficial for treating human circadian-based sleep disorders, as seen in jet lag, in blind people, and in shift workers, as well as treating some non-circadian-based sleep disorders. Studies to dissect the different subtypes of melatonin receptors mediating the actions of this hormone will continue to gain a better understanding of the cellular and molecular actions of melatonin, which will hopefully reveal more about ‘the hormone of darkness’.

D. Scope of the thesis

Circadian rhythms, generated by the internal circadian timing system, are an important aspect of human biology 247. The suprachiasmatic nucleus (SCN) in the anterior hypothalamus is considered to be the endogenous biological clock of the mammalian brain (Chapter 1). The circadian rhythm of the pineal gland hormone melatonin is a reliable output of the SCN. Via melatonin receptors, melatonin feeds back on the SCN and is involved in the regulation of the SCN rhythmicity (Chapter 1). Circadian rhythm disturbances, such as sleep-wake disorders, are often seen in aged subjects, depressed patients and take an even more severe form in Alzheimer patients. In this thesis, we focused, therefore, on the neurobiological basis of these circadian rhythm disorders, by investigating the alterations in the SCN-pineal axis during aging, depression, and AD, in particular in the early AD process. Melatonin levels are decreased in aged subjects and in AD patients (reviewed in 202, 203). Interestingly, previous studies from our group showed that melatonin levels in the cerebral spinal ﬂuid (CSF) were already decreased from the earliest Braak stages

41 chapter 1

(i.e. Braak stages I-II) onwards, when subjects are still cognitively intact 302. In order to investigate the mechanisms that underlie the decreased melatonin levels in the AD process, we studied melatonin synthesis and its noradrenergic regulatory pathway in the pineal gland during the progression of AD (Chapter 2). For the first time, we used the approach to measure simultaneously the precursors (tryptophan, 5-HT, NA), products (melatonin, 5-HIAA, MHPG), and gene expression of important enzymes (TPH, AA-NAT1, HIOMT, MAOA, MAOB) and β1-adrenergic receptor in the melatonin synthesis pathway as well as its noradrenergic regulatory pathway in every postmortem human pineal gland. We observed that the circadian melatonin rhythm disappeared and that the nocturnal melatonin levels were decreased in both early (Braak I-II) and late AD patients (Braak VI), confirming our group’s previous study in the CSF. Moreover, we found that the noradrenergic regulation from the SCN was disturbed, and that monoamine oxidase A (MAOA) activity was increased from the earliest preclinical AD stages (Braak I-II) onwards, which may contribute to the changes in melatonin synthesis. In addition to the pineal gland, many different brain areas of AD patients showed MAOA disturbances 306-308, 416-420. A promoter polymorphism of a variable number tandem repeats (VNTR) in the MAOA gene affects MAOA transcriptional activity in vitro 421, 422. In Chapter 3 we assessed the effects of the MAOA-VNTR polymorphism on MAOA transcriptional levels and MAOA activity in the pineal gland as endophenotype, as well as its involvement in the reduced melatonin production in AD. We found that MAOA-VNTR indeed affected MAOA activity in the AD pineal gland, although higher MAOA activity did not result in lower pineal melatonin production in AD. Therefore, the up-regulation of MAOA does not contribute significantly to the decreased melatonin levels in AD, which supports the possibility that the SCN was the crucial component for the alterations we found in the pineal gland in the AD process. Recently, a set of clock genes, the molecular component of the clock, was revealed 101. Clock genes Per1 and Cry2 in the pineal gland of rodents are controlled directly by the SCN via noradrenergic regulation 179, 423, 424. In line with the studies on melatonin synthesis, we investigated the alterations of clock gene oscillation in the pineal gland during the AD process (Chapter 4). Our studies (Chapter 2 and 4) thus combine methods of high performance liquid chromatography (HPLC), radioimmunoassay (RIA) and quantitative polymerase chain reaction (Q-PCR), and yield a systematic investigation of the molecular mechanisms underlying the circadian function of the human pineal gland and the alterations they undergo in Alzheimer’s disease. We found that in both early (Braak I-II) and end stage AD patients (Braak V-VI) the rhythmic expression of clock genes was lost, as well as the correlation between hPer1 and hβ1- ADR, which indicated a disrupted SCN control. Based on our results, we proposed

42 general introduction that the cellular (melatonin level) and molecular (clock genes) changes in the pineal gland in the AD process may be due to a disrupted SCN control. This hypothesis was strongly supported by our subsequent animal experiments, as the alterations in the AD pineal were, remarkably, mirrored in the rat pineal gland devoid of SCN control by SCN lesion or superior cervical ganglionectomy (Chapter 4). Finally, a functional disruption of the SCN was observed from the earliest preclinical AD stages onwards, as shown by decreased vasopressin mRNA levels, a clock-controlled major output of the SCN (Chapter 4). The physiological actions of melatonin are largely based on its action on the level of the hypothalamus and pituitary, which is for a large part mediated by melatonin receptors (Chapter 1). In Chapter 5 we studied, therefore, for the first time, the distribution of MT1 receptor, i.e. the most abundantly expressed melatonin receptor subtype, in the human hypothalamus and pituitary. Moreover, the colocalization of MT1 with AVP, oxytocin, and corticotropin-releasing hormone (CRH) was studied in the human hypothalamus in order to investigate the possible role of the MT1 in various regulations of the hypothalamic physiology. Melatonin is involved in the regulation of circadian rhythms, for a main part by feeding back on the SCN via melatonin receptors (Chapter 1). Melatonin levels are decreased during aging, AD and depressive disorders. Supplementary melatonin is considered to be a potential treatment for circadian rhythm disturbances in these conditions. Using immunocytochemistry, we therefore studied whether the MT1 receptor in the SCN is affected during aging and the progression of AD (Chapter 6), as well as in depression (Chapter 7). An age-related decrease of MT1 in the SCN was observed, which was more severe in late AD patients, but not present in the earliest stages of AD. In contrast, when compared to non-depressed controls, an up-regulation of the MT1 receptor in depression was observed. In Chapter 8, we discussed confounding factors that may be involved in the individual variability in human postmortem studies, i.e. pre-mortem factors, such as age, gender, medicative history and agonal state, and postmortem factors, such as intervals between death and autopsy, and storage time in the freezer or in paraffin. Moreover, we discussed the differences in the molecular mechanisms involved in the regulation of pineal AA-NAT and clock gene oscillation between several species, i.e. avian, rodent, ungulate and human. Furthermore, possible mechanisms for this early-affected SCN in the progression of AD were proposed. Finally, the clinical significance of our studies is discussed, and future experiments are proposed.

Chapter 2

Molecular mechanism changes underlying decreased reduced pineal melatonin levels in Alzheimer disease: alterations in preclinical and clinical stages

Ying-Hui Wu, Matthijs G. P. Feenstra, Jiang-Ning Zhou, Rong-Yu Liu, Javier Sastre Toranõ, Hendrikus J. M. Van Kan, David F. Fischer, Rivka Ravid, and Dick F. Swaab.

The Journal of Clinical Endocrinology & Metabolism 88 (12): 5898-5906, 2003

Abstract A disturbed sleep-wake rhythm is common in Alzheimer Disease (AD) patients and correlated with decreased melatonin levels and a disrupted circadian melatonin rhythm. Melatonin levels in the cerebrospinal fluid (CSF) are decreased during the progression of AD neuropathology (as determined by the Braak stages), already in cognitively intact subjects with the earliest AD neuropathology (Braak stages I-II) (preclinical AD). To investigate the molecular mechanisms behind the decreased melatonin levels, we measured monoamines and mRNA levels of enzymes of the melatonin synthesis and its noradrenergic regulation in pineal glands from 18 controls, 33 preclinical AD subjects, and 25 definite AD patients. Pineal melatonin levels were highly correlated with cerebrospinal fluid melatonin levels. The circadian melatonin rhythm disappeared because of decreased nocturnal melatonin levels in both the preclinical AD and AD patients. Also the circadian rhythm of β1-adrenergic receptor mRNA disappeared in both patient groups. The precursor of melatonin, serotonin was stepwise depleted during the course of AD, as indicated by the up-regulated monoamine oxidase A (MAOA) mRNA and activity (5-hydroxyindoleacetic acid: serotonin ratio). We conclude that a dysfunction of noradrenergic regulation and the depletion of serotonin by increased MAOA result in the loss of melatonin rhythm already in preclinical AD.

Introduction Sleep-wake rhythm disruption and other circadian disturbances are commonly seen in Alzheimer disease (AD) patients 425 and are the most frequent reason for nursing home placement, in fact more often than cognitive impairment 165. The circadian disturbances in AD are accompanied by decreased melatonin levels and a disrupted circadian melatonin rhythm 167, 299, 301. In addition to melatonin’s ability to regulate circadian rhythms 426, 427, melatonin has also been demonstrated to be a potent antioxidant and neuroprotector against oxidative stress and β-amyloid toxicity 267, 268, 428. A recent

47 chapter 2 study reported that melatonin increases survival and inhibits oxidative and amyloid pathology in a transgenic mouse model of AD 304. The decreased levels of melatonin in AD may thus be involved in the pathogenesis of AD. Indeed, recently we have found that melatonin levels in postmortem cerebrospinal fluid (CSF) decrease with the progression of AD neuropathology (as determined by the Braak stages) 302. Inter- estingly, CSF melatonin levels are already reduced in preclinical AD subjects that are cognitively still intact subjects and have only the earliest sign of AD neuropathology (Braak stages I-II) 302. These findings suggest that reduced melatonin levels may serve as an early marker for the very first stages of AD that could so far not be monitored in any other way. However, the mechanisms behind decreased melatonin levels in AD are not clear. Circulating melatonin levels are derived primarily from the pineal gland 429. Tryp- tophan (Trp) is taken up from the circulation and converted to serotonin (5-HT) by tryptophan hydroxylase (TPH). 5-HT is metabolized by the rate-limiting enzyme N-acetyltransferase (NAT) to N-acetyl-5-hydroxytryptamine (NAS), in turn by hydroxyindole-o-methyltransferase (HIOMT) into melatonin. Following its synthesis, pineal melatonin is passively secreted into the circulation. 5-HT can also be oxidized by monoamine oxidase A (MAOA) into 5-hydroxyindoleacetic acid (5-HIAA). Me- latonin synthesis is influenced by light and regulated by the biological clock, i.e. the suprachiasmatic nucleus (SCN), through polysynaptic noradrenergic innervation, which involves binding of noradrenalin (NA) to the β1 −adrenergic receptor on the pinealocytes to activate NAT. NA is mainly deactivated by MAOA into 3-methoxy- 4-hydroxyphenylglycol (MHPG) 251, 430-432 (Fig. 1A). In AD the neurons in the SCN become less active 23, 173, and noradrenergic fibers in the pineal seem to show dystrophic changes 288, which suggest the possibility of a disrupted noradrenergic regulation of the pineal in AD. The present study aimed to clarify the molecular mechanisms underlying the decreased melatonin in preclinical AD and clinical AD. We systematically measured the precursors (tryptophan, 5-HT), products (melatonin, 5-HIAA), and mRNA levels of enzymes (TPH, NAT-1, HIOMT, MAOA, MAOB) in the melatonin synthesis, as well as its noradrenergic regulation (NA, MHPG, β1 −adrenergic receptor mRNA levels) of the pineal glands of controls (Braak stage 0), preclinical AD subjects (Braak stages I-II) and AD patients (Braak stage VI) (Fig.1). We conclude that the lost of melatonin diurnal rhythmicity and the decreased nocturnal melatonin levels, result from the dysfunction of noradrenergic regulation (i.e.

β1-adrenergic receptor mRNA), and the depletion of its precursor 5-HT by increased MAOA from the earliest preclinical AD stages onwards.

48 Pineal melatonin and AD

A. Control TRP

MHPG TPH MAOA 5-HT 5-HIAA NA ATP SCN cAMP NAT

�-adrenergic NAS receptor HIOMT Pinealocyte

Melatonin

B. Early AD stage TRP

MHPG TPH MAOA 5-HT 5-HIAA NA ATP SCN cAMP NAT

�-adrenergic NAS receptor HIOMT Pinealocyte

Melatonin

C. Late AD stage TRP

MHPG TPH MAOA 5-HT 5-HIAA NA ATP SCN cAMP NAT

�-adrenergic NAS receptor HIOMT Pinealocyte

Melatonin

Fig. 1 (A) Pineal melatonin synthesis pathway and its noradrenergic innervation in controls (Braak stage 0); (B) In the preclinical “Alzheimer disease” (AD) subjects (Braak stages i-ii), compared to controls, day/night rhythm of β1-adrenergic receptor mRNA disappears which is responsible for the lack of melatonin diurnal rhythymicity. Nocturnal melatonin synthesis decreases, while the oxidation of its precursor 5-HT into 5-HIAA increases as indicated by the up-regulated MAOA. The metabolism of the noradrenergic regulatory system (i.e. MHPG/ NA ratio) remains constant; (C) In AD patients (Braak stage vi), compared to preclinical “AD” subjects, the conversion of Trp into 5-HT decreases, as indicated by the impaired TPH mRNA levels, which further contributes to the decrease of melatonin levels. denotes circadian rhythmicity. represents lack of circadian rhythmicity. Thicker arrows indicate up-regulated pathway. Thinner arrows denote down-regulated pathway.

49 chapter 2

Materials and Methods Subjects Human brain material is obtained via the rapid autopsy system of the Netherlands Brain Bank which supplies postmortem specimens from clinically well documented and neuropathologically confirmed cases. Autopsies were performed according to the ethical code of conduct of the NBB on donors from whom written informed consent has been obtained either from the donor or direct next of kin. Permission was given for a brain autopsy and for the use of the brain tissue and clinical files for research purposes. Post-mortem pineal glands were obtained at autopsy generally between 1 to 12 hours after death and were immediately frozen in liquid nitrogen and kept at -80 °C until assayed. Neuropathology of all subjects was systematically performed as previously described and the Braak staging was applied 177, 433, 434. Pineal glands were studied from 76 subjects: 18 controls without any primary neurological or psychiatric disease and devoid of the AD neuropathologic changes (i.e. Braak stage 0), 33 cognitively intact cases with minor AD neuropathologic changes (i.e. Braak stages I-II), and 25 AD patients with extensive AD neuropathological changes (i.e. Braak stage VI) 433. Subjects in Braak stages I-II have neurofibrillary changes in the transentorhinal region but did not show any clinical symptoms of AD (i.e. preclinical AD subjects) 433, 435. Patients in Braak stages VI have severe neurofibrillary changes in neocortical area and clinically diagnosis probable AD was performed according to the NINCDS-ADRDA (National Institute of Neurological and Communicative Diseases and Stroke/Alzheimer’s Disease and Related Disorders Association) criteria 436 with exclusion of other possible causes of dementia by history, physical examination and laboratory tests (i.e. definite AD patients). Subjects who used β-adrenergic receptor blockers or antidepressants that might affect pineal melatonin levels 261, 437, 438 were not included. The following variables were included in the present study: age, sex, clock time and date of death, post-mortem delay, brain weight, pineal weight, and CSF-pH (i.e. a measure for agonal state) 439 (Table 1). To determine diurnal variations in pineal melatonin synthesis, according to the clock time of death the subjects were grouped into day group (1000h-2200h) and night group (2200h-1000h) because these periods are known to be associated with circadian differences in the levels of melatonin 440. In addition, to investigate the effect of photoperiods on pineal melatonin synthesis, according to the date of death the subjects were grouped into short photoperiod (23 September-21 March) and long photoperiod (21 March-23 September) as described before 441. Age, sex, CSF-pH, day/night distribution, short/long photoperiod distribution were well matched between the three groups (Table 1).

50 Pineal melatonin and AD

Table 1 Clinical and pathological data for the controls and Alzheimer’s disease patients studied (Mean ± SEM)

Group Age Sex Post- Day/ Photo- Brain wt Pineal wt Pineal total CSF pH (y) (m/f) mort. night period (g) (mg) protein con- delay (h) (long/ tent (mg/ short) mg pineal)

Braak 68±2 18 (12/6) 6.8±0.3 11/7 7/11 1306 ± 30 195.6 ± 21 0.097±0.011 6.7 ±0.09 st. 0

Braak 73±1 33 (13/20) 7.7±0.5 14/19 15/18 1244 ± 24 238.3 ± 22 0.094±0.009 6.7 ±0.04 st. I-II

Braak 72±2 25 (10/15) 4.8±0.3 12/13 13/12 1106 ± 27 244.8 ± 25 0.118±0.015 6.6 ±0.06 st. VI

Notes: According to the time of death, subjects were divided into “day group”: 10:00-22:00, and “night group”: 22:00-10:00. According to the date of death, subjects were grouped into “short photoperiod”: 23 September-21 March and “long photoperiod”: 21 March-23 September.

Sample preparation Each frozen pineal gland was weighed and homogenized in liquid nitrogen. The powder from each pineal was divided into two parts. Part one of the homogenized pineal was weighed and suspended in 0.1M perchloric acid (PCA) (5 µl PCA per mg powder), centrifuged at 12000 x g for 15 min. Next, the supernatant was removed and used for total protein measurement 442, and melatonin, Trp, 5-HT, 5-HIAA, NA, MHPG, dopamine (DA), homovanilic acid (HVA) assay. Part two of the homogenized pineal was used to measure TPH, NAT-1, HIOMT, MAOA, MAOB and β1-adrenergic receptor gene expression using quantitative PCR (Fig. 1).

Radioimmunoassay (RIA) Melatonin levels were measured in the extracts of the postmortem pineal gland by a direct RIA 299, 443. The assay was run in a 0.1 M tricine buffer (Sigma, Zwijndrecht, The Netherlands) containing sodium chloride (0.15 M, Merck, Amsterdam, The Neth- erlands) and 0.1% gelatin (Merck) adjusted to pH 7.5. Iodinated melatonin (2-Io- domelatonin, Amersham IM 215, Roosendaal, The Netherlands) was diluted in tricine buffer, at a final concentration of 25,000 cpm/ml. The melatonin antibody (AB/R/O3, stockgrand, Guildford, UK) that was raised in rabbits is known to be highly specific. It cross-reacts with 6-hydroxymelatonin at 5.3% and at less than 0.2% with 6-sulphatoxymelatonin 444. Standards were diluted in tricine buffer in a dilution range from 1 pg/ml to 1000 pg/ml. Each sample of the extract of the pineal gland was diluted 200 times with tricine buffer and 200 µl was pipetted in tubes and 200 µl anti-melatonin (final dilution 1:200,000) was added. The tubes were vortexed and incubated for 72 hours at 4 °C. Bound melatonin was separated by 50 µl of donkey anti-rabbit antise-

51 chapter 2 rum coupled to cellulose (SAC-CEL, IOS, Boldon, UK). Precipitates were counted in a gamma Counter (Cobra 500s, Packard, Groningen, The Netherlands). The intra-assay coeﬃcient was 8.7 %.

High Performance Liquid Chromatography (HPLC) Fluorescence detection measurement Tryptophan was measured in the extracts of the postmortem pineal gland by HPLC. The system consisted of a PU-1580 pump, an LG-1580-02 ternary gradient unit, a DG- 980-50 3-line degasser, an AS-1555 autosampler and a FP-920 ﬂuorescence detector from Jasco (Maarssen, The Netherlands). Separation was performed on an XTerra C18 150 x 4.6 mm (5 µm) column with an XTerra guard column (Waters, Etten-Leur, The Netherlands). Tryptophan was eluted at 1.0 ml/min with a 0.005 M potassium- dihydrogen phosphate pH 3 buﬀer solution and acetonitrile. All solid chemicals were from Sigma (Zwijndrecht, The Netherlands). Detection was performed at an excitation wavelength of 224 nm and an emission wavelength of 348 nm. Data were acquired and calculated with a Millennium32 (version 3.05) chromatographic data system (Wa- ters). The accuracy and reproducibility of the method were both above 97%. Due to the technical limitation, we were not able to determine the NAS concentration in the supernatant with this system.

Electrochemical Detection Measurement 5-HT, 5-HIAA, NA, MHPG, DA, HVA were measured in the extracts of the postmortem pineal gland by HPLC as described before 445, 446. The chemicals used in this method were from Sigma (Zwijndrecht, The Netherlands). The mobile phase consisted of 10.4 mM citric acid, 6.1 mM sodium acetate, 1.60 mM heptanesulphonic acid, 0.4mM EDTA, 11.8 mM sodium nitrate and 12.5% methanol in water. It was pumped (Shimadzu LC-10ADvp, Den Bosch, The Netherlands) with a flow rate of 1.0 ml/min through a high-efficient pulse dampener and reversed phase Supelcosil LC-18 DB, 250 x 4.6 mm (Supelco, Zwijndrecht, The Netherlands), and similar guard column (25 x 4.6 mm). Injector, pulse dampener and columns were kept at 35 °C in the oven compartment of an ANTEC DECADE electrochemical detector workstation (ANTEC, Leiden, The Netherlands). The column was coupled to a Coulochem 5011 detector cell (Interscience, Breda, The Netherlands). The first electrode was operated at 300 mV, the second (measurement) electrode at -300 mV. Data were acquired and calculated with a Shimadzu Class-VP program (5.03 version). The accuracy and reproducibility were >96% and >97.5%, respectively.

RNA isolation and reverse transcription Each frozen pineal gland was homogenized, and total RNA was isolated with TRIzol

52 Pineal melatonin and AD

Reagent (Invitrogen, Breda, The Netherlands) according to the manufacturer’s instruc- tions 447. The RNA pellet was dissolved in diethylpyrocarbonate-treated distilled water and kept at -20 °C. 1 µg total RNA of each sample was denatured at 70 °C for 10 min and annealed to 250 ng random hexanucleotides (Roche, Almere, the Netherlands), then reverse transcribed to cDNA by superscript II reverse transcriptase (Invitrogen) in 5 min at 30 °C, 5 min at 37 °C, and 90 min at 42 °C in a ﬁnal volume of 20 µl, in the presence of RNAse inhibitor (Invitrogen). The cDNA was stored at –20 °C until further use.

Quantitative polymerase chain reaction (QPCR) QPCR was carried out in a final volume of 20 µl in 96-well plates, using the SYBR Green PCR kit (Applied Biosystems, California, USA) containing 2 µl 10 × SYBR Green PCR Buffer, 1.6 µl MgCl2 (25 mM), 1.5 µl dNTP Blend (2.5 mM dATP, 2.5 mM dCTP, 2.5 mM dGTP, 5.0 mM dUTP), 0.14 µl AmpErase UNG (1 U/µl), 0.1 µl Ampli Taq Gold (5 U/µl), 0.5 µl cDNA sample (5 ng total RNA), and 3.0 µl mixture of sense and antisense primers (each primer 2 pmol/µl). Cycling conditions were: 2 min 50 °C; 10 min 95 °C; 40 cycles of 15 sec 95 °C and 1 min 60 °C. The data were acquired and processed automatically by Sequence Detection Software (Applied Biosystems). The key enzyme of melatonin synthesis NAT has NAT-1 and NAT-2 subtypes 448, 449. Our preliminary data showed only NAT-1 transcripts in the human pineal gland, which is in accordance with in situ hybridization studies 450. MAO has two subtypes, MAOA and MAOB. MAOA is responsible for 5-HT and NA metabolisms 451. In the present study, we measured both MAOA and MAOB gene expression. Two reference genes were selected from a study of multiple adult and fetal tissues: EF-1-alpha and E2 ubiquitin conjugating enzyme 452, and were measured in all the samples to normalize expression data. The primers were designed with Primer Express software (Applied Biosystems). The efficiency of each primer pairs was calculated using cDNA dilution curves and linear regression. Details of the primers, the GenBank accession numbers and the efficiency of each primer pairs are given in Table 2. The mRNA expression levels of the two reference genes were highly correlated in our samples (r = 0.913; P < 0.0001) and were similarly expressed in the three groups (P > 0.1, P > 0.17, respectively). The amount of every target gene is calculated by raising the primer efficiency of the gene to the power of – CT (cycle threshold), normalizing this and dividing by the average of the two normalized housekeeping gene expressions.

Statistics Differences of monoamine levels and target gene expression levels among the three groups were tested by the Kruskal-Wallis Test. Differences between groups were tested using the Mann-Whitney U-test. The differences in proportion between males and

53 chapter 2

Table 2 GenBank accession code, sequence of PCR primer pairs for the target genes and reference genes, and ampliﬁcation eﬃciency of each primer airs.

Gene Accession Forward Primer Reverse Primer Amp. Code eﬃc.

NAT-1 D90041 AGATGTGGCAGCCTCTGGAG GCACCTGAGGCTGATCCTTC 1.90

HIOMT U11090 CAGGTGGTGGCATTCTGGTA CCTCGCCTGTCTTCATCCA 2.00

MAOA M68840 TTCTGGCCTGCTGAAGATCAT CCCAGGGCAGTTACTGATGTG 1.95

MAOB M69177 TTTTCAGCAACGGCTCTTGG CACAAGTAGCCCCCTTTTGTG 2.08

β1-adrener- NM_000684 CCCACAATCCTCGTCTGAATC AGGAACATCAGCAAGCCACTC 1.94 gic receptor

TPH X52836 CTCTTAGGTCATGTCCCGCTTT GGAGAATTGGGCAAAACTAGGTT 1.90

EF-1-alpha J04617 AAGCTGGAAGATGGCCCTAAA AAGCGACCCAAAGGTGGAT 1.95

E2 ubiquitin U39317 CTGAAGAGAATCCACAAGGAATTGA CTCCAACAGGACCTGCTGAAC 1.94 females, the number of subjects that died during the day and night, and the number of subjects that died during the short photoperiod and long photoperiod for the three groups were tested by Chi-square. Correlations were analysed by the Spearman correlation test. Diﬀerences were considered statistically signiﬁcant at the P < 0.05 level (two-tailed).

Results No significant difference was found in pineal weight and pineal total protein content between the three sample groups (Braak stage 0, stages I-II and stage VI) (P = 0.62, P = 0.215) (Table 1). No significant correlation was observed between post-mortem delay, brain weight on the one hand, and all the monoamine levels and gene expression levels measured on the other.

Melatonin levels in the CSF closely reflect the pineal melatonin content We compared the data from the 40 subjects whose melatonin levels were measured both in the CSF 302 and in the pineal gland (present study). A highly positive correlation between melatonin levels in the CSF and in the pineal was found (r = 0.83, P < 0.0001, n = 41) (Fig. 2), indicating that the CSF alterations in melatonin levels reﬂected the changes in the pineal melatonin content.

Changes in the pineal melatonin synthesis (Fig. 1) The day/night diﬀerences of melatonin and melatonin/5-HT ratio (i.e. melatonin synthesis activity) found in Braak stage 0 (P = 0.012, P = 0.001, respectively) had

54 Pineal melatonin and AD

1200

1000

800 N = 41 600 R = 0.83 P < 0.0001 400

200

CSF melatonin levels (pg/ml)0

0 100 200 300 400 500 600 Pineal melatonin levels (pg/mg tissue) Fig. 2 The significant positive correlation between melatonin levels in the cerebrospinal fluid (CSF) and in the pineal gland. disappeared in Braak stages I-II and Braak stage VI (Fig. 3, A and B). No day/night difference of other monoamines or mRNA levels of the enzymes involved in melatonin synthesis was found in the three groups, nor any photoperiodic difference in monoamines or enzymes mRNA levels in melatonin synthesis. Nocturnal melatonin levels decreased in Braak stages I-II and Braak stage VI (Table 3) (Fig. 3A). Nocturnal melatonin/5-HT (i.e. melatonin synthesis activity) decreased in Braak stage VI compared to Braak stage 0 (P = 0.012) (Table 3 and Fig. 3B). NAT-1 mRNA levels showed a trend to increase in Braak stage VI (P = 0.054) (Table 3 and Fig. 3C), while HIOMT mRNA levels seemed to decrease in Braak stage VI (P = 0.068) (Table 3). 5-HIAA levels, the oxidative product of 5-HT by MAOA, were higher in Braak stage VI compared to Braak stage 0 (P = 0.003) and Braak stages I-II (P = 0.018) (Table 3 and Fig. 3D). Moreover, the MAOA mRNA levels and 5-HIAA: 5-HT ratio (i.e. MAOA activity) were step-wise increased in Braak stages I-II (P = 0.037, P = 0.040, respectively) and Braak stage VI (P < 0.0001, P = 0.007, respectively) compared to Braak stage 0 (Table 3 and Fig. 3, E and F). The 5-HIAA: 5-HT ratio (i.e. MAOA activity) and MAOA mRNA levels correlated positively (r = 0.370, P = 0.001, n = 76). In addition, MAOB mRNA levels were increased in Braak stage VI compared to Braak stage 0 (P = 0.024) and Braak stage I-II (P = 0.015) (Table 3). No correlation between the 5-HIAA: 5-HT ratio and MAOB mRNA levels was found in the three groups. The concentration of Trp, the precursor of 5-HT and melatonin, was higher in Braak stage VI compared to Braak stage 0 (P = 0.025) and Braak stages I-II (P = 0.004) (Table 3 and Fig. 3G). The mRNA levels of TPH, the key enzyme for the conversion

55 chapter 2

p A 400 p=0.012 p=0.018 B p=0.001 =0.012 p=0.042 0.030

) 350

u 0.025 day

s night i 300

- 0.020

250 5

( 200 n

s 0.015

e 150

M 0.010

n 100

a 0.005

e 50

M 0 0.000 Braak 0 Braak I-II Braak VI Braak 0 Braak I-II Braak VI p=0.003 C 1.6 D p=0.018 3000

)

i e 2500

t r 1.2

/ 2000

(

0.8 s

l 1500

a l 1000

0.4 A

H T 500

N 0.0 0 Braak 0 Braak I-II Braak VI Braak 0 Braak I-II Braak VI

E F 0.40 p=0.007 P<0.001 P=0.006

n 1.6

0.30 s e P=0.037

r T p=0.04

x 1.2

n A 0.20

g I 0.8

0.10 e r 0.4

0.00 M 0.0 Braak 0 Braak I-II Braak VI Braak 0 Braak I-II Braak VI G H p=0.020 p=0.022 5000 p=0.004

)

e p=0.027

s o 1.6

i 4000

e 1.2 g 3000

(

v 0.8

2000 i

p 1000 H 0.4

r T 0 0.0 Braak 0 Braak I-II Braak VI Braak 0 Braak I-II Braak VI

56 Pineal melatonin and AD

Table 3 Mean concentrations of monoamines and mean target gene relative expression levels in the pineal glands of Braak stages i-ii and Braak stage vi groups, and the percentages of Braak stage 0 group (controls) Braak stages I-II Braak stage VI Diﬀerence betw. Braak stages Mean±SEM % Mean±SEM % I-II and VI Trp 2996±531 87 5890±844 170 * ** 5-HT 7348±1131 82 9152±1204 102 Melatonin # 46.0±13.2 20 * 35.3±16.8 15 * 5-HIAA 1336±267 142 2593±466 276 ** * 5-HT/Trp 4.5±1.0 109 2.4±0.5 60 Melatonin/5-HT 0.012±0.003 59 0.007±0.004 32 * 5-HIAA/5-HT 0.21±0.03 162 * 0.31±0.05 238 **

NA 61.4±8.9 124 63.9±7.6 129 MHPG 112.5±13.8 88 179.0±29.1 139 * * MHPG/NA 4.2±0.9 100 3.6±0.6 86

DA 16.2±5.5 107 12±1.7 79 HVA 185.8±16.6 110 198.3±25 116 HVA/DA 31.3±7.0 136 22.4±3.8 96

TPH 1.35±0.25 86 0.80±0.17 51 * * NAT-1 1.02±0.09 94 1.41±0.18 129 MAOA 1.14±0.11 146 * 1.52±0.72 195 * * * * MAOB 0.93±0.15 109 1.41±0.19 166 * * HIOMT 0.98±0.10 77 0.76±0.07 60 β1-adrenergic rec. 1.06±0.10 108 1.13±0.10 122

Note: Mean values of monoamines are expressed as picograms per milliliters (Mean ± SEM). Genes relative expression data are expressed as normalized date (Mean ± SEM). * p<0.05, * * p<0.01. % = percentage of the respective mean values of the control group (Braak stage 0). # nighttime value.

Fig. 3 (A) Day/night diﬀerence of melatonin is present in Braak stage 0, but disappears in Braak stages i-ii and Braak stage vi. Nocturnal melatonin levels are decreased in Braak stages i-ii and Braak stage vi; (B) Day/night diﬀerence of melatonin/5-HT (representing the melatonin synthesis activity) is present in Braak stage 0, but is lost in Braak stages i-ii and Braak stage vi. Melatonin/5-HT decreases in Braak stage vi compared to Braak stage 0; (C) NAT-1 gene expression tends to be increased in Braak stage vi (Kruskal-Wallis Test, p=0.054) (D) The levels of 5-HIAA, the oxidative product of 5-HT, are elevated in Braak stagevi, (E) 5-HIAA/5-HT, representing the activity of MAOA, is increased in Braak stages i-ii and Braak stage vi; (F) MAOA gene expression is stepwise increased in Braak stages i-ii and Braak stage vi, (G) Tryptophan levels are increased in Braak stage vi; (H) The gene expression of TPH is reduced in Braak stage vi.

57 chapter 2 of Trp to 5-HT, were lower in Braak stage VI compared to Braak stage 0 (P = 0.022) and Braak stages I-II (P = 0.027) (Table 3 and Fig. 3H). No signiﬁcant diﬀerence of 5- HT or the 5-HT: Trp ratio was found between the three groups (P = 0.193, P = 0.326, respectively) (Table 3).

Dysregulated noradrenergic system (Fig.1)

A day/night difference of β1 −adrenergic receptor mRNA levels was present in Braak stage 0 (P <0.001), but disappeared in Braak stages I-II and Braak stage VI (P = 0.135, P = 0.174, respectively) (Fig. 4). There were no day/night differences in NA or MHPG concentrations in the three groups. No photoperiodic difference in either of these factors was found in the three groups. MHPG concentration in Braak stage VI was higher than Braak stage I-II (P = 0.001) and insignificantly higher than Braak stage 0 (P = 0.09). No significant difference of

NA, MHPG: NA ratio (i.e. NA metabolic activity) and β1 −adrenergic receptor mRNA levels (P = 0.47, p=0.26, P = 0.437, respectively) was found between the three groups (Table 3 and Fig. 5).

β1-adrenergic receptor mRNA levels were correlated with melatonin levels (r = 0.45, P < 0.0001, n = 76) and melatonin/5-HT (i.e. melatonin synthesis activity) (r = 0.37, P < 0.001, n = 76).

n Day Night

essio 2.0 r

xp 1.8 e

e 1.6 p<0.001

elativ 1.4 r

e 1.2 gen

r 1.0

0.8 ecepto

r 0.6 c gi

r 0.4

ene 0.2 r d a

- 0.0 Braak 0 Braak I-II Braak VI

Fig. 4 Day/night rhythm of β1-adrenergic receptor gene expression in Braak stage 0 disappears in Braak stages i-ii and Braak stage vi. Bar indicates the median of β1-adrenergic receptor gene expression. Each single point indicates the β1-adrenergic receptor gene expression of one subject. Note the considerable variations within the group, especially in Braak stages i-ii and Braak stage vi.

58 Pineal melatonin and AD

Dopaminergic system No day/night or photoperiodic differences in either of DA, HVA concentrations nor HVA: DA ratio (i.e. DA metabolic activity) were present in the three groups. No significant differences of these parameters were found in the three groups indicating that the dopaminergic system is not involved in the changes of melatonin in AD (Table 3).

Discussion We have taken a comprehensive approach to study changes in the melatonin synthesis pathway during Alzheimer’s disease. As could be expected, this approach for instance yields a signiﬁcant positive correlation between the mRNA level of MAOA and the ratio of the substrate and product of MAOA, 5-HT and 5-HIAA. Moreover, we observed positive correlations between steps in the metabolic pathway that are much more distant such as between the mRNA level of the β1 −adrenergic receptor and melatonin. This suggests that melatonin synthesis is mainly regulated by noradrenergic regulation in humans and justiﬁes our approach. More novel observations were made in our study. In both Braak stages I-II and Braak stage VI, there was a shift in melatonin synthesis pathway: an increased oxidation of 5-HT to 5-HIAA with upregulated MAOA and an impaired conversion of 5-HT to melatonin, in addition to the dysregulated

β1 −adrenergic receptor mRNA, which are responsible for the decreased nocturnal melatonin synthesis and the lost of melatonin diurnal rhythm in the preclinical AD subjects and AD patients (Fig.1, B and C). The decreased TPH mRNA levels in Braak stage VI may further decrease melatonin synthesis in AD patients (Fig. 1C). Also, the

Braak 0 Braak I-II

)

e p=0.001

u Braak VI s 200 200

t 160 160

/ 120 120

( 80 80 H

e 40 40 G

8 8 A

P 6 6

H 4 4

, 2 2

A 0 0

N -2 -2 NA MHPG MHPG/NA

Fig. 5 MHPG levels in Braak stage vi are higher than in Braak stage i-ii (p=0.001) and insignificantly higher than in Braak stage 0 (p=0.09). No difference in NA levels and MHPG/NA (reflecting the metabolic activity of noradrenergic system) between Braak stage 0, Braak stages i-ii and Braak stage vi.

59 chapter 2 observed highly positive correlation between melatonin levels in the pineal and in the CSF indicates that the decreased CSF melatonin levels in AD are due to the reduced pineal melatonin synthesis rather than the dilution of the CSF in AD 453 (Fig. 2). The diurnal rhythm of pineal melatonin disappeared and the nocturnal melatonin levels decreased in Braak stages I-II and Braak stage VI (Fig. 3A), which is in full agreement with the reduced melatonin levels in preclinical AD subjects 302 and AD patients 167, 297. Moreover, these findings may explain why earlier studies 299, 301 did not find melatonin circadian rhythm in aged controls which were not Braak staged and will thus have been combinations of Braak stage 0 controls and preclinical AD subjects. Although the latter group did not show the clinical symptoms of AD, the diurnal melatonin rhythm was already strongly diminished (Fig. 3A). The day/night rhythm of melatonin synthetic activity (i.e. melatonin/5-HT) found in Braak stage 0 had disappeared in Braak stages I-II and Braak stage VI (Fig. 3B). However, neither AD-related changes nor day/night rhythms of NAT-1 mRNA or HIOMT mRNA were found in the present study. These observations are supported by studies in the rhesus macaque 253, which showed that NAT activity plays a crucial role in the melatonin rhythmicity, while NAT activity is not regulated by changes in NAT mRNA 252, but rather by the posttranscriptional control, e.g. proteasomal proteolysis 454. Whether the posttranscriptional regulation of NAT is affected in the progression of AD remains to be clarified. In contrast to the decreased melatonin production from 5-HT, the oxidation of 5-HT to 5-HIAA was strongly and stepwise increased, as indicated by the elevated MAOA activity (i.e.5-HIAA: 5-HT ratio) and mRNA levels in Braak stages I-II and Braak stage VI (Fig. 1, B and C, and 3, E and F). Increased MAOA activity and mRNA levels as found in the pineal gland in the present study seem to be a general phenomenon in AD, since it was also reported in the cortex, thalamus, hypothalamus and white matter of AD patients 306-308. Interestingly, MAOA gene polymorphisms are suggested to be associated with an increased susceptibility for AD 309. The elevated 5-HIAA levels found in the pineal of AD patients differ from the decreased 5-HIAA levels in cortex, amygdala and caudate nucleus of AD patients 416, 455. This difference may be related to the fact that the predominant distribution of 5-HT in the pineal is cytosolic whereas 5- HT is stored in subcellular vesicles in the midbrain system 456. Therefore, in the pineal, 5-HT may be more vulnerable to the oxidation by MAOA to 5-HIAA than in the rest of the brain. Our data suggest that reduced melatonin production in preclinical AD subjects and AD patients may be due to the depletion of its precursors 5-HT, caused by the upregulation of MAOA. In fact, MAOA inhibitors significantly increase serum melatonin levels in human and rodents 437, 438, 457, 458.

In the present study, a day/night rhythm of β1−adrenergic receptor mRNA with elevated levels at night, present in Braak stage 0, was absent in Braak stages I-II and

60 Pineal melatonin and AD

Braak stage VI (Fig. 5). This finding suggests that the dysregulation of β1−adrenergic receptor mRNA is the basis of the lack of day/night rhythms of melatonin we observed in Braak stages I-II and Braak stage VI. It has been shown in rats that the levels of pineal β1−adrenergic receptor mRNA are decreased and its diurnal rhythm is abolished upon removal of the sympathetic innervation from the SCN 459, 460. Our earlier observations have revealed a marked decrease of vasopressin expression neuron numbers, activity and circadian rhythmicity in the SCN of AD patients 23, 169, 173. Taking these data together, we hypothesize that the circadian fluctuations of the SCN are affected already in the earliest preclinical stages of AD, which results in a dysregulation of β1 −adrenergic receptor mRNA and thus in a decrease of nocturnal melatonin synthesis and the disappearance of the diurnal melatonin rhythm. Studies on the circadian rhythm of the SCN in the first Braak stages should be performed to confirm this idea. No day/night rhythm of NA, MHPG or MHPG/NA in the pineal gland was found either in Braak stage 0, stages I-II or stage VI. This may well be explained by the fact that NA and MHPG levels in the homogenized pineal we measured are the dilution of the levels in the noradrenergic terminal which directly reflect the day/night stimulus from the SCN. MHPG concentration was increased in AD patients probably because of the increased MAOA, while the metabolism of the NA system remained constant, as indicated by the constant MHPG/NA in Braak stages I-II or Braak stage VI (Table 3) (Fig. 4). Although the groups did not show a difference in possible confounding factors such as age, gender, CSF pH (a measure for agonal state) 439, 461, a limitation of the present postmortem study is that it is not known whether the clinical condition of the patients in the various groups, including the circumstance and the time of death, may have influenced the results.

In summary, the dysregulation of pineal β1 −adrenergic receptor mRNA, and the increased MAOA activity and mRNA levels, are held responsible for the disappearance of the melatonin diurnal rhythm and the decrease of nocturnal melatonin synthesis in preclinical AD subjects and AD patients. In addition, the decreased TPH mRNA levels may further contribute to this change in AD patients (Fig. 1). Our finding of a lack of circadian β1 −adrenergic receptor mRNA rhythm in Braak stages I-II suggests that the first alterations may take place in the SCN in the earliest preclinical stages of AD pathology. These findings support the possibility of reduced melatonin levels as an early marker for the onset of AD and provide a basis for a mechanism behind bright light therapy to restore circadian rhythm disorders in AD 322. Furthermore, as the loss of melatonin rhythmicity already occurs in people with early AD-neuropathology, before clinical symptoms occurs, it may be beneficial to supplement melatonin in case of decreased

61 chapter 2 nocturnal melatonin levels in order to slow down the development of AD. The recent ﬁnding that melatonin increases survival and inhibits amyloid pathology in an Alzheimer mouse model 304 supports this possibility. Whether the increased MAOA, a general phenomenon in the AD brain 306-308, contributes to the symptoms of AD, e.g., to depression or to the pathogenesis of AD, demands further investigation. In that case, therapeutic use of MAOA inhibitors in AD may be considered.

62 Chapter 3

A promoter polymorphism in the monoamine oxidase A gene is associated with the increased MAOA activity in Alzheimer’s disease

Ying-Hui Wu, David F. Fischer, Dick F. Swaab

Submitted

Abstract Monoamine oxidase A (MAOA) is thought to be involved in the pathogenesis of mood disorders and Alzheimer’s disease (AD). MAOA activity and gene expression have been found to be up-regulated in different brain areas of AD patients, as well as in the pineal gland, which may contribute to the reduced melatonin production in AD. A promoter polymorphism of a variable number tandem repeats (VNTR) in the MAOA gene has been shown to affect MAOA transcriptional activity in vitro. Here we examined in 63 aged controls and 44 AD patients the effects of the MAOA-VNTR on MAOA gene expression and activity in the pineal gland as endophenotypes, and on melatonin production. We found AD patients carrying long MAOA-VNTR genotype (consisting of 3.5- or 4-repeat alleles) showed higher MAOA gene expression and activity than the short-genotyped (i.e. 3-repeat allele) AD patients. Moreover, the AD-related up-regulation of MAOA showed up only among long-genotype bearing subjects. There was no significant effect of the MAOA-VNTR on melatonin production either in controls or in AD patients. Our data suggest that the MAOA-VNTR is associated with MAOA dysfunction in AD. Endophenotypes such as activation of MAOA may be linked to co-morbidities of AD, for instance depression.

Introduction Monoamine oxidase A (MAOA) is an important catabolic enzyme that regulates monoamine transmitter levels in the central nerve system, including serotonin, dopamine and noradrenalin. The gene encoding MAOA is located on human chromosome Xp11. It is a candidate gene for human neurological diseases and psychiatric and behavioral traits (see review 462). MAOA is suggested to be involved in the pathogenesis of Alzheimer’s disease 306, 307, 418-420. Increased activity and gene expression levels of MAOA have been observed in locus coeruleus, hypothalamus, parietal cortex, occipital cortex, temporal cortex and frontal cortex of AD patients 306-308, 416, 417, which are also reported in the pineal gland of AD patients in our recent study 254. Pineal gland mainly secretes melatonin, which exhibits a circadian pattern and is involved in the regulation of circadian rhythms (see reviews 251, 431). Melatonin production is dramatically reduced in AD patients, which

65 chapter 3 may be responsible for their circadian rhythm disturbances (reviewed in 202, 203). In the pineal gland MAOA oxidizes serotonin (5-HT), the precursor of melatonin 251, into 5-hydroxyindoleacetic acid (5-HIAA). Treatment of MAOA-selective inhibitor has shown to result in an increase of melatonin production in human, probably due to increased availability of serotonin 437. Contrary, the up-regulation of MAOA in the pineal gland of AD patients may decrease the availability of serotonin from melatonin synthesis pathway, and contribute to the reduction of melatonin in AD 254. Recently, a frequent polymorphism of a variable number tandem repeats (VNTR) in the promoter region of MAOA has been identified by Sabol et al., which consists of a 30-bp repeated sequence present in 3, 3.5, 4, or 5 copies 421. The MAOA-VNTR polymorphism has shown to affect MAOA transcription and hence activity in vitro experiments 421, 422, 463. The MAOA-VNTR alleles with 3.5 and 4 repeats of a 30-bp sequence transcribe MAOA more efficiently than those with 3 repeats in different cell lines and human male skin fibroblasts 421, 422, 463. Functional promoter polymorphisms may confer susceptibility to disease or disease-related disturbances, for example in the APP or presenilin genes in Alzheimer disease (AD) (for review see 464). In the present study, we assessed the functional effects of the MAOA-VNTR polymorphism on MAOA transcriptional levels and MAOA activity in the pineal gland in control and AD as endophenotypes 465. Moreover, we investigated the possible role of the MAOA-VNTR in the melatonin synthesis. Our findings suggest that the MAOA- VNTR is associated with the increased MAOA activity in AD.

Experimental Procedure Subjects Human brain material of caucasian subjects was obtained via the rapid autopsy system of the Netherlands Brain Bank (NBB), which supplies postmortem specimens from clinically well-documented and neuropathologically conﬁrmed cases. Autopsies were performed according to the ethical code of conduct of the NBB on donors from whom or from whose direct next of kin written informed consent had been obtained for a brain autopsy and for the use of the brain tissue and clinical ﬁles for research purposes. Systematic neuropathology of all subjects was performed as previously described 177 and the Braak staging was applied 433, 434. Postmortem cerebellum and pineal gland from 130 subjects were collected in our studies: 73 cognitively intact controls without any primary neurological or psychiatric disease, with no or very few AD neuropathological changes (i.e. Braak stage 0 and Braak stages I-II) 433, and 57 AD patients with the clinical diagnosis “probable AD” according to the NINCDS-ADRDA criteria 436, and with extensive postmortem AD neuropathological changes (i.e. Braak stage V-VI) 433. They were matched for sex, age, the CSF-pH (which measures agonal state) 439. Moreover, as pineal melatonin

66 MAOA polymorphism and AD synthesis exhibits a day/night rhythm 202, 254, patients and controls were also matched for day/night distribution according to their time of death (10am-10pm: day).

Sample preparation DNA was isolated from each piece (20 – 60 mg) of frozen cerebellum using the QIA- amp DNA mini kit (QIAGEN, Westburg, b. v., The Netherlands) for genotype analysis. The sample preparation for frozen pineal gland was described before 254. Brieﬂy: it was weighed, homogenized in liquid nitrogen. Part of the homogenized pineal was weighed, suspended into 0.1M perchloric acid (PCA) (5 µl PCA per mg tissue powder), and centrifuged. The supernatant was used for measuring melatonin by radioimmunoassay, and for measuring serotonin (5-HT), and the metabolite of serotonin, 5-hydroxyindoleacetic acid (5-HIAA) by high performance liquid chromatography (HPLC). Part two of the homogenized pineal was used to isolate RNA and to measure MAOA gene expression using quantitative PCR.

Genotype analysis MAOA VNTR genotyping assay The following primer set for MAOA VNTR polymorphism in promoter region has been used: forward: GCCCAGGCTGCTCCAGAAA and reverse: CGGGACCT- GGGCAGTTGT. PCR was carried out in a final volume of 50 µl containing 100 ng genomic DNA, 100 µM dNTPs, 5 pmol of each primer, and 2 mM MgCl2, 0.25 U Taq DNA polymerase (HT, Biotechnology Ltd, Cambridge, UK). Cycling conditions were: 94 °C for 2 min, followed by 35 cycles of 94 °C (20 sec), 58 °C (30 sec), and 72 °C (30 sec), and 72 °C for 7 min. The PCR products were separated by electrophoresis on 3% Nusieve GTG agarose gels (Rockland, ME, USA) and visualized by ethidium- bromide staining. We used a 25 bp DNA ladder (Invitrogen, Breda, the Netherlands) to distinguish the size of PCR products. Sequence analysis on an ABI 3730 (Applied Biosystems, Foster City, USA) confirmed the obtained PCR products as being specific for the MAOA promoter polymorphism studied.

Apolipoprotein E (ApoE) genotyping assay ApoE is the most well-known polymorphism that associates with AD 466. In this study, served as a negative control for MAOA genotype-endophenotype association, ApoE genotype was determined and its relationship to MAOA expression in control and AD subjects was assessed. The ApoE genotype was determined by PCR ampliﬁcation using the primers: Forward: TCCAAGGAGCTGCAGGCGGCGCA Reverse: ACAGAATTCGCCCCGGCCTGGTACATCGCCA Then the PCR product was digested by CfoI and fragments were separated by electrophoresis in a 5% agarose gel 467. 67 chapter 3

Radioimmunoassay Melatonin levels were measured in the extracts of the postmortem pineal gland by a direct radioimmunoassay as extensively described before 254.

HPLC with electrochemical detection 5-HT, 5-HIAA levels were measured in the extracts of the postmortem pineal gland by HPLC as extensively described before 254. The chemicals used in this method were from Sigma (Sigma, Zwijndrecht, The Netherlands). From the ratio of 5-HIAA to 5- HT, the MAOA activity was estimated.

RNA isolation, reverse transcription and Quantitative PCR (Q-PCR) Total RNA was isolated from each frozen pineal gland with TRIzol Reagent (Invit- rogen). The procedures of reverse transcription and quantitative PCR are extensively described before 254. Brieﬂy: single-stranded copy DNA (cDNA) was synthesized from 1µg total RNA by superscript II reverse transcriptase (Invitrogen). QPCR was carried out using the SYBR Green PCR Master Mix (Applied Biosystems) containing 0.5µl cDNA sample (from 5 ng total RNA), and 3.0 µl mixture of sense and antisense primers (2 pmol/µl each primer). Cycling conditions were: 2 min 50 °C; 10 min 95 °C; 40 cycles of 15sec 95 °C and 1 min 60 °C. The data were acquired and processed automatically by Sequence Detection Software (Applied Biosystems). Two reference genes were selected from a study of multiple adult and fetal tissues: elongation factor 1alpha (EF1) and ubiquitin-conjugating enzyme E2U 452, and were measured in all the samples to normalize expression data. The primers were designed with Primer Express software (Applied Biosystems). The sequences are as follows: MAOA: Forward: TTCTGGCCTGCTGAAGATCAT Reverse: CCCAGGGCAGTTACTGATGTG EF1: Forward: AAGCTGGAAGATGGCCCTAAA Reverse: AAGCGACCCAAAGGTGGAT E2U: Forward: CTGAAGAGAATCCACAAGGAATTGA Reverse: CTCCAACAGGACCTGCTGAAC

Statistics Differences of monoamine levels and target gene expression levels between groups were tested using the Mann-Whitney U-test. Distribution analysis was tested by Chi- square. Correlations were analyzed by the Spearman correlation test. Data are presented as mean ± standard error of the mean (Mean ± SEM). Differences were considered statistically significant at the P < 0.05 level (two-tailed).

68 MAOA polymorphism and AD

Results We found four different alleles corresponding to the 3, 3.5, 4, 5 repeats of the MAOA- VNTR in our samples. In agreement with previous studies 468, 469, the “3” and “4” alleles were most frequent among the samples. Indeed, we only found three individual with 3.5 repeats, one individual with 5 repeats, and none with 2 repeats. Two MAOA-VNTR genotype groups, i.e. the short and long genotype groups, were formed on the basis of the functional experiments, showing that 3.5 and 4 repeats alleles induce higher transcriptional efficiency than the 3 repeats allele 421, 422, 463. The functional effect of 5 repeats allele is contradictory in literature 421, 463, and only one case with 5 repeats allele was found in our samples, therefore 5 repeats allele was excluded in the analysis. The short genotype was thus composed of all alleles with 3 repeats; the long genotype contained the alleles of 3.5 repeats and/or 4 repeats. As the MAOA gene is on the X chromosome, in females one allele should be randomly silenced because of X-inactivation. X-inactivation occurs early during development. A given organ or tissue can thus show expression of one allele exclusively, and not a mixture of both. For MAOA heterozygous females containing both low and high activity alleles, it is uncertain which allele is inactivated in the pineal gland (a relatively small sub-structure of the brain) and which genotype group they should be classified. Therefore we have excluded them (10 control females and 13 AD females) from the analysis. Our genotype-endophenotype association study thus contains 107 subjects, including 63 controls and 44 AD patients, who were matched for age, sex, pH of the cerebrospinal fluid (CSF) and day/night distribution (Table 1). Although postmortem delay (PMD)

Table 1 Clinico-pathological data and MAOA-VNTR genotyping for the controls and Alzheimer’s disease patients in MAOA-VNTR genotype-endophenotype analysis. Control (n=63) AD (n=44) MAOA-VNTR genotype short long short long Number of cases 19 44 12 32 Day/night distribution 8/10 19/25 7/5 13/19 Age (years) 70.5± 2.2 71.2± 1.6 72.3±2.2 73.5±1.8 Brain weight (grams) 1297± 34.5 1274± 18.5 1111± 63.7* 1114± 20.7* PH of CSF 6.63± 0.08 6.55±0.17 6.8±0.09 6.6±0.05 PMD (hours) 7.3±0.8 8.8± 0.8 4.4± 0.4* 5.2± 0.4*

Note: Values are the Mean± SEM. Short genotype contains only 3 or 3/3 alleles; Long genotype comprises only 3.5, and/or 4 alleles. Day: 1000-2200h; night: 2200-1000h. CSF: cerebrospinal ﬂuid; PMD: postmortem delay. * p<0.05 between control subjects and AD patients.

69 chapter 3 was different in control and AD (Table 1), subsequent liner regression analysis revealed that it did not affect MAO gene expression (r = 0.1, P = 0.2) or 5-HIAA:5-HT ratio (representing the MAOA activity) (r = 0.1, P = 0.3). In AD patients, the long genotype showed higher MAOA gene expression than the short genotype (1.35 ± 0.11 and 1.02 ± 0.13, P = 0.036) (Fig. 1A). Moreover, the long genotype presented higher 5-HIAA: 5-HT ratio compared to the short genotype in AD patients (0.35 ± 0.07 and 0.16 ± 0.04, P = 0.039) (Fig. 1B). However, in controls, there was no significant difference of MAOA gene expression between long and short genotypes (1.09 ± 0.09 and 1.23 ± 0.18, P = 0.45) (Fig. 1A), as well as the 5-HIAA: 5-HT ratio (0.17 ± 0.02 and 0.21 ± 0.05, P = 0.25) (Fig. 1B).

A B 0,45 1,4 0,40 n o i 1,2 s 0,35 s e T r 1,0 p H 0,30 - x 5 e / 0,25 e 0,8 A n A I e 0,20 g H

0,6 - A 5 0,15 O

A 0,4

M 0,10 0,2 0,05 0,0 0,00 Short Long Short Long Short Long Short Long Controls AD patients Controls AD patients

C 140 ) e u s s

i 120 t g m / 100 g

p Fig. 1 A. Monoamine oxidase A (MAOA) ( s

l gene expression; B. The ratio of 5-HIAA to 80 e

v 5-HT (represents the activity of MAOA); C. e l

n Melatonin levels in the pineal gland of con- i 60 n

o trols and AD patients containing short or t a l long genotype of MAOA-VNTR (sample size

e 40

M seen in table 1). * p<0.05. 20

0 Short Long Short Long Controls AD patients

70 MAOA polymorphism and AD

We found that the long genotype-bearing AD patients had higher MAOA gene expression compared to the long genotype-bearing controls (1.35 ± 0.11 and 1.09 ± 0.09, P = 0.02) (Fig. 1A). 5-HIAA: 5-HT ratio was also increased in the long genotype-bearing AD patients compared to the long genotype-bearing controls (0.35 ± 0.07 and 0.17 ± 0.02, P = 0.002) (Fig. 1B). However, there was no difference of MAOA gene expression between the short genotype-bearing AD patients and the short genotype-bearing controls (1.02 ± 0.13 and 1.18 ± 0.15, P = 0.6) (Fig. 1A), as well as 5-HIAA: 5-HT ratio (0.18 ± 0.04 and 0.21 ± 0.05, P = 0.9) (Fig. 1B). MAOA gene expression was positively correlated with 5-HIAA: 5-HT ratio both in controls (r = 0.42, P = 0.001) and AD patients (r = 0.41, P = 0.01). As the precursor of melatonin, serotonin, is catabolized by MAOA in the pineal gland, we examined the effect of the MAOA-VNTR genotype on melatonin levels. Melatonin levels in AD patients (32.0 ± 7.5 pg/mg tissue) were significantly lower than those in control subjects (85.6 ± 17.4 pg/mg tissue) (P = 0.04). There was no significant difference of melatonin levels between short and long genotypes in either controls (85.9 ± 33.6 and 89.8 ± 25.1 pg/mg tissue, respectively, P = 0.5) or in AD patients (45.9 ± 18.8 and 33.7 ± 10.7 pg/mg tissue, respectively, P = 0.3) (Fig. 1C). As a negative control for MAOA genotype-endophenotype association, we assessed the relationship between ApoE polymorphisms (chromosome 19) and MAOA expression. Six genotypes of ApoE were observed: 2/2, 3/2, 3/3, 4/2, 4/3 and 4/4. ApoE genotype did not affect MAOA gene expression, MAOA activity or melatonin production, neither in controls nor in AD patients (data not shown). AD patients had more frequent ApoE-ε4 allele-containing genotypes (frequency 67%) compared to controls (frequency 32%) (P < 0.0001), indicating an association between ApoE genotype and AD, as widely reported before 466.

Discussion The present study demonstrates that the promoter MAOA-VNTR polymorphism directly affects MAOA gene expression and activity in AD patients, and that it may thus modulates the susceptibility of AD patients to the MAOA dysfunction. Previously we found a significant increase of MAOA gene expression and activity (estimated by 5-HIAA: 5-HT ratio) in the pineal gland of AD patients 254, which has also been reported in other brain areas, including locus coeruleus, hypothalamus and frontal cortex 306, 308, 416, 417. In the present study by using MAOA gene expression and activity levels in the pineal gland as endophenotypes, we assessed the role of the MAOA-VNTR on AD-related MAOA disturbances. Studies in different cell lines and human skin fibroblasts showed that alleles with 3.5- or 4-repeats of the MAOA-VNTR are transcribed more efficiently than those with 3 repeats 421, 422, 463. Here in agreement with these in vitro experiments, we found that AD patients carrying long MAOA-

71 chapter 3

VNTR genotype (consisting of 3.5- or 4-repeats alleles) showed higher MAOA gene expression and activity in the pineal gland than the short-genotyped (i.e. 3-repeats allele) AD patients. Moreover, the AD-related increase of MAOA gene expression and activity showed up only among long genotype-bearing subjects. Our data thus suggest that the MAOA-VNTR polymorphism has a profound effect on the expression of MAOA in the AD brain, and that the long genotype of MAOA-VNTR may predispose AD patients to disturbances of MAOA. However, such functional effects of MAOA-VNTR were not observed in control human pineal gland in the present study or in control human profontal cortext in a previous report 470. The discrepancy of MAOA-VNTR functional effects between cell lines, AD brain tissue and control brain tissue is actually very interesting. Cell culture is known to impose a state of oxidative stress on cells, meaning an imbalance between the formation and spread of reactive oxygen species (ROS) and its antioxidant defenses (see review 471). Similarly, oxidative stress has been widely reported in Alzheimer’s disease, which may play an important role in the pathology of AD (see review 472, 473). We propose that increased reactive oxygen species may stimulate transcription factors to act through the long repeats MAOA-VNTR and subsequently alter MAOA expression in cell culture and AD brain tissues. No significant functional effect of MAOA-VNTR was observed in control brain tissues, which is likely due to a lack of activation of these transcription factors. Further studies are certainly warranted to test this hypothesis. Significantly, the MAOA promoter VNTR has been shown to be associated with other phenotypes of diseases, such as antidepressant therapeutic response in depression 474, and the severity of autism in autism patients 475. In general, this polymorphism can thus be seen as a genetic modifier of disease (endo)phenotype. Depression is among the most frequent neuropsychiatric comorbidities of Alzheim- er’s disease, affecting up to 50% of AD patients, and it is associated with more severe cognitive decline and greater costs of care 476, 477. MAOA is proposed to be involved in the pathogenesis of depressive disorders, since it is one of the main enzymes involved in the degradation of 5-HT, and MAO inhibitors have been used for decades to treat major depressive disorder. Interestingly, long MAOA-VNTR genotype is suggested to be associated with an increased risk of depressive disorder 474, 478, 479. As the long MAOA-VNTR seems to predispose AD patients to MAOA disturbances, whether the long MAOA-VNTR genotype may increase the susceptibility of AD patients for depressive disorders during the course of AD is a notable issue. In our study the number of AD cases where mood disorders is specifically mentioned is small, and the diagnosis of depression needs careful, prospective studies 416, 480. Further investigations are thus warranted to resolve this question, which might be of diagnostic and potentially also of therapeutic and prognostic value for depression in AD patients. Alterations of MAOA activity have shown to affect melatonin production in the hu-

72 MAOA polymorphism and AD man pineal gland 437, 438. Abundant studies described dramatically decreased melatonin levels in AD (see reviews 202, 203), which was proposed to be caused by two possible mechanisms. One is the dysfunction of the rhythmic control for the pineal melatonin synthesis from the biological clock of the brain, the suprachiasmatic nucleus (SCN). The other is the up-regulated MAOA in the pineal gland of AD patients 254, which might deplete 5-HT from melatonin synthesis. Notably, the long-genotyped AD patients who had higher MAOA activity did not show lower melatonin levels, compared to the short-genotyped AD patients. Our data may suggest that the up-regulation of MAOA might not be a major factor for the reduced melatonin production in AD. In other words, the dysfunction of the SCN-control may be more dominant for the decreased melatonin production in AD. A previous study reported an association between Alzheimer’s disease and MAOA dinucleotide repeats-(GT)n polymorphism (MAOA-GT) 309. This MAOA-GT polymorphism is located in the intron 2 of MAOA gene and in strong linkage disequilib- rium with the promoter MAOA-VNTR polymorphism 421. These observations raise the possibility of an association between MAOA-VNTR polymorphism with the predisposition to Alzheimer’s disease. Due to the limited number of subjects, such an association could not be concluded from the present study, and future investigations with suﬃcient number of cases are thus needed. In summary, the present study indicates that the long VNTR genotype in the MAOA promoter may increase the susceptibility to AD-related MAOA dysfunction.

73 Chapter 4

Pineal clock gene oscillation is disturbed in Alzheimer’s disease, due to functional disconnection from the “master clock”

Ying-Hui Wu, David F. Fischer, Andries Kalsbeek, Marie-Laure Garidou-Boof, Jan van der Vliet, Caroline van Heijningen, Rong-Yu Liu, Jiang-Ning Zhou and Dick F. Swaab

The FASEB Journal 20:1874-1876, 2006

Abstract The suprachiasmatic nucleus (SCN) is the “master clock” of the mammalian brain. It coordinates the peripheral clocks in the body, including the pineal clock that receives SCN input via a multisynaptic noradrenergic pathway. Rhythmic pineal melatonin production is disrupted in Alzheimer’s disease (AD). Here we show that the clock genes hBmal1, hCry1 and hPer1 were rhythmically expressed in the pineal of controls (Braak 0). Moreover, hPer1 and hβ1-adrenergic receptor (hβ1-ADR) mRNA were positively correlated and showed a similar daily pattern. In contrast, in both preclinical (Braak I-II) and clinical AD patients (Braak V-VI), the rhythmic expression of clock genes was lost, as well as the correlation between hPer1 and hβ1-ADR mRNA. Intrigu- ingly, hCry1 mRNA was increased in clinical AD. These changes are probably due to a disruption of the SCN control, as they were mirrored in the rat pineal deprived of SCN control. Indeed, a functional disruption of the SCN was observed from the earliest AD stages onwards, as shown by decreased vasopressin mRNA, a clock-controlled major output of the SCN. Thus, a functional disconnection between the SCN and the pineal from the earliest AD stage onwards could account for the pineal clock gene changes and underlie the circadian rhythm disturbances in AD.

Introduction In mammals, the master circadian clock is located within the suprachiasmatic nucleus (SCN) of the hypothalamus, which receives environmental light-dark information and orchestrates circadian rhythms at the organismal level 247,53. Molecular components of the circadian oscillator in mammals are a set of clock genes that involve intracellular transcriptional/translational feedback loops with negative (Per1-3, Cry1-2) and positive limbs (Bmal1 and Clock) 103, 247. Recent mammalian clock gene studies have revealed molecular clocks in many other brain regions and peripheral tissues, such as the pineal gland and liver, that are probably synchronized by the master clock in the SCN 122, 125, 481. Human clock genes are also expressed widely in the brain 137, although an analysis of rhythmic expression has so far only been reported in peripheral tissues such as oral mucosa, skin and peripheral blood mononuclear cells 138, 140, 142.

77 chapter 4

A major output of the SCN in mammals, including human, is the circadian rhythm of melatonin synthesis in the pineal gland, which is involved in the regulation of the circadian system 251, 426, 431. Sympathetic innervation of the mammalian pineal is activated at night via a multisynaptic pathway from the SCN to release noradrenalin

(NA), which acts on the β1 −adrenergic receptor (β1 −ADR) of the pinealocyte to trigger the cAMP signaling pathway 459 and thus leads to the activation of melatonin biosynthesis 252, 482. Interestingly, clock gene Per1 is rhythmically expressed in the rodent pineal under the same noradrenergic control from the SCN as the one that regulates melatonin synthesis 423, 424, 483, 484. Thus, the molecular clock of the rodent pineal seems to be synchronized by the central clock in the SCN. Although the role of the pineal molecular clock has not been fully elucidated, its involvement in the gated expression of N-acetyltransferase, the rhythm-generating enzyme of the melatonin biosynthesis has been proposed in rodents 485. Alzheimer’s disease (AD) patients often have disturbed sleep–wake behavior, which is one of their major clinical problems 189. A deterioration of neuronal function in the SCN has been reported in clinical AD patients 23, 169, 173. Moreover, melatonin levels are dramatically decreased, and the circadian rhythm in melatonin is lost in AD patients 254, 299, 301. These changes are considered to be responsible for the circadian disorders in AD. Recently we reported that the circadian noradrenergic regulation of the pineal by the SCN is affected in both preclinical AD subjects (cognitively intact subjects with the earliest AD neuropathological changes, i.e. Braak stages I-II) and clinical AD patients (Braak stage VI), and results in a loss of the circadian rhythm of melatonin synthesis 254. In the present study, we hypothesized that (1) day-night synchronization of the clock gene oscillation in the human pineal gland is affected early on in the AD process; (2) the SCN is affected from the earliest AD stages onwards, which is responsible for the changes in pineal clock genes. To test the first hypothesis, we studied the diurnal rhythm of clock gene expression in the pineal gland of unaffected controls (Braak stage 0), preclinical AD subjects (Braak stage I-II) and clinical AD patients (Braak stage V-VI) by quantitative PCR. To test the second hypothesis, we determined the expression of vasopressin mRNA in the SCN, a clock-controlled major output of the SCN in controls, preclinical and clinical AD subjects by means of quantitative in situ hybridization. In order to find experimental support for the second hypothesis, the effect of a loss of SCN control on the pineal clock gene oscillation was investigated by measuring pineal clock gene diurnal expression in the rat before and after disruption of the SCN-pineal functional connection, either by superior cervical ganglionectomy (SCG-X) or by lesioning the SCN (SCN-X).

78 pineal clock gene and AD

Materials and Methods Subjects Human post-mortem brain material was obtained from the Netherlands Brain Bank (NBB), which supplies postmortem specimens from clinically well-documented and neuropathologically confirmed cases. Autopsies were performed on NBB donors from whom written informed consent had been obtained for a brain autopsy and for the use of the material and of the clinical data for research purposes. Pineal glands from 68 subjects were studied: 24 controls (without any primary neurological or psychiatric disease and devoid of AD neuropathological changes in the brain, i.e. Braak stage 0), 22 preclinical AD subjects (cognitively intact with minor AD neuropathological changes, i.e. Braak stages I-II), and 22 late clinical AD patients [clinically met the NINCDS-ADRDA criteria 433, 436 and with extensive AD neuropathological changes, i.e. Braak stages V-VI]. To determine and compare diurnal variations of pineal clock genes expression in Braak stages 0, I-II and V-VI, subjects were grouped into four time bins: 1000-1600; 1600-2200; 2200-0400; 0400-1000 (i.e., day period, day/night transition period, night period and night/day transition period, respectively) based on their clock time of death, and matched according to their distribution in the time bins (Table 1). For the human SCN study, 9 controls in Braak stage 0, 9 preclinical AD subjects in Braak stages I-II, and 9 clinical AD patients with Braak stages V-VI were collected and matched for age, gender and day/night distribution (day: 10AM- 10PM) 173 according to their clock time of death (Table 2). The following variables were included in the statistical analysis: age, sex, clock time of death, postmortem delay, brain weight, pineal weight, pH of CSF [a measure for agonal state 439], fixation time and the cause of death (Table 1, 2). There was no difference in pH of CSF, pineal weight and fixation time in control, preclinical or clinical AD groups (all P>0.05). The Spearman correlation test showed that the differences of brain weight and postmortem delay among groups (P=0.03, P=0.04, respectively) did not seem to affect the results we obtained (all P>0.05).

Animals Experiments were conducted on adult male Wistar rats (200-300 g) (12h light / 12h dark schedule, food and water ad libitum). All experiments were performed in accordance with the guidelines on the care of experimental animals of the Animal Ex- perimentation Committee of the Netherlands Institute for Neuroscience (NIN) and approved by the European Communities Council Directive. Rats were divided into 5 groups: intact controls (n=48), sham SCN-lesions (n=6), sham SCG-lesions (n=6), SCN-lesions (SCN-X) (n=36) and SCG-lesions (SCG-X) (n=12). Control rats (6 per time point) were sacrificed at different time points of a day: ZT2, 6, 10, 14, 18 and 22 (ZT12 is the time when the light is switched off). Using quantitative PCR we revealed

79 chapter 4

Table 1 Clinical and pathological data for the controls, preclinical and clinical Alzheimer’s disease patients whose pineal was studied (Mean ± SEM) Braak stage Time period Nr. of Age pH of CSF PMD Brain weight Pineal cases (year) (hour) (g) weight (g) Braak stage 0 1 (0400-1000) 8 66± 3.1 6.60± 0.13 8:15± 0:30 1302.8± 59.7 0.18± 0.04 (Control) 2 (1000-1600) 6 65± 4.5 6.77± 0.15 7:05± 0:18 1335.8± 46.2 0.21± 0.14 N=24 3 (1600-2200) 6 66± 4.3 6.67± 0.18 5:20± 0:24 1294.7± 50.8 0.15± 0.24 4 (2200-0400) 4 66± 7.3 6.27± 0.24 8:38± 1:44 1253.3± 47.0 0.28± 0.51 Mean±SEM 65± 2.1 6.61± 0.08 8:03± 0:48 1300.7± 26.4 0.20± 0.02

Braak stages I-II 1 (0400-1000) 6 69± 3.8 6.56± 0.09 7:00± 0:39 1174.0± 55.9 0.23± 0.08 (Preclinical AD) 2 (1000-1600) 4 69± 2.4 6.80± 0.20 7:28± 1:25 1155.8± 39.8 0.21± 0.05 N=22 3 (1600-2200) 6 74± 1.9 6.46± 0.03 6:37± 0:33 1267.8± 59.0 0.25± 0.07 4 (2200-0400) 6 78± 3.5 6.72± 0.13 6:50± 1:21 1173.0± 68.3 0.21± 0.03 Mean±SEM 73± 1.7 6.63± 0.06 6:56± 0:28 1196.0± 29.4 0.23± 0.03

Braak stages V-VI 1 (0400-1000) 6 74± 3.8 6.52± 0.18 6:10± 1:12 1150.0± 59.5 0.20± 0.05 (Clinical AD) 2 (1000-1600) 5 79± 4.7 6.72± 0.15 4:10± 0:16 1067.2± 55.1 0.22± 0.04 N=22 3 (1600-2200) 6 72± 3.9 6.59± 0.13 4:04± 0:12 1116.0± 82.9 0.26± 0.07 4 (2200-0400) 5 77± 5.3 6.67± 0.09 4:44± 0:36 1055.2± 65.8 0.25± 0.09 Mean±SEM 75± 2.1 6.62± 0.06 4:44± 0:21 1098.0± 32.9 0.23± 0.03

Abbreviations: CSF cerebrospinal fluid; PMD: postmortem delay that rat Cry1 and Per1 were diurnally expressed in the pineal, with a peak at ZT18 and a trough at ZT8 (see Results). Bmal1 mRNA also showed a significant diurnal pattern, with a peak at ZT2. Bmal1 mRNA was significantly higher at ZT18 than at ZT8 (see Results). We therefore chose to sacrifice the SCN-X and SCG-X rats at ZT18 and ZT8. Sham lesioned rats did not differ from intact control rats as far as clock gene expression was concerned and so were added to the control group.

Suprachiasmatic Nucleus Lesion (SCN-X) SCN-X was performed in 36 rats as previously described 97, 98. In order to make a preselection of the eﬀective SCN lesions we measured water intake during a 3 wk period after a 2 wk recovery period. The rats that drank >33% of their 24 h water intake during 8 h of the 12 h light period (from ZT2 to ZT10) were considered arrhythmic 486. This resulted in a group of 14 animals, which were decapitated and processed for immunohistochemistry. The histological control of vasopressin (AVP) and vasoactive intestinal polypeptide (VIP) staining in the SCN area or SCN target areas revealed nine animals with complete SCN-X 97, 98.

80 pineal clock gene and AD Cardiac failure, epilepsy failure, Cardiac pneumonia Dehydration, Esophagus carcinoma Esophagus Mammacarcinoma vocal the right cord of Carcinoma septic shock and cardiac arteriosclerosis, Pneumonia, infarction myocardial failure, Heart carcinoma lung insufficiency, Respiratory edema, heart tamponage lung infarction: Myocardial insufficiency Respiratory pneumonia failure, Cardiac Myocardial infarction, lung cancer lung infarction, Myocardial secondary CARA severe to failure Cardiac decompensation cardiac Hypoglycaemia, Pneumonia Anemia dehydration Hypertension, origin unknown of Fever cachexia Dehydration, Pneumonia embolism pulmonary cordis, Decompensation dehydration Pneumonia, Lung cancer, pneumonia cancer, Lung insufficiency respiratory death, Sudden infarction Myocardial Cardiac arrhythmia Cardiac hypotension failure, Cardiac Cause of death Cause of 8:25 8:00 PMD 02:45 02:45 01:35 13:50 06:25 04:45 08:00 08:00 06:25 08:30 06:00 03:15 10:20 10:20 09:10 13:30 05:30 04:55 04:30 04:10 05:00 05:11 05:11 07:00 04:00 <09:00 <50:00 <48:00 7 7 6 na na na na na na na 7.2 6.4 6.7 6.6 6.8 6.5 6.5 6.3 6.6 6.4 6.9 6.7 6.6 6.4 7.1 6.5 6.7 CSF pH of pH of 925 961 2220 1250 1216 1499 1454 1344 1317 1135 1074 1385 1067 1355 1255 1440 1442 1168 1315 1219 1032 1055 1110 1030 1088 1315 1043 (gram) Brain weight weight Brain 0 52 32 27 28 28 34 60 32 28 31 41 34 63 56 29 24 30 31 44 32 34 11 26 32 Fix 122 137 (hour) 5:00 1:15 8:00 9:10 4:00 8:30 1:45 5:45 1:30 21:00 17:01 13:00 08:00 12:45 16:00 10:00 13:50 15:00 13:00 12:15 18:15 14:00 18:15 12:10 15:30 19:30 23:55 death Clock of time f f f f f f f f f f f f f f f f f m m m m m m m m m m m m Sex 61 63 63 65 70 73 74 78 85 69 71 72 75 77 78 78 86 87 66 69 76 78 80 81 82 83 85 Age (year) NBB 92-042 96-010 95-092 95-051 93-061 93-073 95-106 93-139 94-074 96-057 98-126 95-054 98-095 98-097 95-093 94-076 95-016 96-078 92-099 93-008 93-026 96-029 93-050 93-087 94-110 93-040 95-038 (Braak stage 0) stage (Braak I-II) stages (Braak (Braak stages V-VI) stages (Braak Preclinical AD Clinical AD Controls Clinical and pathological data for controls, preclinical and clinical Alzheimer’s disease patients whose SCN was studied (Mean ± SEM) SEM) ± (Mean studied was SCN whose patients disease Alzheimer’s clinical and preclinical controls, 2 for data Table pathological and Clinical Group obstructive disease pulmonary CARA: Chronic available; PMD: postmortem delay; not na: fluid; cerebrospinal time; CSF: fixation Fix:

81 chapter 4

- Amp. Amp. 1.91 1.95 1.93 1.91 1.94 1.95 1.91 1.95 1.97 1.99 1.97 1.99 1.98 1.97 efficiency TGGTCTGCCATTGGATGATCT TTGCCCCTTAGTCAGGAACCT CCCGCCAACTGCAGAATCT AGCAAAAATCGCCACCTGTT AGGAACATCAGCAAGCCACTC AAGCGACCCAAAGGTGGAT CTCCAACAGGACCTGCTGAAC AAACTCTTCATTGTTGTAGGTGTTGC GGTCTTTGAGATGCTGCATGG GAGAACCGTGGCTGTTGTTTC TGTCGGGCTAGATGGTGGAT CGTAGCCCAGCCAGTTGAAG TGTCTGCCTCATGTCACGAAC TGTCATTTGGCCCCATTATTG Reverse Primer Forward Primer Forward CCAGAGGCCCCTAACTCCTC TTGGCAAAATGTCATGAGCAC CAGTGCTCCTGTTCCTGCATC TCCGCTGCGTCTACATCCT CCCACAATCCTCGTCTGAATC AAGCTGGAAGATGGCCCTAAA CTGAAGAGAATCCACAAGGAATTGA TAGCCAGCCTGAGGTCTTTCA AATTTACCAATCCCACAAGGCA GTCCAGGGATGCAGCGTCT GGACAGGCTTTCCGTGGATT GGCCAACGTGGTGAAAGCT GCAAGCCCATGTGTGTTGAA CCAGCACAGTGTTCAGCAGGT NM_001178 NM_004898 NM_002616 NM_004075 NM_000684 J04617 U39317 NM_012780 NM_021856 XM_340822 NM_198750 NM_012701 NM_031001 Accession Code Accession BC063162 Gene Clock Per1 Cry1 receptor β1-adrenergic 1 alpha factor Elongation enzyme E2 ubiquitin-conjugating Clock Per1 Cry1 receptor β1-adrenergic 1 alpha factor Elongation enzyme E2 ubiquitin-conjugating Bmal1 Bmal1 RAT Human ficiency of each primer pair. primer each ficiency of GenBank accession code, sequence of PCR primer pairs for the HUMAN and RAT target genes and reference genes, and amplification ef 3 amplification and genes, reference and Table genes target RAT and the HUMAN for pairs PCR primer code, of sequence GenBank accession

82 pineal clock gene and AD

Superior cervical ganglia ectomy (SCG-X) Animals (n=12) were deprived of the sympathetic innervation of the pineal gland by bilateral removal of the SCG 487 10 days before they were sacriﬁced. The success of the surgery was checked by observing the ptosis of both eyelids, resulting in a group of 10 rats.

Pineal RNA isolation, reverse transcription and quantitative polymerase chain reaction (QPCR) Each frozen human or rat pineal gland was homogenized, and total RNA was isolated with TRIzol Reagent (Invitrogen, Breda, the Netherlands). cDNA was synthesized using superscript II reverse transcriptase (Invitrogen) in 5 min at 30 ˚C, 5 min at 37 ˚C, and 90 min at 42 ˚C. We used elongation factor 1 alpha (EF-1-alpha) and E2 ubiquitin conjugating enzyme (Ube2d2) to normalize the target gene expression data. The primers were designed with Primer Express software (Applied Biosystems, Foster City, Cali- fornia, USA). The efficiency of each primer pair was calculated using cDNA dilution curves and linear regression. Details of the primers, the GenBank accession numbers and the efficiency of each primer pair are given in Table 3. QPCR was performed using the SYBR Green PCR kit (Applied Biosystems) and an Applied Biosystems Model ABI 5700 Prism Sequence Detection System. An RT-polymerase chain reaction (RT-PCR) volume of 20 µl was used. All samples were run in duplicate using an annealing temperature of 60 ˚C. The amount of every target gene is calculated by raising the primer efficiency of the gene to the power of minus cycle threshold (– CT), normalizing this, and dividing it by the average of the two normalized housekeeping gene expression levels 254.

In situ hybridization and quantitative analysis of AVP mRNA in the SCN In situ hybridization was performed on every ﬁftieth (6 µm) section of the human postmortem SCN as extensively described before 173. We used an IBAS-KAT image analysis system for the quantitative analysis of the in situ signal of the AVP mRNA in the SCN as described 173.

Statistical analysis Differences among the groups were statistically evaluated by the Kruskal-Wallis multiple comparison test. Differences between groups were tested by the Mann-Whit- ney test. Diurnal rhythmicity of clock gene expression was determined first by the Kruskal-Wallis test over the four time bins. If a significant difference was identified among the four time bins, each combination of two time bins was then compared by the Mann-Whitney test. Correlations were analyzed by the Spearman correlation test. Statistical significance was considered at the P<0.05 level (two-tailed). Data are expressed as mean ± standard error of the mean (SEM).

83 chapter 4

Results Diurnal rhythmic expression of hBmal1, hCry1, hPer1 is lost in the pineal gland of both preclinical (Braak stages I-II) and clinical AD subjects (Braak stages V-VI) We found a significant diurnal rhythmic profile of clock genes hBmal1, hCry1 and hPer1 in the pineal gland of control subjects, as well as hβ1-adrenergic receptor (hβ1- ADR) mRNA (P=0.01; P=0.04; P=0.03; P=0.006, respectively)(Fig. 1 a, b, c, e). There were no significant daily variations in hClock gene expression (Fig. 1 d, p=0.76). Gene expression of hBmal1 showed a trough at the 1000-1600 time bin (0.86 ± 0.07) compared to the other three time bins (all P<0.02). hCry1 had a nocturnal peak at 2200-0400 (1.31 ± 0.08) which represents a 75% increase compared to the 1000-1600 time bin (0.75 ± 0.12) (p=0.01). Two significant daily peaks in hPer1 expression were detected at 0400-1000 (1.24 ± 0.16) and at 1600-2200 (1.22 ± 0.19) (i.e. during the light-dark transition periods), which represents an increases of about 85% compared to the 1000-1600 time bin (0.67 ± 0.11, both P=0.01) (Fig. 1 c). Diurnal variations of hβ1-ADR expression were similar to the hPer1 diurnal expression pattern: also with two peaks in the light-dark transition periods 0400-1000 (1.14 ± 0.13) and 1600-2200 (1.09 ± 0.20), compared to the 1000-1600 (0.70 ± 0.04) and the 2200-0400 time bin (0.44 ± 0.23) (all P<0.05, respectively) (Fig. 1 e). Moreover, hβ1-ADR and hPer1 mRNA were positively correlated (r=0.50, P=0.02, n=24). Intriguingly, diurnal rhythmic expression of hBmal1, hCry1, hPer1, and hβ1-ADR mRNA was lost, in both preclinical AD (Braak stages I-II) (P=0.20, P=0.45, P=0.82, P=0.35; respectively) and clinical AD (Braak stages V-VI) (P=0.30, P=0.20, P=0.11, P=0.25; respectively) (Fig. 1). No significant correlation was found between hβ1-ADR and hPer1 gene expression, either in Braak stages I-II (r=-.09, P=0.68, n=22) or Braak stages V-VI (r=0.25, P=0.26, n=22). Remarkably, we found that the daytime (0400-1000 and 1000-1600) levels of hCry1 mRNA were significantly increased in Braak stages V-VI (1.53 ± 0.17) compared to those in Braak stage 0 (0.94 ± 0.09) (P=0.004) and in Braak stages I-II (0.99 ± 0.14) (P=0.014), while no such change was observed between Braak stages 0 and I-II (P=0.9) (Fig. 1 b). There was no significant difference of hβ1-ADR, hPer1, hBmal1, hClock

Fig. 1 Daily gene expression pattern of hBmal1, hCry1, hPer1, hClock, hβ1-adrenergic receptor (hβ1-ADR) in the human pineal gland of aged controls (Braak stage 0), preclinical AD (Braak stages I-II) and clinical AD (Braak stages V-VI). The significant daily variation of hBmal1, hCry1, hPer1, hβ1-ADR mRNA only appears in Braak stage 0 (aged controls). * Denotes among Braak stage 0 the values significantly higher than 1000-1600 (a, b, c), or significantly higher than 1000-1600 and 2200-0400 (e). # Denotes hCry1 mRNA levels are higher in Braak stages V-VI than Braak stage 0 and I-II (b). ## indicates hCry1 mRNA levels are higher in Braak stages V-VI and Braak stages I-II than Braak stage 0 (b).

84 pineal clock gene and AD

a 2.2 b 2.2 # 2.0 2.0

1.8 1.8

1.6 1.6 1

l ## 1 a

1.4 y 1.4 r m C B * * 1.2 * * 1.2 h h 1.0 1.0

0.8 0.8

0.6 0.6

0.4 0.4

0.2 0.2

0 0 4-10 10-16 16-22 22-04 4-10 10-16 16-22 22-04

2.0 2.0 d c 1.8 1.8

1.6 1.6 k 1 c r 1.4 1.4 o l e * * P 1.2 C 1.2 h h

1.0 1.0

0.8 0.8

0.6 0.6

0.4 0.4

0.2 0.2

0 0 4-10 10-16 16-22 22-04 4-10 10-16 16-22 22-04 e 2.0

1.8 Braak 0 1.6 Braak I-II

R 1.4 Braak V-VI D A - 1.2 * * 1-ADR 1 β � 1.0 h 0.8

0.6

0.4

0.2

0 4-10 10-16 16-22 22-04

85 chapter 4

b Day (ZT 8) a 1.6 Night (ZT 18) 1.6 * * 1.4 * * * 1.4 * 1.2 1.2 1.0 1.0 0.8

Cry1

Cry1 0.6 0.8 0.4 0.6 0.2 11 11 5 4 5 5 0.0 0 2 4 6 8 1012141618202224 CON SCN-X SCG-X c d 2.0 * 2.4 1.8 1.6 2.0 * * 1.4 1.2 1.6 1.0 1.2 0.8

Per1

Per1 * 0.6 0.8 0.4 0.2 0.4 11 11 5 4 5 5 0.0 e 0 2 4 6 8 10 12 14 16 18 20 22 24 CON SCN-X SCG-X 1.4 * f 1.4 * * 1.3 1.2 1.2 * * 1.0 1.1 * 1.0 0.8

1-ADR

0.9 1-ADR 0.6

�

� 0.8 0.4

0.7 0.2 11 111 5 4 5 5 0.6 0 2 4 6 8 10 12 14 16 18 20 22 24 0.0 CON SCN-X SCG-X g 1.4 h

1.3 * 1.2 * 1.2 * 1.0 1.1 * 0.8 1.0 0.6

0.9 Bmal1 Bmal1 0.4 0.8 0.2 0.7 11 11 5 4 5 5 0.0 0 2 4 6 8 10 12 14 16 18 20 22 24 CON SCN-X SCG-X ZT Fig 2 86 pineal clock gene and AD mRNA levels among Braak stages 0, I-II and V-VI, either in the average level of 24 hours (P=0.18; P=0.36; P=0.49, P=0.26, respectively), or in the average level of any time bins (all P>0.05).

Rat pineal gland deprived of SCN control showed similar clock gene alterations as the human pineal gland in AD As the pineal gland is synchronized to the environment through the SCN, we investigated the eﬀect of the loss of SCN control on the pineal clock gene oscillation. We compared pineal clock gene diurnal expression in the rat before and after disruption of the SCN-pineal functional connection, either by superior cervical ganglionectomy (SCG-X) or by lesioning the SCN (SCN-X). In control rat pineal gland, Cry1, Per1, Bmal1 and β1-ADR mRNA showed clear diurnal rhythms (P<0.001; P<0.001; P=0.01; P=0.02) (Fig. 2), but not Clock (P=0.55). Moreover, Per1 and β1-ADR mRNA were highly correlated during 24 hours (r=0.61, P<0.001, n=58), or at their trough and peak times (ZT 8 and ZT18) (r=0.76, P<0.001, n=22). Following experimental denervation, the day-night (ZT 8 and ZT 18) diﬀerences in Per1, Cry1, β1-ADR and Bmal1 gene expression completely disappeared in both SCN-X and SCG-X rats (Fig. 2 b, d, f, h). Moreover, the correlation between Per1 and β1-ADR mRNA on ZT8 and ZT18 was lost, in both SCN-X (r=0.58, P=0.10; n=9) and SCG-X rats (r=0.46, P=0.17, n=10). Strikingly, daytime (ZT8) Cry1 mRNA levels were higher in SCN-X (0.83 ± 0.06) (P=0.036) and SCG-X rats (1.00 ± 0.04) (P=0.003) compared to control rats (0.66 ± 0.04) (Fig. 2 b). Remarkably, these changes in denervated rat pineal mirrored the clock gene alterations in the AD pineal, suggesting that the synchronization of the pineal through output from the SCN is lost in both preclinical and clinical AD.

Fig. 2 Daily expression patterns of clock genes in control rat pineal glands (a, c, e, g), and effects of suprachiasmatic nucleus lesion (SCN-X) and superior cervical ganglionectomy (SCG-X) on the daily fluctuations of pineal clock genes (b, d, f, h). (a) Cry1; * P<0.05 vs. ZT 6, 8, 10. (c) Per1; * P<0.05 vs. ZT 2, 6, 8, 10. (e) β1-ADR; *P<0.05 vs. ZT 8, 10. (g) Bmal1; * P<0.05 vs. ZT 6, 8, 10. SCN-X, SCG-X and their controls (CON) were sampled at ZT 8 (day) and ZT 18 (night), the nadir and peak of the Cry1, Per1 and β1-ADR diurnal profile, respectively. (b) Cry1; * P<0.05 vs. control ZT 8. (d) Per1; * P<0.05 vs. control ZT 8. (f) β1- ADR; * P<0.05 vs. control ZT 8 and SCG-X ZT18. (h) Bmal1; * P<0.05 vs. control ZT 8. The numbers in the bars indicate the number of animals in each group.

87 chapter 4

Decreased AVP mRNA levels in the SCN of both preclinical and clinical AD subjects AVP is a major rhythmic neuropeptide output of the SCN clockwork. It regulates the rhythm of activity within the SCN and induces the rhythmicity in other brain regions 94-96, 488. A significant decrease of AVP mRNA in the SCN was observed already from the earliest preclinical AD stages (Braak stages I-II) onwards. The total masked area of silver grains, as an estimate of total amount of AVP mRNA in the SCN, was decreased by 45% in Braak stages I-II (4334 ± 1099 µm2) (n=9) (P=0.038) and strongly reduced, by 70%, in Braak stages V-VI (2381 ± 871 µm2) (n=9) (P=0.004), as compared to Braak stage 0 (7861 ± 1302 µm2) (n=9) (Fig. 3 a). There was no significant difference in the AVP mRNA amount between Braak stages I-II and Braak stages V-VI (P=0.085) (Fig. 3 a). The total number of AVP mRNA-expressed cell profiles in the SCN markedly decreased in Braak stages V-VI (10270 ± 2140), i.e. 20% of that in Braak stage 0 (53968 ± 14192) (P=0.001), and 30% of that in Braak stages I-II (34228 ± 6393) (P=0.003) (Fig. 3 b). No difference was found between Braak stages 0 and I-II as far as the number of AVP mRNA-expressed cell profiles was concerned (P=0.40) (Fig. 3 b). These data indicate that the activity of the SCN, but not the cell number, is reduced very early on in the AD pathogenesis.

Fig. 3 Total amount of AVP mRNA (estimated by total mask area of silver grains, µm2) and total number of AVP mRNA expressing cell profiles in the SCN in Braak stage 0 (aged controls), Braak stages I-II (preclinical AD) and Braak stages V-VI (clinical AD stage). (a) Total amount of AVP mRNA in the SCN. (b) Total number of AVP mRNA expressing cell profiles in the SCN. Note that in Braak stages I-II the total AVP mRNA levels are already decreased while the number of AVP mRNA expressing cell profiles is not significantly different. * indicates p<0.05; ** represents p<0.01.

88 pineal clock gene and AD

Discussion This is the first study to show a rhythmic expression of a number of clock genes hB( - mal1, hCry1 and hPer1) in the human pineal gland of controls, i.e. subjects without neurological disease or AD neuropathological changes. Furthermore, we found that the rhythmic expression of pineal clock genes was lost in both preclinical and clinical AD patient groups, which may be due to a decreased output of the master clock, the SCN. We observed, in the rat pineal, a 24 h rhythm in the expression of the clock genes Per1, Cry1, Bmal1, but not in the expression of Clock, which is in agreement with previous studies in the rat pineal 423, 424, 489. Moreover, the temporal phasing of Per1 and Cry1 mRNA daily rhythm in rat pineal is in a good accordance with previous rat studies with the same 12/12 light/dark conditions 423, 424. Bmal1 mRNA showed a peak at ZT2 in our study rather than at ZT5 as reported in a previous study, which may be attributed to the difference between the two studies in light/dark conditions (LD 12/12 instead of LD 8/16, respectively) 489. Our observation of a diurnal rhythmic expression of hPer1, hCry1, hBmal1 mRNA in the human pineal gland is in agreement with these animal studies. However, the amplitude of clock gene daily rhythm is generally lower in the human pineal gland than in the rat pineal gland. Similarly, sheep pineal gland also showed lower amplitude in the diurnal rhythm of Per1 mRNA compared to rodents 490. These data indicate that there are species differences in diurnal expression of pineal clock genes which deserve further investigation. In rodent, the pineal clock gene Per1 is under the noradrenergic control of the SCN through a β-adrenergic cAMP signaling pathway 108, 423, 424, 484. β1-ADR acts as a key factor in this cAMP signaling pathway and is directly linked to the multisynaptic noradrenergic pathway derived from the SCN 459, 491. The highly positive correlation between Per1 and β1-ADR mRNA observed in the rat pineal, that actually disappeared in the SCN or SCG lesioned rat pineal (both lacking the noradrenergic SCN control of the pineal gland), may thus reflect the SCN regulation of Per1. Intriguingly, in the pineal gland of human control subjects, hPer1 and hβ1-ADR mRNA were also positively correlated and showed a similar daily expression pattern. Moreover, functionally active cAMP-responsive elements (CREs) have been found in the promoter region of the hPer1 gene of human 135. It therefore seems likely that the clock gene hPer1 in the human pineal gland is also controlled by the SCN via the β-adrenergic cAMP-signaling pathway. Thus, by activatinghPer1 , the sympathetic input could potentially trigger the clockwork oscillation in the human pineal gland. Interestingly, the diurnal rhythmic expression of hPer1, hCry1, and hBmal1 was lost in both preclinical and clinical AD, which suggests that pineal clock gene oscillation is disrupted very early on in the AD process. Moreover, the positive correlation between hPer1 and hβ1-ADR mRNA, which probably indicates SCN control of hPer1,

89 chapter 4 is disturbed in both preclinical and clinical AD. The pineal gland itself does not suffer from AD-neuropathology 286, nor does it show alterations in calcium deposition or total protein content in AD, compared to aged controls 254, 305. Therefore, we propose that the changes in clock gene expression in the pineal gland in the AD process are due to a disrupted SCN control. This possibility is strongly supported by our animal experimental data. The rat pineal that was deprived of SCN control showed alterations of clock gene expression that were remarkably similar to those we observed in the AD pineal. This holds for the loss of rhythmic clock gene expression, the loss of correlation between Per1 and β1-ADR mRNA, and in particular for the increased Cry1 mRNA. Although the mechanism underlying the increased hCry1 in AD pineal is not yet clear, our studies suggest that it may be due to a complete lack of sympathetic input. It has been shown that CRY1 protein is able to regulate Per1 transcription by binding both to activators (CLOCK and BMAL1) and inhibitors (PER1, PER2) in the circadian feedback loop 109. We hypothesize that if the sympathetic regulation of Per1 expression from the SCN is lacking, CRY1 may be needed to act as the main regulator for the transcription-translation loop that leads to the increased Cry1 gene expression as observed in late AD and denervated rat pineal. Moreover, the recent identification of the orphan nuclear receptor REV-ERB-alpha as a negative regulator of Cry1 104, 111 suggests that this signaling pathway might also be involved. The pineal clock genes have been suggested to gate melatonin synthesis through the cAMP signaling cascade in the pineal gland and retina of rodents 485, 492. cAMP signaling can influence both transcription and post-transcriptional regulation (e.g. proteasomal degradation and inhibition/activation switch) of AA-NAT (the rate-limiting enzyme for melatonin synthesis) and thus the melatonin production 454, 493. No- tably, the gated regulation of melatonin synthesis has also been observed in humans. In healthy human subjects, evening administration of β-ADR agonist or antagonist strongly affected plasma melatonin concentrations 494, whereas daytime administration had no such effect 495, 496. In rodents, melatonin synthesis is controlled on both AA- NAT transcriptional and post-transcriptional levels, whereas in primates it is probably mainly controlled on the post-transcriptional level of AA-NAT regulation 253, 254. Thus, we hypothesize that the pineal clock genes gate the regulation of melatonin synthesis through a cAMP-dependent post-transcriptional pathway in humans. Further studies are needed to elucidate this mechanism. We and others previously reported typical cytoskeletal alterations and degenerative changes in the SCN, both on AVP protein and mRNA levels 23, 169, 173, 174. Remarkably, in the present study, we found that the AVP mRNA levels in the SCN decreased from the earliest AD neuropathological stages onwards. AVP is a major rhythmic neuropeptide output of the SCN clockwork and regulates the rhythm of activity within the SCN and induces the rhythmicity in other brain regions 94-96, 488. The AVP neuropeptide pattern

90 pineal clock gene and AD

SCN change :

AVP

SCG Pineal changes:

β1 ADR Clock genes Melatonin and

Circadian Disturbances

Fig. 4 Schematic overview of the functional connectivity between the SCN and the pineal and its changes from the earliest neuropathological stages of Alzheimer’s disease onwards. A decrease of vasopressin mRNA levels, indicating a functional disruption of the SCN, leads to a functional disconnectivity between the SCN and the pineal. This may cause a desynchronization of the pineal clock gene oscillation and melatonin production to the environmental cues and contributes to the circadian rhythm disturbances in AD. The dark arrow-lines indicate the pathway from the SCN to the pineal, including the retina-hypothalamic tract, which sends environmental light information directly to the SCN. The changes in the SCN and the pineal shown in the scheme concern mRNA level, except for melatonin. = a decrease of levels. = daily rhythm; = disrupted daily rhythm.

91 chapter 4 in the SCN is driven by AVP gene expression 92, which is strictly controlled by the molecular clock in the SCN, as its rhythm had disappeared and its mRNA levels were dramatically decreased in the SCN of Clock mutated mice 93. Therefore, our data suggest that the SCN has a diminished output and a disrupted clock function from the earliest AD stages onwards. Moreover, it supports the possibility that the SCN control of the pineal gland is disturbed very early on in the AD process. The mechanism(s) underlying such early functional changes of the SCN in AD certainly deserve further study. The finding, that human pineal clock gene changes in the AD process are mimicked in the rat pineal by SCN lesion or SCG-ectomy, strongly suggests that the functional connection between the SCN and the pineal gland is affected in both preclinical and clinical AD stages. Other functional disturbances of brain network connections could underlie the cognitive deficits in Alzheimer’s disease. Indeed, a loss of functional connectivity between prefrontal cortex and hippocampus has been reported recently in living Alzheimer patients 497, and a loss of synapses has been proposed to be one of the earliest changes in the disease process 498. Synchronization of the pineal clock gene oscillation to environmental cues has provided us with a rare opportunity to study functional connectivity in the post-mortem brain. In mammals, in addition to the pineal gland many peripheral tissues express oscillating clock genes 122, 499, which may be synchronized by the master clock in the SCN and contribute to many aspects of circadian physiology. The alterations of the master clock in AD that disrupts clock gene oscillation in the pineal gland could also affect clock gene oscillation in other peripheral tissues and in this way contribute to the circadian disorders in AD. Considerable variability was found between subjects within the groups. Ante- and postmortem confounding factors such as age, sex, agonal state, and clock time of death, that may contribute to such variations 500, 501, were excluded as well as possible in the present study by matching. Information on the exact influence of each of these factors on clock gene mRNA levels is, however, still very limited. In conclusion, the decreased activity of the SCN—already present at the moment of the occurrence of the very first tangles in the transentorhinal cortex (Braak stage I)—most probably affects pineal clock gene synchronization. The loss of functional connectivity between the master clock (the SCN) and peripheral clocks (e.g. the pineal gland) may underlie the circadian disturbances in the course of AD (see Fig. 4). Ap- parently, the circadian system is extraordinarily vulnerable to AD pathogenesis. We propose that the circulating melatonin levels and their daily rhythmicity may thus provide information about the very first AD stages that cannot be monitored in any other way at this moment.

92 Chapter 5

Distribution of MT1 melatonin receptor immunoreactivity in the human hypothalamus and pituitary gland: colocalization of MT1 with vasopressin, oxytocin, and corticotropin-releasing hormone

Ying-Hui Wu, Jiang-Ning Zhou, Rawien Balesar, Unga Unmehopa, Aimin Bao, Ralf Jockers, Joop Van Heerikhuize, Dick F. Swaab

Journal Comparative Neurology, 499:897-910, 2006.

Abstract Melatonin is implicated in numerous physiological processes, including circadian rhythms, stress and reproduction, many of which are mediated by the hypothalamus and pituitary. The physiological actions of melatonin are mainly mediated by melatonin receptors. We here describe the distribution of the melatonin receptor MT1 in the human hypothalamus and pituitary by immunocytochemistry. MT1 immunoreactivity showed a widespread pattern in the hypothalamus. In addition to the area of the suprachiasmatic nucleus (SCN), a number of novel sites, including the paraventricular nucleus (PVN), periventricular nucleus, supraoptic nucleus (SON), sexually dimorphic nucleus, the diagonal band of Broca, the nucleus basalis of Mey- nert, infundibular nucleus, ventromedial and dorsomedial nucleus, tuberomamillary nucleus, mamillary body and paraventricular thalamic nucleus, were observed to have neuronal MT1 receptor expression. No staining was observed in the nucleus tuberalis lateralis and bed nucleus of the stria terminalis. The MT1 receptor was colocalized with some vasopressin (AVP) neurons in the SCN, colocalized with some parvocellular and magnocellular AVP and oxytocine (OXT) neurons in the PVN and SON, and colocalized with some parvocellular corticotropin-releasing hormone (CRH) neurons in the PVN. In the pituitary, strong MT1 expression was observed in the pars tuberalis, while a weak staining was found in the posterior and anterior pituitary. These ﬁndings provide a neurobiological basis for the participation of melatonin in the regulation of various hypothalamic and pituitary functions. The colocalization of MT1 and CRH suggests that melatonin might directly modulate the hypothalamus- pituitary-adrenal axis in the PVN, which may have implications for stress conditions such as depression.

Introduction The pineal hormone melatonin is rhythmically produced under the control of the central biological clock of the brain, the suprachiasmatic nucleus (SCN), located in

95 chapter 5 the hypothalamus, via a multi-synaptic sympathetic pathway 251. Melatonin acts as a hormonal message of the photoperiod to regulate circadian and seasonal rhythms 57, 502. Moreover, melatonin is involved in numerous physiological responses, including sleep regulation, reproduction, stress, mood, body temperature and cognition 57, 318, 331-334. Many of these physiological processes are mediated by the hypothalamus and pituitary gland. In mammals, the physiological functions of melatonin are largely mediated by two G protein-coupled melatonin receptors, i.e. MT1 and MT2 that are able to couple to distinct signal transduction cascades, whose activation may lead to different cellular responses (see review 201, 503). MT1 is considered to be the most widely expressed melatonin receptor subtype in mammalian brain 64. Numerous animal studies have shown melatonin binding or the MT1 receptor mRNA in discrete hypothalamic regions and pituitary, in particular in the SCN and pars tuberalis (PT) of the pituitary in mammals. The MT1 receptor in the latter two areas is thought to mediate the response to circadian and reproductive function of melatonin (see review 201). However, only very few studies are available on melatonin receptor expression in the human hypothalamus and pituitary. In vitro autoradiography and in situ hybridization showed a low expression of melatonin binding and MT1 mRNA limited in the SCN region and PT in the hypothalamus of human adult and fetus 200, 342, 504. By reverse transcription PCR, MT1 mRNA was detected in some more hypothalamic areas, including the ventromedial nucleus, arcuate nucleus and mamillary nucleus in human fetus 347. However, so far, the exact MT1 receptor expression pattern in the adult human hypothalamus and pituitary at the protein level is entirely lacking. In the present study we, therefore, systematically investigated the MT1 receptor expression in the human adult hypothalamus and pituitary at the protein level by immunocytochemistry. Melatonin was shown to regulate vasopressin (AVP) and oxytocin (OXT) release in the hypothalamus in vitro 505, in vivo in rats (reviewed in 506), and in healthy human subjects 506-508. In order to explore whether melatonin may exert such an effect directly via the presence of the MT1 receptor in the hypothalamic AVP and OXT neurons, colocalization studies of the MT1 receptor with AVP and with OXT in the hypothalamus were performed. Furthermore, melatonin was reported in the rat to modulate the activity of the hypothalamus–pituitary–adrenal (HPA) axis and to reduce corticotropin-releasing hormone (CRH) release in chronic stress and glucocorticoid-induced deterioration 509, 510. Moreover, such an inhibitory effect of melatonin on the HPA axis in human is supported by in a recent study in blind humans, where a single oral dose of melatonin has shown to suppress adrenocorticotrophin (ACTH) and cortisol secretion 264. To investigate whether melatonin may act directly via the MT1 receptor on the CRH neurons of the hypothalamic paraventricular nucleus (PVN), colocalization of the MT1 receptor was studied in the hypothalamus by means of CRH.

96 MT1 in the hypothalamus and pituitary

Subjects and Methods Subjects In the present study we used 9 hypothalami and 8 pituitaries obtained via the rapid autopsy procedure of the Netherlands Brain Bank (NBB) at the Netherlands Institute for Neuroscience (NIN). Permission was obtained for a brain autopsy and for the use of human brain material and clinical information for research purposes. Clinico- pathological data are presented in table 1.

Histology Hypothalamic and pituitary samples were ﬁxed in formalin and embedded in par- aﬃn. Serial 6-µm coronal sections were made from the pituitary gland and from the hypothalamus from the level of the lamina terminalis to the mamillary bodies. Depending on availability, either the left or right hemi-hypothalamus was used. For anatomical orientation every 100th hypothalamic section was collected and mounted on superfrost plus slides (Menzel Gläser, Germany) and subsequently dried for at least 2 days at 37 ºC, followed by Nissl staining (0.5 % thionine in distilled water).

Specificity test of the MT1 receptor antibody A polyclonal antibody specific for the human MT1 receptor was raised in a rabbit, directed against a peptide corresponding to a sequence found in the C-terminal region of this receptor (Ac-YKWKPSPLMTNNNVVKVDSV-COO2, peptide 536) and affinity-purified by affinity chromatography on a glutaraldehyde-activated Sepharose column coupled to peptide 536 via the amide group of its lysine residue 378. This antibody detected the MT1 receptor as a protein with an apparent molecular weight of 60KD in immunoblots after separation by SDS-PAGE. It also specifically precipitated the 2- [125I]iodomelatonin–labeled receptor from MT1-transfected HEK 293 cells 378 and has been used before in human postmortem brain material 511, 512. In the present study, specificity of the antibody (batch number 9/03/00) in our system was further confirmed by (1) staining following solid phase preadsorption of the anti-MT1 antibody with homologous peptide of human MT1 receptor; (2) Western blotting, and (3) staining of sections following the same immunocytochemistry protocol as below but omitting of the primary anti-MT1 antibody. In addition, we assessed the staining pattern in wild-type and target-MT1-disrupted mouse brain (a generous gift from Dr. Weaver and Dr. Rep- pert, for details see 64), following the same staining protocol as used for human material.

1. Spot blot test and solid phase preadsorption A spot blot test and solid phase preadsorption were performed as previously described 513. The synthetic human MT1 peptide was dissolved (500 ng/μl) in IEF medium [10 % glycerol, 10 % dimethylformamide, 2.5 % Nonidet (Sigma)] and spotted on three

97 chapter 5 separate strips of gelatin (0.2 %) coated nitrocellulose (0.1 μm poresize, BA45, Sch- leicher & Schuell, Germany), each containing 10 μg of peptide. Fixation of the peptide to the nitrocellulose was performed overnight with 4 % paraformaldehyde in a press block, followed by rinsing in water (3x10 min), TBS (0.05M Tris, 0.15M NaCL, pH 7.6; 3x10 min) and SUMI [0.05 M Tris, 0.15 M NaCL, 0.25 % gelatin (Merck, Amsterdam, The Netherlands) (wt/vol), 0.5 % Triton (Sigma) (vol/vol), pH 7.6; 3x10 min]. Subse- quently, the anti-MT1 antibody was adsorbed at 1:200 in three cycles, each time by incubation of a strip of spotted nitrocellulose overnight at 4 ºC. As a negative control, we took along a non-relevant peptide [peptides 1-20 of the human synthetic androgen receptor 514] in the adsorption procedure. After the ﬁnal adsorption cycle, binding of the antibody to the peptide was visualized by staining the spotted strips using the avidin-biotin-complex (ABC) method, as described below. Finally, hypothalamic and pituitary sections were immunostained using the adsorbed anti-MT1 antibody solution according to the procedure described below.

2. Western blotting Western blotting was performed on extracts from diﬀerent areas of a fresh frozen hypothalamus from a control subject (51 year-old male, NBB 03-002), including PVN, SON and TM. Protein was run on 7.5 % SDS-PAGE gel and electroblotted onto nitrocellulose membrane (BA45, Schleicher & Schuell, Germany). After blocking with phosphate buﬀer saline containing 0.1 % Tween (PBS-tween), one blot was incubated with the MT1 antibody (1:400 in PBS-tween) for 1 hour at room temperature (RT), and then overnight at 4 ºC. Another blot was incubated with omission of the primary anti-MT1 antibody as negative control. The blots were then rinsed and incubated in the second antibody (porcine-anti-rabbit HRP, Dako, 1:1000 in PBS-tween) for 1h at RT. The bands were detected with chemiluminescence, using Western Lightning Chemiluminescence reagents (PerkinElmer, Boston, USA).

MT1 immunocytochemical staining To localize the MT1 receptor, MT1 immunocytochemistry was performed with the anti-MT1 antibody in the pituitary and on every 100th hypothalamic section taken along the rostrocaudal axis. Deparaffinized and rehydrated sections were pretreated with 3 μg/ml proteinase-K (Invitrogen, Breda, Netherlands) in Proteinase-K buffer (2 mM CaCL2, 0.01 M Tris/ HCl pH 7.5) at 37 ºC for 15 min, and rinsed in glycine buffer (2 g glycine in 1000 ml PBS) for 30 sec to stop the proteinase-K reaction. Sections were then preincubated with TBS-milk [3 % nonfat dry milk powder in TBS (pH 7.6), ELK, Campina] for 1h at room temperature (RT), followed by one hour incubation (RT) with anti-MT1 antibody 1:200 in SUMI-milk [0.25 % gelatin (wt/vol), 0.5 % Triton X-100 (vol/vol) in

98 MT1 in the hypothalamus and pituitary

TBS-milk, pH 7.6] and overnight incubation at 4 °C in a moist chamber. The next day slides were washed once with TBS-milk and twice with TBS, followed by incubation with biotinylated goat-anti-rabbit IgG (Vector Laboratories Inc., Burlingame, CA) at 1:400 in SUMI (1 h, RT). After rinsing in TBS the sections were incubated with ABC (Vector Laboratories, Inc. Burlingame, CA) at 1:800 in SUMI (1 h, RT). Following washes, a tyramid signal ampliﬁcation step 515 was applied by incubating sections with biotinylated tyramide 1:500 in TBS containing 0.01 % H2O2 (Merke) at RT for 15 min. The ABC procedure was then repeated (see above). After rinsing in TBS, visualization of the signal was accomplished by incubation in 3.3’-diaminobenzidine (DAB) (Sigma) 0.5 mg/ml TBS, containing 0.23 % (wt/vol) nickelammoniumsulfate (Merck) and 0.01 % (vol/vol) H2O2 (Merck) for approximately 20 min (RT). The reaction was stopped in distilled water, and slides were dehydrated in an ascending series of alcohol and xylene and coverslipped in Entellan (Merck).

Immunofluorescence double staining and confocal laser scanning microscopy To further characterize the MT1 immunoreactive cells in the SCN, PVN and SON, we performed immunofluorescence double staining of MT1-AVP, MT1-OXT and MT1-CRH, respectively. Hypothalamic sections were pretreated as described above in the MT1 immunocytochemical staining. For the combination of MT1 and AVP staining: the primary antibodies were (1) rabbit polyclonal anti-MT1 antibody at the optimal dilution of 1:100; (2) mouse monoclonal anti-AVP (VP III-D-7) at 1:50 dilution, which recognizes Phe in position 3 as the most important determinant in the AVP ring (a generous gift from Dr. A. Silver- man to Dr. F.W. Van Leeuwen, Netherlands Institute for Brain Research, Amsterdam, The Netherlands; for details see 516). Only processed AVP can be recognized by this antibody 517. For the combination of MT1 and OXT staining, the primary antibodies were (1) anti-MT1 antibody (1:100); (2) mouse monoclonal anti-OXT (A-I-28) at 1:100 dilution, which has three different antigenic determinants on the OXT molecule: the Ile in position 3, Pro in 7 and Leu in 8 (a generous gift from Dr. A. Silverman to Dr. F.W. Van Leeuwen; for details see 516, 518). For the combination of MT1 and CRH staining: the primary antibodies were (1) anti- MT1 antibody (1:100); (2) rat monoclonal anti-CRH antibody ‘PFU 83’ (IgG2a sub- class) at dilution of 1:50,000, which aims at the C-terminal part (amino acids 38–39) of rat/human CRH protein (kindly donated by Professor F.J.H Tilders, Vrije Universiteit, Amsterdam, The Netherlands). A radioimmunoassy (RIA) has proved it to be highly specific, with no significant binding of pituitary hormones and hypothalamic peptides other than CRH 519. Moreover, the specificity of this antibody has been confirmed previously in our laboratory by a spot blot test and solid phase preadsorption 520, 521.

99 chapter 5

Antibody incubation took place in 3 % milk-SUMI for 1 h at RT, followed by overnight incubation at 4 °C. After rinses in TBS-milk and TBS, the secondary antibodies were applied. Within the MT1-AVP combination, MT1 was visualized in green with streptavidin-coupled Cy2 after tyramide signal amplification, using biotinylated anti- rabbit antibodies and ABC. AVP was detected in red with anti-mouse Cy3 conjugated antibodies. On sections for MT1-OXT and MT1-CRH colocalization, the detection strategy was reversed: MT1 was visualized in red with anti-rabbit Cy3 antibodies, while OXT and CRH were detected in green with streptavidin-coupled Cy2 after tyramide signal amplification. Fluorochrome-conjugated antibody incubations were performed in SUMI for 5 h at RT, followed by an overnight incubation at 4 °C (all fluorochrome-conjugated antibodies were from Jackson Laboratories, BarHarbor, ME). Finally, sections were rinsed in TBS, covered with Vectashield H-1000 (Vector Laboratories) and analyzed on a Zeiss 510 confocal laser scanning microscope (CLSM) equipped with lasers emitting at 488 and 543 nm to excite FITC/Cy2 and TRITC/Cy3, respectively (Carl Zeiss B.V., Sliedrecht, The Netherlands). Contrast and brightness of light microscopic images were improved by using Corel PHOTO-PAINT 9.0. Images acquired with the confocal microscope were minimally altered in Adobe Photoshop 7.0 in dynamic range.

Results Specificity test In the spot blot test, anti-MT1 antibody recognized its blocking synthetic peptide on nitrocellulose paper by diminishing staining after each time of adsorption until a negative staining occurred on the strips. As a negative control, the antibody did not bind

Fig 1 MT1 receptor immunoreactivity in the human supraoptic nucleus (SON) neurons before (A) and after (B) solid phase preadsorption of the anti-MT1 antibody with the homologous peptide. Note that the cytoplasmic staining of the MT1 receptor is completely eliminated after preadsorption. Scale bar represents 50µm.

100 MT1 in the hypothalamus and pituitary to spots containing the androgen receptor 1-20 peptide, and the nitrocellulose strips showed identical negative staining irrespective of the times of adsorption. Staining with preadsorption of anti-MT1 with the synthetic human MT1 peptide eliminated any positive staining in the hypothalamic sections (Fig. 1) and pituitary sections. Staining western blots of the human hypothalamus with anti-MT1 receptor revealed bands of the expected molecular weight 60 KD for the MT1 receptor protein in different areas of the human hypothalamus, i.e. SON, PVN and TM (Fig. 2). Staining western blots with omission of primary anti-MT1 antibody did not show any bands (Fig. 2). Moreover, omission of the primary anti-MT1 antibody showed no staining in the hypothalamic and pituitary gland sections. Taken together, our data in addition to the available literature 378, supported the specificity of the anti-MT1 antibody in the human brain. We characterize the specificity of the anti-human MT1 antibody further by assessing the staining pattern in the target-MT1-disrupted mouse brain provided by Dr. Weaver, which, to our knowledge, is the only MT1-knockout mouse up to now 64. We observed a clear staining in both wide-type and MT1-knockout mouse brain (data not shown). This result, however, does not add any value, neither positive nor negative, to the specificity of our antibody, as this type of MT1-knock mice is not a good model for testing our antibody. This target-MT1-disrupted mouse only deleted exon1 from the MT1 gene, whereas exon 2 is still present 64, which may well be capable of producing non-functional truncated proteins, and be stained by our antibody that is raised against the C-terminal encoded by exon 2 378.

Fig 2 Western blots of diﬀerent areas of human hypothalamic tissue samples, i.e. supraoptic nucleus (SON), tuberomamillary nucleus (TM) and paraventricular nucleus (PVN), by using anti-MT1 antibody results in bands of the molecular weight 60 KD. Staining western blots with omission of anti-MT1 receptor antibody (right panel) did not show any bands.

101 chapter 5 - Found dead in bed Found based a pulmo on hypertension, Severe pulmonary the of rupture secondary traumatic to Haematothorax accident) car a (after ventricle left infarction Heart failure. organ Sepsis multi with Bronchopneumonia/bronchitis disorder Bipolar hematoma, Subdural Cause of death death Cause of Respiratory insufficiency due to massive hemorrhagic hemorrhagic massive to due insufficiency Respiratory virus immunodeficiency Human bronchopneumonia, Severe prostate cancer with metastasis. with cancer Severe prostate carcinoma a bladder of Bone metastases possibly aorta. the ascending A dissection of Type coma. stopped, was Medication disease Alzheimer’s Bronchopneumonia, Dysrhythmia. disease Parkinson embolism. pulmonary failure, Heart failure emphysema/heart pulmonary Terminal nary capillary hemangiomatosis 5:55 7:45 8:15 30:00 16:46 PMD 10:00 05:15 02:55 05:30 05:45 07:30 <8:15 <17:00 <41:00 <46:25 <10:37 <.16:30 31 45 45 72 98 45 34 31 31 30 31 31 31 31 31 130 1390 (days) Fixation

July July May April Month Month August August August August August of death of February February December December November November November November NA NA 4:00 8:45 0:30 night 18:35 06:23 11:20 08:00 07:30 17:55 18:00 16:00 05:15 13:30 14:00 of death of Clock time f f f f f f f f m m m m m m m m m m Gender

25 32 32 32 33 39 40 69 70 53 61 68 69 73 74 76 78 (y) Age Age Area studied Area tuberalis pars Hypothal., Hypothal. tuberalis pars Hypothal., tuberalis pars Hypothal., tuberalis pars Hypothal., tuberalis pars Hypothal., tuberalis pars Hypothal., tuberalis pars Hypothal., tuberalis pars Hypothal., Pituitary Pituitary pituitary tuberalis, Pars Pituitary Pituitary Pituitary Pituitary Pituitary Note: NBB number: Netherlands Brain Bank number; PMD postmortem delay. Fixation: ﬁxation time. f: female, m: male. NA: not available. not NA: m: male. f: female, time. ﬁxation Fixation: Bank number; PMD postmortem delay. Brain number: NBB Netherlands Note: Table 1 Table subjects of data Clinico-pathological NBB number 01-072 92-037 96-239 02-006 98-031 99-071 99-125 97-157 00-090 95-102 96-081 02-014 95-103 95-101 96-082 96-075 96-084

102 MT1 in the hypothalamus and pituitary

MT1 receptor expression in the human hypothalamus MT1 immunoreactivity (ir) was observed as non-nuclear granular staining localized in the cytoplasm of neurons, and it was not found in glia cells. MT1-ir neurons were localized throughout the human hypothalamus, but in diﬀerent amounts in diﬀer- ent regions and only in part of the neurons. The results showing the distribution of MT1-ir are summarized in Table 2 and Figures 3-6. Strong cytoplasmic MT1-ir was observed in the PVN (Fig. 4A), periventricular nucleus (PeVN) and SON (Fig. 4B). Both parvocellular and magnocellular neurons in the PVN and SON were positive for MT1 staining. Neurons stained with medium intensity were found in the SCN (Fig. 4C), sexually dimorphic nucleus of the preoptic area (SDN-POA) (Fig. 4D), diagonal band of Broca (hDBB/DBB) (Fig. 4E), nucleus basalis of Meynert (NBM) (Fig. 4F), infundibular nucleus (IF)(arcuate nucleus) (Fig. 5A), TM (Fig. 4G) and nucleus of the mamillary body (MB) (Fig. 5B). Weak MT1-ir staining was observed in the lateral hypothalamic area (LHA), ventromedial nucleus (VMN) (Fig. 5C) and dorsomedial nucleus (DMN) (Fig. 5D). No staining was observed in the nucleus tuberalis lateralis (NTL) (Fig. 4H) and bed nucleus of the stria terminalis (BST).

Table 2 MT1 receptor immunoreactivity intensity in diﬀerent areas of the hypothalamus NBB SCN PVN PeVN SON SDN DBB NBM TM INF ME VMN NTL LHA MB nr. & DMN 01-072 +-++ +++ +++ +++ +-++ +++ ++ ++ ++ ++ + - + + 92-037 ++ +++ ++ +++ + ++ ++ ++ +-++ +-++ + - - ++ 96-239 + ++ ++ ++ + ++ + ++ ± ± - - - ± 02-006 +++ +++ +++ ++ ++ +++ ++ ++-+++ ++ + + - + ++ 98-031 ++ +++ +++ +++ +++ +++ ++ +++ ++ ++ - - - ++ 99-071 ++-+++ +++ +++ +++ ++ +++ ++ ++ ++ ++ ++ - ++ ++ 99-125 +-++ ++ ++ ++ +-++ ++ - + ++ + - - - + 97-157 ++ +++ ++ +++ ++ +++ ++ ++-+++ ++ ± ± - - + 00-090 ++ ++ ++ ++ - + - + ± ± - - - ±

Scales: - no staining; ± very weak staining; + weak staining; ++ median intensive staining; +++ Strong staining. SCN: suprachiasmatic nucleus, PVN: paraventricular nucleus, PeVN: periventricular nucleus, SON: supraoptic nucleus, SDN-POA: sexually dimorphic nucleus of the preoptic area or intermediate nucleus of the anterior hypothalamus-1; DBB: diagonal band of Broca or vertical limb of the diagonal band of Broca (Ch2), hDBB: horizontal limb of the diagonal band of Broca (Ch3), NBM: nucleus basalis of Meynert (Ch4), TM: tuberomamillary nucleus, INF: infundibular or arcuate nucleus, ME: median eminence, VMN: ventromedial hypothalamic nucleus, DMN: dorsomedial hypothalamic nucleus, NTL: nucleus tuberalis lateralis, LHA: lateral hypothalamic area, MB: mamillary body.

103 chapter 5

A. B.

FO PNN

FO PeVN

TM DMV DBB PVN NBM VMN SDN OT

hDBB 3V PeVN SON NTL

INF 3V SCN OX

PT/PV C.

OT 3V FO

Fig 3 Schematic representation of the distribution of the MT1 receptor from rostral to caudal through the human hypothalamus. The density of the dots indicates the median values of the MT1 staining intensity. For abbreviations in all ﬁgures, see the list of abbreviation.

104 MT1 in the hypothalamus and pituitary

Fig 4 The MT1 receptor immunoreactivity in the A: paraventricular nucleus (PVN); B: supraoptic nucleus (SON); C: suprachiasmatic nucleus (SCN); D: sexually dimorphic nucleus of the preoptic area (SDN-POA); E horizontal diagonal band of Broca (hDBB), F: nucleus basalis of Meynert (NBM). G: tuberomamillary nucleus (TM) H: nucleus tuberalis lateralis (NTL). Note the positive MT1 staining in the cytoplasm of neurons in all the above nuclei except the NTL. Scale bar indicates 50 µm. 105 chapter 5

Fig 5 MT1 receptor immunoreactivity in A: the infundibular nucleus (IF); B: mamillary body (MB); C: ventromedial nucleus (VMN); D: dorsomedial nucleus (DMN); E: paraventricular nucleus of the thalamus (PV/PT); F: median eminence (ME). Note the positive MT1 staining in the cytoplasm of neurons in all the brain areas and in the beaded nerve ﬁbers in the ME (ﬁgure F). Scale bars represent 50 µm.

Medium to strong MT1-ir was also observed in the paraventricular nuclei of the thalamus (PVT) (Fig. 5E). In the median eminence (ME) MT1-positive cells and beaded ﬁbers were found (Fig. 5F). MT1-ir was also observed in cells of the ependymal layer of the third and lateral ventricular wall, as well as in vascular smooth muscle cells and endothelial cells.

106 MT1 in the hypothalamus and pituitary

Fig 6 MT1 receptor immunoreactivity in the pituitary gland. A: strong staining in the pars tuberalis (PT) cells of the pituitary; B: weak granular staining in the gland cells of the anterior pituitary. Scale bars in A and B represent 25 µm.

Immunocytochemical MT1 staining in the pituitary MT1-ir distribution in the pituitary of 7 subjects is summarized in Table 3. Strong MT1-ir was observed in the pars tuberalis along the pituitary stalk up to the bottom of the ME of all subjects, in which this structure was available (n=8) (from one hypothalamus the pituitary stalk was absent)(Fig. 6A). Seven out of the eight pituitaries we studied showed weak MT1 staining, both in the posterior and anterior pituitary. The zona intermidia of the pituitary was not stained by MT1. In the anterior pituitary gland, weak granular cytoplasmic staining was observed in the gland cells (Fig. 6B). In the posterior part, MT1-ir showed up very weakly in nerve ﬁbers.

Table 3 MT1 receptor immunoreactivity intensity in the pituitary gland. NBB number Pars tuberalis Subject No. Anterior pituitary Posterior pituitary 97368 ++ 93-203 + + 97406 +++ 95-101 ± ± 99035 +++ 95-102 + ± 99205 +++ 95-103 + + 00185 ++ 96-081 ± + 01137 +++ 96-082 + + 01323 ++ 96-084 - - 93203 +++ 96-075 ± +

Scales: - no staining; ± very weak staining; + weak staining; ++ median intensive staining; +++ Strong staining

107 chapter 5

In the present study we found that subjects with longer postmortem delay (PMD) (longer than 10 hours) or longer fixation time (longer than 50 days) had similar MT1 receptor staining intensities and distribution patterns in the hypothalamus and pituitary gland compared to those with shorter PMD (shorter than 10 hours) or fixation time (shorter than 50 days). In addition, the 2 old controls (69 and 70 years old) showed a similar MT1 distribution of immunocytochemical staining in the hypothalamus as in the 7 young controls (from 25-40 yours old). We did not find a significant influence of gender, the clock time or season of death on the MT1 receptor distribution in the hypothalamus and pituitary gland.

Colocalization of MT1 with AVP, MT1 with OXT and MT1 with CRH in the anterior hypothalamus The intensity of the ﬂuorescent signal of the MT1 and AVP-ir neuronal cell bodies ranged from weak to very strong in the SCN, PVN and SON. The same holds true for OXT-ir neuronal cell bodies in the PVN and SON, and CRH-ir neuronal cell bodies in the PVN. We observed cytoplasmic colocalization of the MT1 and AVP in some of the neurons in the SCN, and in parvocellular and magnocellular neurons of the PVN and SON (Fig. 7). Colocalization of MT1 and OXT was also observed in part of the parvocellular and magnocellular neurons of the SON and the PVN (Fig. 8). Colocalization of MT1 and OXT seemed less frequent than the colocaliztion of MT1 and AVP in the PVN neurons. Colocalization of MT1 and CRH was observed in part of the parvocellular neurons of the PVN (Fig. 9). There were also single-labeled cells expressing for MT1, AVP, CRH or OXT, respectively, in all the sections.

Discussion The present study describes the immunocytochemical distribution of the melatonin MT1 receptor in the human hypothalamus and the pituitary gland. Moreover, we observed that the MT1 receptor was colocalized with the neuropeptide AVP in the SCN, PVN and SON, with OXT in the PVN and SON, as well as with CRH in parvocellular neurons of the PVN in the hypothalamus. Postmortem delay (PMD) and fixation time appeared not to affect the MT1-ir intensity and distribution pattern. This indicates that the MT1 receptor protein in the SCN is stable under prolonged postmortem and fixation conditions, which is in agreement with a study of melatonin binding in the mouse SCN 522. Previous animal studies have shown a diurnal rhythm of the density of 125I-melatonin binding in the hypothalamus 523, and of MT1 receptor mRNA in the SCN 406, 407. Moreover, aging diminishes melatonin binding and MT1 mRNA in the SCN of rodents 523, 524. In the present study, we did not find a significant effect of the clock time of death or age on the MT1-ir either in the SCN or in other hypothalamic regions. However, as we had limited number

108 MT1 in the hypothalamus and pituitary of cases (2 old and 7 young subjects), and the 7 subjects with a known time of death were not equally distributed over the 24-hour-day, it is not possible to draw any ﬁrm conclusions on these matters. We aim to explore the possible eﬀects of age and the diurnal cycle on MT1-ir in the human SCN into detail in a future separate study.

MT-ir distribution in the human hypothalamus MT1-ir is present as a non-nuclear cytoplasmic staining in the hypothalamic neurons, which is in accordance with previous observations in the human hippocampus 511. In agreement with a previous RT-PCR study which revealed a wide distribution of MT1 mRNA in the fetal human hypothalamus 347, we found in the adult human hypothalamus that the MT1-ir was also widespread, rather than being limited to the SCN area and PT as reported previously on binding and mRNA levels 200, 504. A number of novel sites of the MT1 receptor in the adult human hypothalamus were revealed, including the PeVN, PVN, SON, SDN-POA, NBM, DBB, ME, TM, MB. This more widespread hypothalamic expression pattern of MT1 compared to previous studies is probably due to the high sensitivity and better resolution of the immunocytochemitry in combination with the antigen retrieval procedure we used, as well as the systematic approach we followed. The role of melatonin in the regulation of circadian rhythms is well-deﬁned and it is regarded as the “circadian message” for the brain 502. From our study it has become apparent that the nocturnally restricted melatonin may, via the MT1 receptors that are expressed widely throughout the hypothalamus, not only inﬂuence the SCN rhythmicity, but also reinforce the endogenous circadian rhythmicity on other hypothalamic systems. Indeed, besides its circadian function, melatonin has been implicated in the regulation of sleep, reproduction, stress, mood, body temperature and cognition 57, 318, 331-334, many of which functions are mediated by the hypothalamus. Here we will focus on the MT1-expressing brain areas and neuropeptide systems which are involved in these functions to discuss the potential functional implication of the present study.

The SCN Experiments with transgenic mice with targeted disruption of the MT1 receptor have clearly established that melatonin exerts acute inhibitory effects on the spontaneous neuronal activity in the SCN via the MT1 receptor 64. Our present study shows the presence of the MT1 receptor in the human SCN on the protein level, confirming previous studies on the mRNA and binding level 525, 526, thus supporting the role of the MT1 receptor in the circadian effects of melatonin in humans. Consistent with previous animal studies 527, we found that the MT1 receptor is colocalized with AVP in the SCN neurons, indicating that this subpopulation of neurons is involved in melatonin’s action in the SCN. In addition, in vitro experiments have

109 chapter 5

MT1 AVP MT1 + AVP A B C

SCN

D E F

SON

G H I

PVN

Fig 7 Co-expression of the MT1 receptor and vasopressin (AVP) in part of the neurons in the suprachiasmatic nucleus (SCN) (A, B and C), paraventricular nucleus (PVN) (D, E and F) and supraoptic nucleus (SON) (G, H and I). A, D and G panels show single MT1 fluorescence (green) in the SCN, PVN and SON, respectively; B, E and H show single AVP fluorescence (red) in the SCN, PVN and SON, respectively; C, F and I show double fluorescence of MT1 and AVP (green, yellow and red) in the SCN, PVN and SON, respectively. Scale bar in the SCN (A, B and C) represents 25 µm. Scale bars in the SON (D, E and F) and PVN (G, H and I) represent 50 µm.

shown that melatonin can inhibit the spontaneous and stimulated release of AVP from the SCN 427, 528. The observed colocalization supports this possibility and adds that it may concern a direct melatonin eﬀect on the AVP neurons via the MT1 receptor.

110 MT1 in the hypothalamus and pituitary

OXT MT1 OXT + MT1 A B C

Fig 8 Co-expression of the MT1 receptor and oxytocin (OXT) in part of the neurons in the paraventricular nucleus (PVN). A. single OXT fluorescence (green); B. single MT1 fluorescence (red); C. double fluorescence of OXT and MT1 (green, yellow and red). Scale bar represents 50 µm.

CRH MT1 CRH + MT1 A B C

Fig 9 Co-expression of the MT1 receptor and corticotropin-releasing hormone (CRH) in part of the neurons in the paraventricular nucleus (PVN). A. single CRH fluorescence (green); B. single MT1 fluorescence (red); C. double fluorescence of MT1 and CRH (red, yellow and green). Scale bar represents 25 µm.

MT1 in the PVN and SON and colocalization with AVP, OXT and CRH A strong MT1-ir was found in the human hypothalamic PVN and SON, where a colocalization of MT1 with AVP and MT1 with OXT was observed. The SON and PVN are the major sources for AVP and OXT 529, which regulate water balance, blood pressure, reproduction, and the activity of the HPA axis (for review see 6). In plasma and urine, AVP and OXT normally show a circadian rhythm, with higher levels at night and lower levels during the day 156. The circadian rhythm of AVP and OXT was presumed to result from a direct projection of the SCN to the SON in rat 530. However, such a direct connection has not been found in the human hypothalamus, although

111 chapter 5

SCN fibers come close to the SON 24, 25. Melatonin has been shown to regulate the basal and stimulated neurohypophysial hormone AVP and OXT release in human subjects 507, 508. Our observation of a strong MT1 receptor expression and a colocalization with AVP and with OXT in the SON and PVN, suggests that the nocturnally increased levels of melatonin can act as a direct circadian message to regulate the neuroendocrine activity of the SON and PVN, thus contributing to the diurnal pattern of vasopressin and OXT levels and kidney function, without the need of a direct innervation of the SCN to the neuroendocrine cells. Here the colocaliztion of MT1 and CRH was found in pavocellular neurons of the PVN in the human hypothalamus. Melatonin is believed to exert a stress-protective effect 531, 532 and may be involved in depression (see review 212). Such an anti-stress effect of melatonin is proposed to be based on its inhibitory effect on the HPA axis 509, 510. Melatonin treatment attenuates the chronic stress and high doses of glucocorticoid induced CRH release from the hypothalamus in rat 509, 510. Moreover, a single oral dose of melatonin suppresses ACTH and cortisol secretion in blind humans 264. CRH produced in the PVN is a crucial neuropeptide in the regulation of the HPA axis, the final common pathway in the stress response 533. The observed colocalization of MT1 and CRH in the human PVN suggests that melatonin may act directly on the CRH neurons to modulate the secretory activity of the HPA axis. The molecular mechanism of such an action should be studied further.

MT1 in the NBMC, TM and MBC in relation to cognition and neuroprotection In addition to the MT1-ir and mRNA presence in the hippocampus 369, 512 and cortex 369, we observed the presence of MT1 receptor in neurons of brain areas involved in memory, attention and cognition, such as the NBM 534, 535, TM 536, 537, and MB 538-540. These systems are affected in aging and Alzheimer’s disease (AD) (for review see 6). Melatonin was reported to be an effective free radical scavenger, anti-oxidant and neuroprotector 541, 542. Some of these effects are membrane melatonin receptor-dependent 543, 544, and some are receptor-independent 545. Melatonin production is decreased during aging and even more so in AD 203. Moreover, melatonin production is significantly decreased in the aged cognitively intact subjects with the earliest AD neuropathological changes, i.e. a few tangles in the brain 546, compared to aged subjects devoid of any AD neuropathology 254. Administration of melatonin increases the survival and inhibits oxidative and amyloid neuropathology in Alzheimer transgenic mice, suggesting the involvement of decreased melatonin in the pathogenesis of AD 304. Indeed, melatonin supplementation has shown beneficial effects on cognition in AD patients 26, 163, 547, 548. Our study suggests that such valuable effects of melatonin may be partially mediated via the MT1 receptor expressed in the NBM, TM and MB.

112 MT1 in the hypothalamus and pituitary

MT1 receptor expression in the IFN, ME and pituitary gland in relation to reproduction The role of melatonin in reproductive processes has been known for a long time (for review see 251, 331). Melatonin probably regulates reproduction via melatonin receptors at different levels of the hypothalamus-pituitary-gonadal axis: the hypothalamus, pituitary, gonads and reproductive tissues (see review 331). The MT1 receptor is present in the preoptic and/or mediobasal hypothalamus of various mammals 549, and shows up in the pars tuberalis (PT) of the anterior pituitary of all mammals studied 63, 550, 551. Since the Siberian hamster, which lacks a functional MT2 receptor, still shows seasonal reproduction, an essential role is proposed for the MT1 receptor in reproductive processes 525. In agreement with previous mammalian studies, the present study shows that the MT1 receptor is expressed in neurons of the preoptic and mediobasal hypothalamus, such as the sexually dimorphic nucleus of the preoptic area, infundibular nucleus and median eminence in the human hypothalamus. Moreover, strong MT1 receptor expression was observed in the pars tuberalis of the pituitary. In addition, a weak granular MT1 staining was observed in the gland cells of the anterior pituitary and in the nerve fibers of the posterior pituitary, which agrees with the low expression level of MT1 reported in the pituitary of adult rats 552, 553. Our observations thus suggest that melatonin might not have a strong direct effect on the pituitary itself, but rather may exert its reproductive effects via MT1 receptors in the hypothalamus. In conclusion, the expression of MT1, the major melatonin receptor in the mammalian brain, is expressed in many nuclei of the human hypothalamus, which suggests that melatonin may directly modulate a vast array of hypothalamic physiological processes via this receptor.

113

Chapter 6

Decreased MT1 melatonin receptor expression in the suprachiasmatic nucleus in aging and Alzheimer’s disease

Ying-Hui Wu, Jiang-Ning Zhou, Joop Van Heerikhuize, Ralf Jockers, Dick F. Swaab

Neurobiology of Aging, 2006 Jul 10; [Epub ahead of print]

Abstract The pineal hormone melatonin is involved in the regulation of circadian rhythms and feeds back to the central biological clock, the hypothalamic suprachiasmatic nucleus (SCN) via melatonin receptors. Supplementary melatonin is considered to be a potential treatment for aging and AD-related circadian disorders. Here we investigated by immunocytochemistry the alterations of the MT1 melatonin receptor, the neuropeptides vasopressin (AVP) and vasoactive intestinal peptide (VIP) in the SCN during aging and AD. We found that the number and density of AVP/VIP-expressing neurons in the SCN did not change, but the number and density of MT1-expressing neurons in the SCN were decreased in aged controls compared to young controls. Furthermore, both MT1-expressing neurons and AVP/VIP-expressing neurons were strongly diminished in the last neuropathological stages of AD (Braak stages V-VI), but not in the earliest stages (Braak stages I-II), compared to aged controls (Braak stage 0). Our study suggests that the MT1-mediated eﬀects of melatonin on the SCN are disturbed during aging and even more so in late stage AD, which may contribute to the clinical circadian disorders and to the eﬃcacy of therapeutic melatonin administration under these conditions.

Introduction The suprachiasmatic nucleus (SCN) in the anterior hypothalamus is considered to be the self-sustained master circadian pacemaker. Via neuronal and/or hormonal pathways the SCN coordinates circadian rhythms in a variety of biochemical, physiological, and behavioral processes, including the rhythm of pineal melatonin production 53. Melatonin is involved in the regulation of circadian rhythms and appears to feedback to the SCN via two specific, high-affinity G protein-coupled melatonin receptors, i.e. 62 MT1 (also called Mel1a) and MT2 (or Mel1b) (reviewed in ). Physiologically, the MT1 receptor mediates the acute inhibitory action of melatonin on the SCN 64, which may be important for defining SCN sensitivity to phase-shifting stimuli, and may contribute to the regulation of sleep. The MT2 receptor is reported to mediate the phase shift effect of melatonin on the SCN of mammals 65, 66. The circadian effects of melatonin

117 chapter 6 have led to substantial therapeutic applications for jet lag, shift work, blindness, and some circadian-based sleep disorders 6, 60. Specific 125I-melatonin binding sites have been revealed in the SCN area of mammals, including human 200. The MT1 receptor is the primary subtype of melatonin receptors as targeted disruption of the MT1 receptor eliminates detectable 125I-melatonin binding from the mouse brain 64. In addition, MT1 receptor mRNA has been detected in the post-mortem human SCN in an overlapping pattern with the distribution of 125I-melatonin binding, while MT2 subtype was not detected, probably due to a very low expression level in the human hypothalamus 347, 504. Sleep disruptions, nightly restlessness and other circadian rhythm disturbances are frequently seen in elderly and even more severely so in Alzheimer’s disease (AD) patients 148, 189. These disturbances actually are the main reasons for hospitalization of AD patients 165. Our previous studies showed degeneration of the SCN during aging, and particularly in AD, reflected by the decrease of the neuropeptide vasopressin at both the protein and the mRNA level 23, 173. Moreover, melatonin levels are decreased during aging and predominantly in AD with flattened circadian rhythm 202, 203, in fact, from the earliest AD neuropathological stages onwards 254, 302. Such changes in the SCN and pineal gland are considered to be the neurobiological basis for the circadian disturbances in aging and AD 202, 203. However, so far no data are available concerning the possible involvement of the MT1 receptor in the SCN in these circadian disorders. In addition, as supplementary melatonin is considered to be a potential treatment for aging and AD-related circadian disturbances, it is a prerequisite to investigate the alterations of the MT1 receptor in the SCN in these processes. In the present study, we studied the alterations of the MT1 receptor in the SCN during aging and the progression of AD by using immunocytochemistry. We found that the MT1 receptor expression in the SCN was decreased in aging. Moreover, MT1 in the SCN was even more strongly diminished in the late neuropathological stages of AD (Braak stages V-VI), but not in the early stages (Braak stages I-II). Our study thus suggests that effects of melatonin on the SCN, mediated by the MT1 receptor, may be disturbed during aging and even more so in late stage AD, which may contribute to the clinical circadian disorders and to the efficacy of therapeutic melatonin administration under these conditions.

Subjects and Methods Subjects Brain material was obtained via the rapid autopsy system of the Netherlands Brain Bank (NBB) at the Netherlands Institute for Neuroscience, in accordance with the formal permissions for a brain autopsy and the use of human brain material and clinical information for research purposes. We studied the hypothalamus of 46 subjects,

118 MT1 in the SCN in aging and AD subdivided into 4 groups: 13 young controls, 11 old controls, 11 preclinical “AD” subjects and 11 late clinical AD patients. Young controls (aged 19~40 years old) and aged controls (61~85 year old) were free of any psychiatric and neurological diseases, and without any AD neuropathological change in the brain (Braak stage 0). Preclinical “AD” subjects (64~87 year old) were cognitively intact subjects but appeared to have minor AD neuropathology (Braak stages I-II). Late clinical AD patients (59~86 year old) fulfilled the NINCDS-ADRDA (National Institute of Neurological and Commu- nicative Diseases and Stroke/Alzheimer’s Disease and Related Disorders Association) criteria 436, i.e. they had a clinical diagnosis of probable AD, excluding other causes of dementia by means of history, physical examination and laboratory tests, and were neuropathologically confirmed with systematic and extensive AD neuropathology (Braak stages V-VI)177, 433. All the groups were matched for sex, season of death (according to their month of death, Summer: Jun-August, Winter: Dec-Feb), day/night distribution (according to their clock time of death, day: 1000-2200 h, night: 2200- 1000 h) (P=0.79, P=0.22 and P=0.46, respectively)(for diurnal and seasonal fluctuations in the human SCN see 169, 170). In addition, aged controls, preclinical “AD” subjects and late clinical AD patients were matched for age (P=0.78). Clinico-pathological data are presented in Table 1. There were significant differences in brain weight, postmortem delay, and fixation time between groups (P=0.02, P=0.01, P=0.01, respectively). However, later linear regression analysis indicated that these parameters did not seem to affect our data comparison (all P>0.05).

Histology Hypothalamic samples were fixed in formalin and embedded in paraffin. Serial coronal sections (6 μm) were made from the level of the lamina terminalis to the mamillary bodies. Depending on availability, either the left or right hemi-hypothalamus was used. For anatomical orientation every 100th section was collected and mounted on superfrost plus slides (Menzel GmbH & Co KG, Baunschweig, Germany) and subsequently dried for at least 2 days at 37 ºC, followed by Nissl staining (0.5% thionine in distilled water). Additionally, in order to delineate the SCN, double immunocytochemical staining for vasopressin (AVP) and vasoactive intestinal polypeptide (VIP) was performed on every 50th section taken along the rostrocaudal axis throughout the complete SCN region. The rostral and caudal border of the SCN was defined as the most rostral and most caudal section that contained one or more AVP or VIP positive cells in the area. The central cross-section containing the maximal SCN was defined as the central section containing the most AVP and VIP positive cells, to which adjacent central SCN sections were taken for the MT1 receptor staining.

119 chapter 6

Table 1 Clinical-pathological data of subjects. - - Myocardial infarct infarct Myocardial death Sudden dead in bed Found Bronchopneumonia/bronchitis respiratory deficiency syndrome, immune Acquired bronchopneumonia to due insufficiency capil based a pulmonary on Pulmonary hypertension, the of rupture secondary traumatic to Haematothorax accident) car a (after ventricle left aortaA thoracic of rupture fractures; Multiple perotonitis Faecal infarction Heart failure organ Sepsis multi with heart failure Acute with infection a pulmonary of a combination Probably in the abdomen abscess an ischemi mesenterial with Complications cystitis, pyelitis Bronchopneumonia, Retroperitoneal chondrosarcoma/septic shock chondrosarcoma/septic Retroperitoneal pneu haemorrhagic fibronous acute and failure Heart coma multiform, Glioblastoma stomach, acute with perforation ileus, with Probable infarct myocardial probably metastasi with cancer Severe prostate carcinoma a bladder of Bone metastases possibly failure Heart cancer Pancreatic Pneumonia Cause of death Cause of monia, next to a B-cell lymphoma to next monia, lary hemangiomatosis lary hemangiomatosis

7:45 8:35 8:00 5:15 8:35 5:55 7:45 6:45 4:20 5:11 19:35 19:35 30:00 85:40 26:45 PMD <8:15 <33:30 <17:00 <41:00 <46:25 <71:30 <16:30 <10:37 <17:45 <20:00 <56:20 (h:min)

51 35 34 82 65 31 31 45 45 72 51 61 98 28 29 36 55 45 34 34 43 35 28 Fix Fix 130 119 1390 (day) (g) 975 925 ND BW 1400 1311 1900 1490 1364 1500 1280 1200 1287 1588 1420 1348 1400 1240 1311 1296 1154 1169 1475 1560 1383 1310 1400

1 5 4 8 1 4 2 8 8 1 1 7 7 9 7 7 10 11 10 12 12 12 11 11 Month Month of Death of

ND ND 0:25 8:30 7:30 4:00 8:45 0:30 6:23 2:55 8:00 8:40 16:55 18:35 19:20 17:30 17:55 23:15 16:30 21:00 11:20 11:00 16:45 12:45 of death of Clock time f f f f f f f f f f f f f f m m m m m m m m m m m m Sex

32 69 19 20 21 24 25 32 32 32 33 36 36 39 40 61 61 63 64 65 69 70 72 78 83 85 (y) Age Age NBB 81-017 94-040 01-009 97-173 01-072 92-037 96-239 02-006 98-031 84-002 91-009 99-071 99-125 96-081 99-033 94-114 01-004 98-035 97-157 00-090 98-101 99-116 82-019 94-074 Old control Group Median (Braak stage 0) stage (Braak Median Young control Young

120 MT1 in the SCN in aging and AD 3:55 Primary pancreatic tail carcinoma Primary pancreatic arrest Cardiac accident Cerebrovascular unexpected in sleep death Pneumonia, Bronchopneumonia at located secondary a blockage to pancreatitis Acute bronchopneumonia by accompanied papil the Vater in the groin Hemorrhage complication: PTCA Post Myocardial infarction: lung oedema, heart tamponade oedema, heart tamponade lung infarction: Myocardial secondary CARA, coupled severe to failure Cardiac dehydration cachexia/ with carcinoma pancreas Septic Metastasized shock; arres Cardiac deterioration Sudden Deterioration pneumonia Aspiration convulsions/dehydration Pneumonia/epileptic Dehydration Pneumonia Pneumonia arrest Cardiac deterioration physical general with Dementia pneumonia Cachexia, pneumonia Dehydration; Cause of death Cause of

8:30 9:10 2:40 7:25 3:55 3:40 5:55 5:30 3:24 4:00 4:20 3:40 3:10 3:15 4:15 12:50 12:50 22:10 19:15 10:45 PMD <63:00 <65:30 <53:15 (h:min)

31 43 28 43 31 34 56 77 47 35 34 29 30 37 44 27 33 45 32 28 26 31 Fix Fix 147 ND (day) (g) 825 ND BW 1265 1150 1417 1030 1265 1074 1352 1067 1079 1235 1078 1379 1074 1050 1218 1219 1150 1390 1068 1065 1315 1085

6 4 1 5 8 5 1 9 2 4 3 1 7 2 9 7 3 6 11 12 12 12 Month Month of Death of

2:00 2:00 3:30 8:30 9:45 8:25 5:45 0:05 4:15 21:00 16:00 13:50 16:30 11:45 13:15 16:40 13:25 20:35 15:45 17:55 22:00 19:30 of death of Clock time f f f f f f f f f f f f f f f m m m m m m m m m Sex 69 72 64 65 65 69 69 72 72 76 77 82 87 59 64 64 66 66 69 75 76 82 83 86 Age (y) Age NBB 93-127 97-042 98-003 96-057 99-101 95-054 98-034 98-183 97-156 98-016 98-049 97-033 94-012 94-082 92-099 95-014 95-087 94-126 99-138 96-068 93-040 91-097 Clinical AD Group I-II) disease; lung non-speciﬁc delay; time; PMD: postmortem CARA: chronic Fix: ﬁxation weight. Brain BW: Bank number. Brain Netherlands y: angioplasty; m: male; f: female. coronary years; transluminal Percutaneous PTCA: (Braak stages stages (Braak Median stage (Braak V-VI) Median Preclinical AD

121 chapter 6

AVP and VIP immunocytochemical double staining The combined immunocytochemical staining for AVP and VIP followed the protocols as previously described with some minor modiﬁcations 23, 310, which are as follows: deparaﬃnized and rehydrated sections were 1) microwave-treated for antigen retrieval at 90 ºC for 15 min in TBS (0.05 M Tris, 0.15 M NaCL, pH 7.6); 2) incubated with rabbit anti-AVP antibody (Truus, 29-01-1986) at 1:800 and rabbit anti-VIP antibody (Viper, 18-09-86) at 1:1000 in Supermix [SUMI: 0.25% gelatin (wt/ vol), 0.5% Tri- ton X-100 (vol/vol) in TBS)] for one hour at room temperature (RT), followed by overnight incubation at 4 ºC; 3) washed in TBS and incubated with goat-anti-rabbit serum (Akon) at 1:100 in SUMI (1 h, RT); 4) washed in TBS and incubated with peroxidase-antiperoxidase (PAP) at 1:1000 in SUMI (1h, RT); 5) washed in TBS and incubated with 3.3’-diaminobenzidine (DAB) (Sigma) 0.5 mg/ml in TBS, containing

0.23% (wt/vol) nickelammoniumsulfate (Merck) and 0.01% (vol/vol) H2O2 (Merck) for approximately 20 min (RT) for visualization of the signal; 6) washed in distilled water, dehydrated in an ascending series of alcohol and xylene and coverslipped in Entellan (Merck).

MT1 immunocytochemical staining From each subject two consecutive central sections containing the maximal SCN area were stained with a polyclonal affinity-purified specific anti-MT1 receptor antibody (provided by Dr. Ralf Jockers, Institute Cochin, Department of Cell Biology, Paris, France). The anti-MT1 receptor antibody was raised in a rabbit, directed against a peptide corresponding to a sequence found in the C-terminal region of this receptor (Ac-YKWKPSPLMTNNNVVKVDSV-COO2, peptide 536), and affinity-purified 378. This antibody detects the MT1 receptor as a protein with an apparent molecular weight of 60KD in immunoblots after separation by SDS-PAGE378 . It also specifically precipitates the 2-[125I]iodomelatonin–labeled receptor from MT1-transfected HEK 293 cells 378 and has been used before in human postmortem brain material 512. For our studies we also performed staining with omission of the primary anti-MT1 antibody or preadsorption with the homologous peptide 513, both completely eliminated any MT1 staining in the human hypothalamus. Moreover, staining of Western blots of the human hypothalamus with anti-MT1 receptor revealed bands of the expected molecular mass 60 KD for the MT1 receptor protein. These data further supported the specificity of the antibody in our system. Deparaffinized and rehydrated sections were stained for the MT1 receptor as follows: 1) pretreated with 3 μg/ml proteinase-K (Invitrogen, Breda, Netherlands) in Proteinase-K buffer (2mM CaCL2, 0.01M Tris/HCl pH 7.5) at 37 ºC for 15 min; 2) rinsed in glycine buffer (2 g glycine in 1000 ml PBS) for 30 sec to stop the proteinase-K reaction; 3) pre-incubated with TBS-milk [3% nonfat dry milk powder in TBS,

122 MT1 in the SCN in aging and AD

ELK, Campina] for 1h at RT; 4) incubated with anti-MT1 puriﬁed antibody at 1:200 in Supermix-milk (3% nonfat dry milk powder in SUMI, pH 7.6) 1 h at RT, and followed by overnight incubation at 4 ºC in a moist chamber; 5) washed once with TBS- milk and twice with TBS, followed by incubation in biotinylated goat-anti-rabbit IgG (Vector Laboratories Inc., Burlingame, CA) at 1:400 in SUMI (1 h, RT), 6) washed with TBS and incubated with avidin biotin complex (ABC) (Vector Laboratories, Inc. Burlingame, CA) at 1:800 in SUMI (1 h, RT) and rinsed twice in TBS; 7) incubated in biotinylated tyramide 1:500, 0.01% H2O2 (Merke) in TBS at RT for 15 min, a signal ampliﬁcation procedure described in detail by Adams 515; 8) rinsed twice in TBS and incubated in ABC complex 1:800 in SUMI (1h RT) and then again rinsed twice in TBS; 9) a similar procedure of visualization was performed as described in Section 2.3.

Image analysis By using an IBAS image analysis system, we determined the number of AVP/VIP- expressing cells (i.e. cells that are positive for AVP and/or VIP) and the number of MT1-expressing cells in the central SCN section, without information regarding the clinicopathological data of subjects. This image analysis system (ImagePro v4.5, Me- dia Cybernetics, Silver Spring, USA) was connected to a color camera (JVC KY-F55 3CCD) and a plain objective microscope (Zeiss Axioskop with Plan-NEOFLUAR Zeiss objectives). For each section to be analyzed, an image covering the SCN (2.5x objective) was loaded into the IBAS and displayed on the computer monitor. The X-Y scanning stage coordinates of the position of this image in the section were stored. In the image of an AVP/VIP double stained central SCN section (2.5x objective), we manually drew an outline of the SCN. This outline was also transferred to the image of the adjacent MT1 stained section (2.5x objective), which is exactly matched with the AVP/VIP staining section image, for defining its SCN area. After SCN area out- lining, a grid of rectangular areas superimposed the image, each covering a full-size image on the computer screen at 40x objective. From this grid all fields were selected. On the basis of the pixel positions of the selected fields in the image and the stored scanning stage position of the 2.5x image, a list of scanning stage coordinates was calculated to put the selected fields in view when looking through a 40x objective. Next, the 40x objective was positioned in front and under program control the scanning stage moved to the calculated positions. The positive cells that presented with a nucleolus in the field and did not touch an exclusion line were counted and stored in the analysis system. The positive cell number was thus determined in the central SCN section of each subject. The number of positive cells per unit SCN volume was estimated as the density of positive cells in the SCN, with the section thickness (6 μm) taken into account.

123 chapter 6

Statistical Analysis Differences among groups were tested using the Kruskal-Wallis and Mann-Whitney test. Differences between day and night within groups were tested with the Mann- Whitney U test. Differences of distributions in gender and day/night period between groups were tested by Chi-square. Linear regression analysis was performed to study the effects of postmortem delay (PMD), brain weight and fixation time on the data set, using the non-parametric Spearman correlation coefficient. Values of P<0.05 were considered to be significant (2-tailed). All values are expressed as mean ± standard error of the mean (Mean ± SEM).

Results Distribution of MT1-immunoreactive (MT1-ir) neurons in the anterior hypothalamus MT1-ir cells were widely distributed over the human anterior hypothalamus. MT1-ir was observed as granular cytoplasmic staining in the neurons, not in glia cells. MT1-ir neurons were not only present in the area of the SCN, but also in the paraventricular nucleus (PVN), supraoptic nucleus (SON), periventricular nucleus (PeVN), sexually dimorphic nucleus of the preoptic area (SDN-POA), diagonal band of Broca (DBB), horizontal limb of the diagonal band of Broca (hDBB), and nucleus basalis of Meyn- ert (NBM) in the anterior hypothalamus at a medium to strong staining level. Weak staining was observed in the lateral hypothalamic area (LHA). The overall intensity of the MT1 staining in the SCN was lowest in the late stage AD patients and highest in young controls (Fig. 1).

Decreased MT1-ir neurons in the SCN in aging and in AD Both the number (129.0 ± 26.0) and density (25,946 ± 4,996 neurons/mm3) of MT1- ir neurons in the central SCN section were reduced in elderly controls compared to young controls (225.8 ± 25.6) (47,555 ± 6,818 neurons/mm3) by 1.75 fold (P=0.02) and by 1.8 fold (P=0.03), respectively (Fig. 2). Compared to the age-matched elderly controls (Braak stage 0), preclinical AD subjects (Braak stages I-II) did not show a significant difference in the number of MT1-ir neurons (132.5 ± 19.6) (P=0.8) or in the density of MT1-ir neurons (23,597 ± 3,850 neurons/mm3) (P=0.6) in the central SCN section (Fig. 2). However, the late stage clinical AD patients (Braak stages V-VI) showed dramatic decreases in the number of MT1-ir neurons (26.1 ± 7.8) and in the density of MT1-ir neurons (9,275 ± 3,163 neurons/mm3) in the central SCN section compared to the age-matched elderly controls (P=0.001, P=0.004, respectively) and preclinical AD subjects (P=0.001, P=0.009, respectively) (Fig. 2). We did not detect a significant day/night difference in the number or the density of MT1-ir neurons in the central SCN section in any of the four groups (all P > 0.1).

124 MT1 in the SCN in aging and AD

Fig. 1 Melatonin receptor MT1 immunocytochemical staining in the SCN of (a) a young control, (b) an aged control (Braak stage 0), (c) a subject in early AD stages (Braak stages I-II; preclinical AD), (d) a late stage AD patient (Braak stages V-VI). Note that the MT1 receptor staining intensity in the SCN is decreased in the aged control and more strongly diminished in the late stage AD patient. Scale bar indicates 25 µm.

AVP and/or VIP immunoreactive (AVP/VIP-ir) neurons in the SCN in aging and AD No signiﬁcant diﬀerence was found in the number of AVP/VIP-ir neurons, or in the density of AVP/VIP-ir neurons in the central SCN section, or in the surface area of the central SCN between young controls (150 ± 15.7) (30,558 ± 4,604 neurons/ mm3)(0.92 ± 0.08 mm2), elderly controls (Braak stage 0) (130 ± 15.5) (28,527 ± 3,670 neurons/mm3) (0.83 ± 0.11 mm2), and preclinical AD subjects (Braak stages I-II) (137 ± 14.8) (27,100 ± 3,700 neurons/mm3) (0.94 ± 0.13 mm2) (P=0.6, P=0.5, P=0.9; respectively) (Fig. 3). However, the late stage clinical AD patients (Braak stages V-VI) did show dramatic decreases in the number of AVP/VIP-ir neurons (51 ± 7.8), in the density of the AVP/VIP ir neurons in the central SCN section (16,305 ± 2,940 neurons/mm3), and in the surface area of the central SCN compared to the age-matched elderly controls (P=0.0004, P=0.01, P=0.04; respectively) and preclinical AD subjects

125 chapter 6

a b p=0.02 250 60000 p=0.03 tion ec ) 3 p=0.001 50000 N s

200 mm SC

er p=0.004 p=0.001 al

(p 40000 tr

150 ty en si p=0.009

r c 30000 en pe

100 n d r.

ro 20000 u n n ne ro 50 ir

eu 10000 T- r n M -i 0 0 MT Young Old EarlyAD LateAD Young Old EarlyAD LateAD

Fig. 2 (a) MT1-immunoreactive (MT1-ir) neuron number per central SCN section and (b) MT1-ir neuron density in the central SCN section in young controls, aged controls (Braak stage 0), subjects in early AD stages (Braak stages I-II; preclinical AD) and late stage AD patients (Braak stages V-VI). Note that both MT1-ir neuron density and number are decreased in the old controls and preclinical “AD” subjects, and are more dramatically decreased in the late stage clinical AD patients.

(P=0.0003, P=0.02, P=0.01; respectively) (Fig. 3). No signiﬁcant day/night diﬀerence in the number or the density of AVP/VIP-ir neurons in the central SCN section was found in any of the four groups (all P > 0.05).

Discussion Our study provides the ﬁrst immunohistochemical evidence for the presence of the melatonin MT1 receptor in the human SCN, and the cellular distribution of MT1 at the protein level. Furthermore, we found a decreased MT1 receptor expression in the SCN during aging, which is even more pronounced in late stage Alzheimer’s disease patients. We observed that both the number and the density of MT1-ir neurons were decreased in the human SCN in aged control subjects compared to young controls. These results are in a good agreement with previous animal studies, which showed an age-related decline of 2-[125]iodomelatonin binding and MT1 mRNA in the SCN of rat and mice 524, 554. Moreover, we found that not only the number and density of MT1-ir neurons, but also the MT1 staining intensity was greatly decreased in the SCN in late stage AD patients (Braak V-VI) compared to age-matched controls (Braak 0). Such changes were not present in preclinical AD subjects (Braak I-II) compared to age-matched controls (Braak 0). This indicates that the decrease of MT1 receptor in the SCN is not an early event during the process of AD. Interestingly, in contrast to the MT1-ir neurons, the number and the density of AVP/VIP-ir neurons in the central SCN did not decrease in aged controls (mean age:

126 MT1 in the SCN in aging and AD

n a b ctio 180 se ) 3 35000

CN 160 mm l S

140 er 30000 ra (p nt

120 ty 25000 ce si er 100 en 20000 . p n d nr 80 ro 15000

60 eu uron

r n 10000 ne 40 -i ir IP V

P- 5000 20 / VI VP P/ 0 A 0 Young Old Early AD Late AD Young Old Early AD Late AD AV c 2 1.2 m ) (m

N 1.0 C S l ra

t 0.8 en c

e 0.6 h t f

a o 0.4 re

e a 0.2 ac rf u S 0.0 Young Old Early AD Late AD

Fig. 3 (a) Vasopressin (AVP) and/or vasoactive intestinal peptide (VIP) immunoreactive (AVP/VIP-ir) neuron number per central SCN section, (b) AVP/VIP-ir neuron density in the central SCN section and (c) the surface area of the central SCN in young controls (Braak stage 0), subjects in early AD stages (Braak stages I-II; preclinical AD) and late stage AD patients (Braak stages V-VI). Note that these parameters only decrease in the late stage AD patients compared the other groups. *denotes p<0.05 compared to the other three groups.

70-year-old) compared to young controls (mean age: 30-year-old). Since in an AVP- VIP double stained human SCN section, the majority of the immunoreactive cells are AVP-expression neurons (about 80%) 555, our results are in full agreement with a previous study showing that the number of AVP-expressing neurons in the SCN did not change until the age of 80 years 23. Our study thus suggests that the MT1 receptor expression is disrupted earlier than the expression of neuropeptides AVP and VIP in the SCN during aging. Actually, functional disturbances of SCN circadian activity may start already around the age of 50, reﬂected in a disruption in the circadian rhythmicity of the AVP-expressing neurons in the SCN 169, while the number of AVP-expressing neurons is still intact 23. Thus the age-related decrease of the MT1 receptor in the SCN may be due to a functional disruption of the SCN rather than, to the loss of SCN neurons. Furthermore, consistent with previous studies 23, 173, 310, a marked decrease of AVP/VIP-ir neurons in the SCN was observed in late stage AD patients, which

127 chapter 6 accompanied a strong decline of MT1-ir neurons. This can be explained in different ways. In the first place it may indicate a prominent loss of SCN neurons in late AD patients, with concomitant loss of MT1 receptors in their SCN. Alternatively, it may indicate that both the MT1 receptor expression and the expression of the neuropeptides AVP and VIP are decreased in SCN neurons, but that the neurons still exist in late AD patients. In support of the latter hypothesis, we did observe in the SCN of late AD patients, in contrast to other groups, much more negative neurons than positive neurons for AVP/VIP and MT1 expression. Moreover, it is possible to improve circadian rhythmicity in late AD patients by increasing the intensity of environmental light that will reactivate SCN neurons 322, 556-558, an effect that has been confirmed in a recent longitudinal placebo-controlled trial of our group (Van Someren et al., unpublished data). Taken together, we prefer the latter explanation. Melatonin exerts an acute inhibitory effect on the SCN neurons mediated by the MT1 receptor 64, by which melatonin may help to define the neuronal sensitivity to phase-shifting stimuli that normally keep the biological clock in tune with the environment. Since both the endogenous melatonin levels (see reviews 202, 203) and the MT1 receptor in the SCN are decreased during aging, and even more so in late AD patients, the neurobiological effects of melatonin, including the circadian effects on the SCN mediated by the MT1, may be affected in elderly and even more severely in late AD patients, and contribute to their enhanced incidence of circadian rhythm disturbances, such as sleep-wake disorders 189. Interestingly, although the density and number of the MT1-ir neurons in the SCN of aged controls are only half of those in young subjects, exogenous melatonin has still shown to be effective in aged subjects with insomnia (see review 559). This indicates that the expression of MT1 receptors in the SCN of aged controls is still sufficient to respond to exogenous melatonin. This may also be the case in preclinical AD subjects who have similar expression levels of the MT1 receptor in the SCN as elderly controls, but this possibility should be confirmed in clinical trials. Previously in some pilot studies and case reports, melatonin has shown some benefit for circadian disturbances in Alzheimer demented patients, which, however, was not confirmed in recent well- performed placebo-controlled clinical trials of melatonin 319, 320. In fact, these negative results may be partly explained by our study: the number of MT1 receptor containing neurons in the SCN of late clinical AD patients (Braak stage V-VI) is only 10% of those in age-matched control subjects, which may well be too few to mediate an efficient action of the melatonin supplement. Animal studies have shown that peripheral melatonin levels have auto-regulatory effects on its receptor expression in the SCN, i.e. decreasing circulating melatonin levels by pinealectomy increases melatonin receptor expression in the SCN, whereas a single melatonin injection results in decreased melatonin receptor expression in the

128 MT1 in the SCN in aging and AD

SCN 400, 403. Since melatonin levels decrease in the course of aging and even more dramatically so in Alzheimer’s disease (see reviews 202, 203), one would expect an increase of the MT1 receptor in the SCN during aging and AD, according to the auto-regulatory effect of melatonin. However, we observed the opposite. Our study suggests that either this auto-regulatory feedback loop does not exist in human, or it is disturbed during aging and AD, possibly due to degeneration of the SCN neurons. Melatonin receptors in the SCN of rodents have shown diurnal rhythms and are influenced by light/dark conditions 401, 403. We therefore matched the subjects for clock time of death and season of death to obtain a similar environmental light/dark exposure between the groups, so that the day/night cycle could not influence our group comparison. Although in the present study we did not find a significant diurnal pattern of the MT1 receptor in the SCN of young or aged controls, due to the limited number of subjects and large individual variations, a definitive conclusion on the daily rhythm is not possible. In our previous studies on the number of AVP cells in the SCN, measurements in the central section containing the maximal SCN area reflected the data obtained by measurements throughout the SCN 23. Therefore, in the present study, we measured the MT1-ir neurons numbers and density only in the central SCN section, which already showed a significant diminishment of the MT1 receptor during aging and an even stronger decrease in AD. Moreover, since the volume of the SCN decreases in old controls over 80 years and even more so in AD patients, while the SCN cell density stays the same 23, MT1 expression integrated over the whole SCN would show an even more pronounced change with aging and AD than reported here. The MT2 receptor is another subtype of melatonin receptor, and was reported in the SCN of rodents mediating the phase shift effect of melatonin 65, 66. It was shown that the phase shift responsiveness to melatonin is identical in young and aged mice, which suggests a functionally intact MT2 receptor in the SCN during aging 524. In the human brain, MT2 receptor expression has only been reported in the hippocampus, which is reduced in AD patients 391. It has never been detected in the human SCN, possibly due to an extremely low expression level 347, 504. Whether and how the MT2 receptor expresses in the human SCN and changes during aging and AD needs further study. In summary, the MT1 receptor in the SCN is decreased in aged subjects and more dramatically so in late stage clinical AD patients. The reduction in MT1 receptor may contribute to the circadian disorders during aging and AD.

129

Chapter 7

Increased number of neurons expressing melatonin receptor MT1 in the suprachiasmatic nucleus in depression, and its relation to age at onset and disease duration

Ying-Hui Wu, Jiang-Ning Zhou, Ai-Min Bao, Ralf Jockers, Joop Van Heerikhuize, Dick F. Swaab

Abstract The pineal hormone melatonin is involved in the regulation of circadian rhythms. This regulation partly depends on its feedback to the central biological clock of the brain, the suprachiasmatic nucleus (SCN) via melatonin receptors. Melatonin circulating levels are decreased in depressed patients, who frequently suffer from circadian rhythm disturbances, such as sleep disorders. Administration of melatonin is considered a potential effective therapy for these circadian disorders in depression. In the present study we investigated the alterations of the MT1 receptor, the most abundantly expressed melatonin receptor subtype in the SCN in depression by immunocytochemistry. 14 depressed patients (7 major depression, 7 bipolar depression) and 14 controls were selected and matched for age, sex and clock time at death. We found that in depressed patients the MT1-immunoreactive neuron density and numbers in the central SCN were more than 1.8 times higher than those in controls. Moreover, the MT1-expressing neuron density and numbers are negatively correlated with the age at onset, while positively correlated with the disease duration in depressed patients. Our study suggests an enhanced protein expression of MT1 receptor in the SCN of depressed patients, which may be accumulatively increased during the course of disease. This may contribute to the efficacy of melatonin administration in depression.

Introduction The suprachiasmatic nucleus (SCN) in the hypothalamus is considered to be the central biological clock of the mammalian brain, generating and coordinating circadian rhythms throughout the body, including that on melatonin synthesis in the pineal gland via a multi-synaptic pathway 53. Melatonin feeds back on the SCN via melatonin receptors and influences circadian rhythms 560. In mammals there are two G-protein coupled melatonin receptors, i.e. MT1 and MT2. Both are present in the SCN of rodents, where they mediate the acute inhibitory effect and phase shift effect of melatonin, respectively 63, 64, 67. However, in the human SCN only MT1 mRNA was observed, whereas MT2 mRNA was not detected by in situ hybridization or RT-PCR, possibly due to its extremely low expression levels 347, 504.

133 chapter 7

Abundant evidence has shown that the circadian system plays a role in the pathogenesis of depression. There is a diurnal variation in the depressive state205, 206, and 90 % of the patients with major depression suﬀer from sleep disturbances 204. A large number of studies have, moreover, observed a reduced daily production of melatonin, with a reduced and phase-advanced nocturnal peak in both patients with major depressions, bipolar and seasonal depression 213, 561-566. Decreased nocturnal melatonin levels were associated with depressed mood and other clinical manifestation 213, 567. Furthermore, melatonin administration has been suggested as a potential treatment for mood disorders, which appeared to improve their sleep disorders, such as insomonia, early waken and short sleep length 214, 547, 568-572, and alleviate the depressed mood 214, 547, 568, 573. However, the possible involvement of the MT1 receptor in the SCN, which mediates melatonin’s circadian action, has so far not been studied in this disorder. In the present study we aimed to investigate the alterations of the MT1 receptor in the SCN in depression using immunocytochemistry with an anti-MT1 receptor antibody. Moreover we evaluated the eﬀects of age at onset and disease duration on the MT1 receptor expression in the SCN of depressed patients.

Subjects and Methods Subjects 28 brain autopsy samples were used, of which 14 were patients clinically followed for mood disorders with an age range of 51–80 years, and 14 were control subjects matched for sex, age, clock time of death, seasonal distribution (according to the month of death, Summer: Jun-Aug; Winter: Dec-Feb), brain weight, postmortem delay and fixation time (Table 1). Brain material was obtained via the rapid autopsy system of the Netherlands Brain Bank (NBB) at the Netherlands Institute for Neuroscience (NIN), in accordance with the formal permissions for a brain autopsy, the use of human brain material and clinical information for research purposes. The DSM-IIIR/DSM-IV criteria were used for the diagnosis of major depression (MD) and bipolar disorder (BD) at any time during life. The criteria for the presence, duration and severity of symptoms of either MD or BD, as well as exclusion of other psychiatric and neurological diseases, were systematically scored by two qualified psychiatrists (Dr W.J.G. Hoogendijk or Dr E. Vermette). Seven patients fulfilled the criteria for MD and seven fulfilled the criteria for BD (Table 1). 4 MD patients and 6 BD patients suffered from their last episode just before death. In 2 BD patients, the last episode was a manic episode (Table 1). An overview of psychiatric medication in the past and in the last month before death for depressed patients and controls is given in Table 1. The medical records did not reveal any alcohol or other drug abuse among subjects of either group unless stated otherwise (C3). The absence of neuropathological changes, both in the patients with mood disorders and in the controls,

134 MT1 in the SCN in depression was conﬁrmed by systematic neuropathological investigation 177 by Dr W. Kamphorst (Free University Amsterdam, The Netherlands).

Histology Hypothalamic samples were fixed in formalin and embedded in paraffin. Serial coronal sections (6 μm) were made from the level of the lamina terminalis to the mamillary bodies. Depending on availability, either the left or right hemi-hypothalamus was used. For anatomical orientation every 100th section was collected and mounted on superfrost plus slides (Menzel GmbH & Co KG, Baunschweig, Germany) and subsequently dried for at least 2 days at 37 ºC, followed by Nissl staining (0.5 % thionine in distilled water). Additionally, immunocytochemical staining for vasopressin (AVP) and vasoactive intestinal polypeptide (VIP) was performed simultaneously on every 50th section taken along the rostrocaudal axis throughout the complete SCN region in order to delineate the SCN. The rostral and caudal border of the SCN was defined as the most rostral and most caudal section that contained one or more AVP or VIP positive cells in the area. The central cross-section containing the maximal SCN was defined as the central section containing the most AVP and VIP positive cells, to which adjacent central SCN sections were taken for the MT1 receptor staining.

AVP and VIP immunocytochemical staining The combined immunocytochemical staining for AVP and VIP followed the protocols as previously described with some minor modiﬁcations 23, 310. Deparaﬃnized and rehydrated sections were microwave-treated (for antigen retrieval) at 90 ºC for 15 min in TBS (0.05 M Tris, 0.15 M NaCL, pH 7.6). Sections were then incubated with rabbit anti-AVP antibody (Truus, 29-01-1986) at 1:800 and rabbit anti-VIP antibody (Viper, 18-09-86) at 1:1000 in Supermix [SUMI: 0.25 % gelatin (wt/vol), 0.5 % Triton X-100 (vol/vol) in TBS)] by one hour incubation (RT) followed by 4 ºC overnight. The next day after rinsing in TBS the sections were incubated with goat-anti-rabbit serum (Akon) at 1:100 in SUMI (1 h, RT). After rinsing in TBS the sections were incubated with peroxidase-antiperoxidase (PAP) at 1:1000 in SUMI (1 h, RT). Af- ter rinsing in TBS, visualization of the signal was accomplished by incubation with 3.3’-diaminobenzidine (DAB) (Sigma) 0.5 mg/ml in TBS, containing 0.23 % (wt/vol) nickelammoniumsulfate (Merck, Amsterdam, The Netherlands) and 0.01 % (vol/vol)

H2O2 (Merck) for approximately 20 min (RT). The reaction was stopped in distilled water, and slides were dehydrated in an ascending series of alcohol and xylene and coverslipped in Entellan (Merck).

135 chapter 7

yes yes yes yes yes yes manic during during depr. Died

PHT Li Li, BZD BZD BZD BZD med. taken (last month) (last PhT Li. Hal, MAOA-I none BZD, MAO-I, none LI, PHT Psychiatric Psychiatric Morphine, BZD Morphine, none morphine BZD,

Li, Hal, TCA, TCA, Li, Hal, TCA Li Li, BZD TCA, MAO-I, BZD, MAO-I, Psychiatric Psychiatric BZD BZD BZD BZD BZD BZD LI BZD medication taken (past) taken none none BZD, 8 3 1 8 1 4 2 Episode Clinicopathological information Respiratory insufficient, lung emphysema lung insufficient, Respiratory peritonitis Faecal hematoma Subdural head carcinoma Pancreas Pneumonia bronchopneumonia, carcinoma, Lung abuse) (alcohol infarction myocardial intestinal to ileus due arrest, Cardiac haemorrhage myocardial ileus, probably Probable infarction after dysrhythmia and Hypotension hypernatremia with intoxication lithium metastasis with cancer Severe prostate pneumonia ischemia, Cerebral deficiency, G6PD infarction, Myocardial aerophagia, insufficiency Respiratory thromboses mesenterial

28 61 30 43 35 32 43 36 48 45 38 31 26 55 Fix Fix (day)

1:00 4:20 4:50 8:35 5:00 5:55 8:30 75:00 71:00 41:00 10:20 16:00 13:30 20:00 PMD (h:min)

nd nd (g) 975 1390 1348 1424 1310 1424 1250 1490 1169 1475 1074 1109 Brain Brain weight weight

4 2 9 1 1 5 2 5 8 1 of of 11 10 10 11 death Month

0:59 4:40 5:00 2:45 2:55 14:00 17:30 16:45 12:00 11:30 16:15 16:00 22:30 21:00 death death Clock time of of time f f f f m m m m m m m m m m Sex

36 78 63 64 69 69 65 51/41 68/42 61/50 70/35 70/45 71/53 71/65 Age at at Age death/ at onset at Table 1 subjects Table control and patients depressed of material brain Postmortem No. D1b C1 D2b C2 D3b C3 D4b C4 D5b C5 D6b C6 D7 C7

136 MT1 in the SCN in depression

no no yes yes yes Yes manic during during depr. Died

Li, HAL, ZUC Li, HAL, SSRI Li, TCA TCA, MIA, BZD, Haldol, SSRI, Morphine ZUC BZD, SSRI, BZD BZD BZD med. taken (last month) (last none morphine BZD, Psychiatric Psychiatric none Hal, BZD Hal, Morphine none

MIA, TCA MIA, PHT, BRO, TCA, CAR BZD, SSRI none SSRI None Psychiatric Psychiatric MAP BZD BZD BZD medication taken (past) taken none none none none 3 4 4 1 4 3 1 Episode - - - - - Clinicopathological information Myocardial infarction Myocardial carcinoma Pancreas Pneumonia metas multiple with Mammacarcinoma pyelonephri septic shock, failure, Heart insuﬃciency, Septic respiratory shock, coma cystitis, pyelitis Bronchopneumonia, embo lung accident, vesicular Cerebral (pons) hemorrhage Brain ischemia Cardiac vascularCerebral accident abdo acute vascularCerebral accident, hanging by Suicide decompen cardial CARA, emphysema, tases in the cerebral meninx tases in the cerebral sation tis either secondarymen perforation to intestines or stomach lism,

39 56 35 28 69 35 54 33 43 38 35 34 Fix Fix 130 147 (day)

7:00 5:55 4:00 9:10 22:00 65:00 28:00 33:25 33:00 16:30 22:10 63:00 62:55 PMD (h:min)

nd (g) 1287 1116 1320 1300 1400 1427 1400 1204 1265 1123 1417 1444 1067 Brain Brain weight weight

1 1 4 7 7 6 2 1 1 6 3 5 of of 12 10 death Month

4:20 2:00 8:00 8:40 2:30 0:30 2:00 19:00 16:30 23:15 21:00 20:45 17:05 14:00 death death Clock time of of time f f f f f f f m m m m m m m Sex

72 65 83 39 65 64 72 72/54 72/53 80/60 45/32 68/32 75/40 74/74 Age at at Age death/ at onset at No. D8 C8 D9 C9 D10b C10 D11 C11 D12b C12 D13b C13 D14 C14 time; ﬁxation Fix, femal; f, patient; depressed D, carbamazepine; CAR, subject; control C, benzodiazepine; BZD, bromperidol; BRO, disorder;phenothiazine; bipolar b: PhT, data; no ND: mianserin; MIA, maprotiline; MAP, inhibitors; oxidase monoamine MAOA-I, male; m, lithium; LI, haloperidol; Hal, ZUC, zuclopenthixol; tricyclic TCA, antidepressants; inhibitors; reuptake PMD: postmortem delay; selective serotonin SSRI,

137 chapter 7

MT1 immunocytochemical staining Two consecutive central cross-sections containing the maximal SCN area from each subject were stained with polyclonal affinity-purified specific antibody anti-MT1 receptor (provided by Dr. Ralf Jockers, Institute Cochin, Department of Cell Biology, Paris, France). A polyclonal anti-MT1 receptor antibody was raised in a rabbit, directed against the C-terminus and affinity purified 378. This antibody detected the MT1 receptor as a protein with an apparent molecular weight of 60 KD in immunoblots after separation by SDS-PAGE. It also specifically precipitated the 2-[125I]iodomelatonin– labeled receptor from MT1-transfected HEK 293 cells 378, and has been used in human postmortem brain material before 511, 512. In our recent studies (Chapter 5), staining with omission of primary antiserum and preadsorption with the homologous peptide eliminated positive staining. Moreover, staining of western blots of the human hypothalamus with anti-MT1 receptor revealed a band of the expected molecular mass of 60 KD for the MT1 receptor protein. These data further supported the specificity of the antibody in the human hypothalamus. Deparaffinized and rehydrated sections were 1) pretreated with 3 μg/ml proteinase-K (Invitrogen, Breda, Netherlands) in Proteinase-K buffer (2mM CaCL2, 0.01M Tris/HCl pH 7.5) at 37 ºC for 15 min; 2) rinsed in glycine buffer (2 g glycine in 1000 ml PBS) for 30 seconds to stop the proteinase-K reaction; 3) pre-incubated with TBS-milk [3 % nonfat dry milk powder in TBS (pH 7.6), Elk, Campina] for 1h at RT; 4) incubated with anti-MT1 purified antibody 1:200 in Supermix-milk [SUMI-milk: 0.25 % gelatin (Merck, Amsterdam, The Netherlands) (wt/vol), 0.5 % Triton X-100 (Sigma) (vol/vol) in TBS-milk (pH 7.6)] 1 h at room temperature (RT), and overnight incubation at 4 C in a moist chamber; 5) washed once with TBS-milk and twice with TBS, followed by incubation in biotinylated goat-anti-rabbit IgG (Vector Laboratories Inc., Burlingame, CA) 1:400 in SUMI (1 h, RT), 6) washed with TBS and incubated with avidin biotin complex (ABC) (Vector Laboratories, Inc. Burlingame, CA) 1:800 in SUMI (1 h, RT) and rinsed in TBS twice; 7) incubated in biotinylated tyramide 1:500, 0.01 % H2O2 (Merke) in TBS at RT for 15 min, a signal amplification procedure described in detail by Adams 515; 8) rinsed in TBS twice and incubated in ABC complex 1:800 in SUMI (1h RT) and rinsed in TBS twice; 9) similar procedure of visualization was performed as above.

Quantitative analysis of MT1-positive cells in the SCN Measurements of the MT1-immunoreactive (MT1-ir) cells in the SCN were made using an image analysis system (ImagePro v4.5, Media Cybernetics, Silver Spring, USA) connected to a camera (JVC KY-F55 3CCD) and plain objective microscope (Zeiss Axioskop with Plan-NEOFLUAR Zeiss objectives). The microscope was equipped with planapo objects and a scanning stage. The area of the SCN was manually outlined

138 MT1 in the SCN in depression at low magnification (2.5 x objective) in an anti-AVP and anti-VIP staining section containing the largest SCN area. The outline of the SCN was transferred to the low magnification (2.5 x objective) image of the adjacent section stained with anti-MT1, and a grid of field was subsequently super-imposed. Then, at high magnification (40 x objective), each field was retrieved at high resolution on the image analysis monitor. The MT1-ir cell profile that presents with a nucleolus in the field was counted and stored in the analysis system. The MT1-ir cell number was determined in the central SCN of each subject. The number of MT1-ir cells per unit SCN volume was estimated as the density of MT1-ir cell in the SCN, with the section thickness (6 μm) taken into account.

Statistical Analysis Mann-Whitney U test (2-tailed) was used to test the (1) differences between the control and depressed groups; (2) differences between day and night groups, according to their time of death (1000-2200 day; 2200-1000 night), within the depression group or within the control group; (3) differences within the depressed patients according to their medication in the last month. The effects of postmortem delay (PMD), fixation time and duration of the depression on MT1 receptor data set was tested by using the Spearman correlation coefficient. Values of P<0.05 were considered to be significant. All values are expressed as mean ± standard error of the mean (Mean ± SEM).

Results The sex (P=0.44), age (P=0.29), day / night distribution (P=0.5), brain weight (P=0.78), post-mortem delay (P=0.48) and ﬁxation time (P=0.25) were well matched for the depression and control groups, which are presented in the Table 1. MT1-expressing cells were distributed widely in the human anterior hypothalamus. MT1 immunoreactivity (MT1-ir) appeared as granular cytoplasmic staining in the neurons, not in glia cells. MT1-ir was also present in nerve ﬁbers. MT1-ir neurons were not only present in the area of the SCN, but also in the paraventricular nucleus, supraoptic nucleus, periventricular nucleus, sexually dimorphic nucleus of the preoptic area, diagonal band of Broca, horizontal limb of the diagonal band of Broca, and nucleus basalis of Meynert in the anterior hypothalamus at median to strong level. Weak staining was observed in the lateral hypothalamic area. The distribution of immunocytochemical reaction product and the appearance of the immunoreactive neurons in patients with depression were similar to the results of the control cases. The intensity of neuronal MT1-ir was stronger in depressed patients than in controls (Fig. 1). The density of the MT1-ir neurons in the central SCN section was increased in depressed patients (38,390 ± 4,957 per mm3) compared to controls (21,009 ± 5,522

139 chapter 7

Fig. 1 MT1 receptor immunocytochemical staining in the SCN of (a) a depressed patient; (b) a control subject. Note that the MT1 receptor staining intensity in the SCN is higher in the depressed patients than in controls. Scale bar indicates 50 µm. per mm3) (P=0.004) (Fig. 2A). The total MT1-ir neurons number in the central SCN section was also significantly higher in depression (251.5 ± 41.4) than in controls (120.1 ± 21.1) (P=0.015) (Fig. 2B). No significant difference was found in the size of the area of the central SCN section between the depression and control group (1.07 ± 0.09 and 0.97 ± 0.07 mm2, respectively; P=0.75). In depressed patients, there was a positive correlation of the density and number of the MT1-ir neurons in the central SCN section with the duration of disease (r=0.63, P=0.016; r=0.57, P=0.034, respectively). Moreover, a negative correlation was observed between the age at onset and the density and number of the MT1-ir neurons in the central SCN section (r=-0.56, P=0.036; r=-o.632, P=0.015, respectively). There was no difference either in the number of MT1-ir neurons or in the density of MT1-ir neurons in the central SCN section between the 6 depressed subjects who had taken Lithium in the last month before death (D1, D4, D5, D7, D10, D12) and 8 other last month Lithium-free depressed patients (P=0.44, P=0.85; respectively). No difference in the number or the density of MT1-ir neurons was observed between 3 depressed patients who had taken selective serotonin reuptake inhibitors (SSRIs) in the last month before death (D11, D13, D14) and 11 other last month SSRIs-free depressed patients (P=0.31, P=0.75; respectively). In addition, the number and density of MT1-ir neurons in 2 patients who were treated with monoamine oxidase inhibitors (MAOIs) in the last month before death (D2, D7) did not differ from 12 other last month MAOIs-free depressed patients (P=0.36, P=0.21; respectively), nor did 6 patients who had taken benzodiazepines during the last month before death (D5, D7, D8, D9, D13, D14) differ from the 6 other last month benzodiazepines-free depressed patients (P=0.9, P=0.9; respectively). No differences in number and density of MT1-ir neurons were observed between 7 patients with major depression and 7 patients with

140 MT1 in the SCN in depression

MT1 cell number (per central SCN section ) MT1 neuron density in the central SCN

300 5 104

250 4 4 10

200 4 3 10 150 4 2 10 100

4 1 10 50

0 0 Control Depression Control Depression

Fig. 2 (a) MT1-immunoreactive (MT1-ir) neuron density and (b) MT1-ir neuron number per central SCN section in the central SCN of controls and patients with depression. Note that both are increased in depressed patients compared to controls *denotes p<0.05 compared to the other three groups. bipolar depression (P=0.41, P=0.25; respectively). The only subject (C3, control) who had alcohol abuse did not show different values compared to other control subjects. No significant correlation was observed either in the control group or in depressed group between postmortem delay and fixation time on the one hand, and the number and density of MT1-ir neuron in the central SCN section on the other (all P>0.05).

Discussion The present study demonstrates for the first time that the MT1 melatonin receptor expression in the SCN neurons is significantly increased in patients with depression. Moreover, the enhancement of the MT1 receptor expression in the SCN of depressed patients is related to the age of onset and disease duration. An auto-regulatory role of melatonin on its receptor expression in the SCN and pars tuberalis of the pituitary has been elucidated in animal experiments 400, 402, 405, 574. Melatonin receptor protein is strongly increased in the SCN by decreasing melatonin production induced by constant light exposure or pinealectomy, and decreased following a single melatonin injection 402, 403. Numerous studies have observed that patients with major depression, bipolar and seasonal depression have a reduced daily production of melatonin, and a decreased and phase-shifted nocturnal peak of melatonin levels 212, 213, 561-566. Our present observation of an increased MT1-ir in the SCN of depressed patients may, because of the animal experiments mentioned above, be considered as a result of an auto-regulatory effect of low-melatonin production in

141 chapter 7 depression. Interestingly, in the present study we also observed a significant correlation between age at onset and depression duration with the density and number of the MT1-ir neurons in the central SCN in depressed patients. These results strongly suggest an accumulative effect of depression process on the up-regulation of the MT1 receptor in the SCN in depression. The hyperactive hypothalamopituitary–adrenal (HPA) axis, a key system in stress responses, is considered to be the ‘final common pathway’ for a major part of the depressive symptomatology 575, 576. This hyperactive HPA-axis may exert an inhibitory effect on the melatonin production via the increased CRH secretion from the hypothalamic pavaventricular neurons in depression 577, 578. Age at onset and depression duration influence the duration of the hyperactivity of the HPA-axis. Long durative hyperactivity of the HPA axis may result in a continuing low melatonin production, which could enhance its auto-regulatory effects on the melatonin receptor expression in the SCN. Although we do not have melatonin levels from these patients, all of these subjects had histories of hospitalization, multiple episodes, indicating relatively more severe depression. Moreover, most of these studied depressed patients had strongly activated CRH expression in the PVN neurons in our previous studies 579-581. Taken together, these findings suggest that HPA axis dysreg- ultion in depression repeatedly give rise to decreased melatonin production, which may result in enhanced MT1 receptor expression in the SCN. This up-regulation of MT1 may reflect a compensatory adaptation of SCN neurons to the lower melatonin levels in the regulation of its circadian actions. This compensatory effect, however, is probably inadequate per se; however, it may considerably contribute to the effectiveness of administration melatonin and melatoninergic agonist in depression. In fact, exogenous melatonin has shown to alleviate the depressive mood, and in particular sleep disorders such as insomonia, early waken and short sleep length in depression 214, 547, 568-572. Our results thus provide further biological basis for the clinical implication of melatonin supplementation for depressed patients. In addition, another possible interpretation of our data should also be considered. A synaptic pathology has been demonstrated in the anterior cingulate cortex 582, hippocampus 583 and associated cortex 584 of depressed patients. Interestingly, we recently found a clear decrease of vasopressin (AVP) mRNA however an increased AVP-immunoreactivity (ir) in the SCN of depressed patients 210. As AVP is normally transported via axons to targeted areas, this phenomenon actually suggests a decreased axonal transport in the AVP neurons of the SCN of depressed patients. MT1 receptor is present both in the neurons and in the nerve fibers of the SCN, however, it is not clear yet if the MT1 is transported through the nerve fibers or just diffused in the fiber without actual functional transport. Previously we have observed a co-localization of MT1 receptor protein with AVP peptide in the neurons and some fibers of the SCN (lab observation). If the MT1 were transported like the AVP peptide in the same SCN

142 MT1 in the SCN in depression

fibers, a disrupted axonal transport activity in depression would affect both proteins; at least we would observe a correlation between these two proteins. In 9 out of the 14 depressed patients we studied in the present study, their AVP mRNA and AVP peptides have also been studied before by our group 210. We did not find a significant correlation between the density and number of MT1-expression neurons in the central SCN on one hand, the AVP mRNA amount and AVP-expression neuron number in the SCN on the other. Thus this result does not support the idea of a transport of the MT1 receptor in the nerve fibers in the SCN. So far whether the increased MT1 in the SCN is a results of decreased axonal transport activity in depression is still in argument. Clearly, further studies are needed to fully elucidate this issue. Clock activity and light / dark cycle have been implicated in the regulation of melatonin receptors in the SCN in mammals 403. As we strictly matched the clock time of death between depressed patients and control subjects, these factors thus did not affect our group comparisons. Diurnal rhythms of melatonin receptor binding in the SCN have been reported in rodents 402, 403, 406. We did not find a significant daily fluctuation of the MT1-ir neurons in the SCN either in controls or in depressed patients. As the number of cases in each group is rather small, and some confounding factors such as medication, antemortem and postmortem states etc (discussed below) might have contributed to the variations of our outcomes, it is not possible to draw a firm conclusion on this point. We attempted to investigate the possible effects of psychiatric medication on the MT1 receptor in the SCN. Many psychiatric medications have shown influences on the SCN. Lithium has shown to affect the SCN neuron firing directly and thus influence circadian rhythms 585, 586. GABA and benzodiazepines have shown to influence the SCN activity in rodents 587. The significant effects of a number of psychiatric drugs which affect monoamine function and thus melatonin production, including MAOIs 588, lithium, propranolol, tricyclic antidepressants amphetamine have also been suggested 589. Here we did not observe any significant effect of these compounds on either the density or the number of MT1-ir neurons in the central SCN section, which indicates that these medications were unlikely to influence our outcome. Postmortem delay and fixation time did not seem to affect the MT1 expression data set in our study. This indicates a considerable stability of the MT1 protein under different postmortem conditions, which is consistent with previous studies showing the melatonin binding in the SCN under prolonged postmortem conditions 522. In conclusion, the MT1 receptor in the SCN is increased in depressed patients. Moreover the enhancement of MT1 receptor is correlated to the age at onset and the disease duration. The up-regulation of the MT1 receptor may contribute to the efficacy of clinical administration of melatonin in depression.

143

Chapter 8

General Discussion

Content A. Methodological considerations 147

B. Comparison of melatonin synthesis between avian, rodent, ungulate and human: 154

C. Alterations in the circadian timing system in the process of Alzheimer’s disease and the 160

D. Clinical implication of the present studies 166

E. Future experiments 169

A. Methodological considerations

A comprehensive approach on human postmortem material In the present thesis we investigated the molecular mechanism of melatonin synthesis and the circadian profiles of clock gene expression in the human pineal gland and how these profiles change in the course of Alzheimer’s disease. For the first time, we used the approach to measure simultaneously the precursors (tryptophan, 5-HT, NA), products (melatonin, 5-HIAA, MHPG), and gene expression of important enzymes (TPH, AA-NAT1, HIOMT, MAOA, MAOB) and β1-adrenergic receptor in the melatonin synthesis pathway as well as its noradrenergic regulatory pathway in every postmortem human pineal gland (Chapter 2). The activity of enzymes is represented by the ratio of product to precursor (see Fig.1 melatonin synthesis and regulatory pathway). Moreover, gene expression levels of a set of clock genes were determined in each human pineal gland (Chapter 4). This approach combines methods of high performance liquid chromatography (HPLC), radioimmunoassay (RIA) and quantitative polymerase chain reaction (Q-PCR), and yields a systematic investigation of the molecular mechanisms underlying the circadian function of the human pineal gland and the alterations they undergo in Alzheimer’s disease. The circadian rhythm of melatonin production is generated within the hypothalamic suprachiasmatic nucleus (SCN), the circadian clock 590. Since we found that melatonin production, melatonin daily rhythm and pineal clock gene oscillation are disturbed already from the earliest Braak stages onwards in the AD process (Chapters 2 and 4), we hypothesized that the “master clock”, the SCN, would be affected already early in the AD process, and that this would be the reason for the changes observed in the pineal gland. Consequently, we performed in situ hybridization to investigate

147 chapter 8 the alterations on the level of mRNA of the main neuropeptide output of the SCN, vasopressin (AVP), in the human SCN during the AD process (Chapter 4). Since melatonin feeds back on the SCN via melatonin receptors (MT1 being the most expressed receptor subtype in the SCN), we also investigated the alterations in the MT1 receptor in the SCN during aging, depression and Alzheimer’s disease by immunocytochemistry (Chapters 6 and 7). These studies were all carried out on the human postmortem pineal gland and hypothalamus, making use of a variety of techniques and methods. The use of human postmortem tissue obtained from the Netherlands Brain Bank (NBB) and the collection of some 2000 serially sectioned human hypothalamus of Prof. D. F. Swaab enables the understanding of the neurobiological and neuropathological processes in health and disease on patient material instead of using models. While researches in

Indole synthesis from tryptophan in pineal gland

Tryptophan TH 5-hydroxytryptophan (5-HTP)

MAO 5-MT Serotonin (5-HT) 5-HIAA + NAT β-adrenergic receptor N-acetyl-5-hydroxytryptamine (NAS) NE MHPG + HIOMT Suprachiasmatic Nucleus Melatonin

Fig. 1 Melatonin biosynthesis and regulatory pathway in the mammalian pineal gland. The production and secretion of melatonin are controlled by the circadian pacemaker, the suprachiasmatic nucleus in the hypothalamus via a multi-synaptic pathway. At night, the SCN stimulates the release of the noradrenaline from the sympathetic nerve ending, and activates beta1-adrenergic receptors in the pinealocytes, which raises cyclic AMP and calcium concentrations and activates NAT and HIOMT, initiating the synthesis and release of melatonin. Abbreviations: 5-HIAA, 5-Hydroxyindole Acetic Acid; 5-HT, 5-hydroxytryptamine; 5-HTP, 5-Hydroxytryptophan; 5-MT, 5-methoxytryptamine; HIOMT, hydroxyindole O-methyltransferase; MAO, monoamine oxidase, MHPG, 3-methoxy-4-hydroxyphenylglycol ; NAT, arylalkylamine N-acetyltransferase; NE, noradrenaline; TH, tryptophan hydroxylase.

148 general discussion cell culture and animal models are useful for the study of symptoms and the building of theories, the postmortem diseased human brain should always be used to validate models since it contains the in vivo clues from the disease processes themselves.

The combination of animal experimental studies and the investigation of human postmortem material Based on observations in the human postmortem pineal gland and on the SCN of controls and AD subjects (Chapters 2 and 4), we hypothesized that a disruption of the SCN would be responsible for the alterations in melatonin synthesis and clock gene oscillation in the pineal gland in the AD process. However, the role of the SCN could only be studied in a correlative way on human postmortem materials. There- fore, we investigated the eﬀect of a loss of SCN control of the pineal molecular clock by comparison of pineal clock gene expression in the rat before and after disruption of the SCN-pineal functional connection, either by superior cervical ganglionectomy (SCG-X) or by lesioning the SCN (SCN-X) (Chapter 4). Our animal experiments thus directly showed the dominant eﬀect of the disruption of the SCN in the regulation of pineal circadian function. Combination of both animal studies and postmortem human researches strengthened our conclusions.

Confounding factors in human postmortem studies In our quantitative human postmortem studies, the abundance of mRNA, monoamines and immunoreactivity exhibit marked individual variation in our studies. Gen- erally, variation can have a biological background and reveal important functional information. However, variation can also be partially due to confounding factors. Previous studies suggest that there are two groups of possible confounding factors that are present in human postmortem material that do not complicate animal experiments, i.e. pre-mortem factors, such as age, gender, ethnicity, medicative history and agonal state, and postmortem factors, such as interval between death and autopsy, and storage time in the freezer or in paraﬃn (501, 591; for review see 461). However, one can deal mostly well with the factors by matching. These confounding factors did not aﬀect the results or conclusions of the present thesis. The confounding factors which may be involved in our studies are discussed below.

Postmortem factors Accumulative evidence demonstrates that postmortem brain tissue is suitable for in vitro metabolic, functional and gene expression studies (501, 592, 593; see review 461). The analysis of monoamines, their metabolites and receptor binding in postmortem brain tissue should account for time, temperature, the handling of frozen tissue, post-

149 chapter 8 mortem delay samples and storage parameters to interpret accurately the effects of neurological disorders on neurotransmitter dynamics 594, 595. RNA can routinely be extracted from frozen autopsy tissue and formalin-fixed, paraffin-embedded autopsy tissue. Although RNA extracted from frozen tissue undergoes some degradation when compared with unfrozen tissue 557, 596, the pool of post-mortem human brain mRNA remains essentially an ordered system, which allows effects due to pathology or gender to be isolated and tested for significance 597. Actually the overall stability of mRNA in postmortem human brain material was shown up to 72-96 hours despite age, gender and diagnosis 593, 598. This stability of RNA is dependent less on factors associated with tissue handling and postmortem-interval and more on factors such as agonal state 599. However, autopsy RNA can be especially vulnerable to degradation during purifica- tion 600, and repeated freezing and thawing are very deleterious 593, 601. Most enzymes are quite stable in autopsy tissue (see review 461). Compared with RNA, the influence of postmortem on protein degradation in human autoptic brain tissue seems to be of minor importance 461, 591. In our studies, we matched these confounding factors as well as possible. Although in some of the studies Alzheimer’s disease patients had shorter postmortem delay (PMD) compared to controls, no significant effect of PMD was observed on our data.

Pre-mortem factors Age Plasma melatonin levels show maximum levels around the ages of 3–7 years and decline subsequently with age, with a major decline occurring before puberty 158, 602- 604. A reduction in melatonin concentrations during aging has also been extensively reported for the pineal gland, in saliva, CSF and also in urine as 6-hydroxymelatonin (reviewed in 202, 203). An age-related decline in SCN neuron activity has been reported previously, e.g for AVP on both the protein and mRNA level in the human SCN 23, 173. Age is apparently an important confounding factor for our studies of the human pineal gland and the SCN, and we have therefore used age-matched control and case subjects in Chapters 2, 3, 4, 6 and 7.

Sex The SCN is involved in reproduction, sexual orientation and sleep (reviewed in 6). A sex difference was found in the shape of vasopressin subdivision of the human SCN 23, as well as in the number of VIP-containing neurons 310, 605. Although melatonin is also involved in reproduction 431, no significant difference in melatonin production between male and female was observed 299, 606. However, another pineal hormone, 5-methoxytryptophol, shows a significant sex difference in children 606. In order to

150 general discussion exclude any possible bias from gender, we used sex-matched control and case subjects in Chapters 2, 3, 4, 6 and 7. In addition, statistical analysis showed that sex is not a determinant factor for our results.

Clock time of death and season of death (circadian and circannual distribution) The SCN is considered to be the endogenous central biological clock. It exhibits circadian and circannual rhythms, receives environmental light stimulation and coordinates ubiquitous circadian rhythms in the body 6, 20, 53. Melatonin synthesis in the pineal gland is rhythmically controlled by the SCN 53. Circadian and seasonal fluctuation in melatonin levels are reported in human 202, 441, and they are chronobiological indicators for the SCN function. We presume that our subjects had lived under a 24 hour light/ dark cycle, just like daily life. Subjects who died at similar clock time and month may share the similar environmental light/dark condition. By matching month and clock time of death as well as possible for the postmortem studies on the pineal (Chapters 2 and 4) and the SCN (Chapters 4, 6 and 7), we were able to exclude the influence of circadian time and light. A caveat may be that the light intensity to which the subject was exposed at the time of death is for obvious reasons unknown and uncontrollable, which might have contributed to the intercase variations. In fact, we observed significant day/night fluctuations of melatonin production (Chapter 2), β1-adrenergic receptor (β1-ADR) mRNA (Chapter 2), and gene expression of a series of clock genes (hPer1, hCry1, hBmal1) (Chapter 4) in the pineal gland of control subjects. These observed daily profiles had disappeared in the pineal gland of preclinical and clinical AD subjects (Chapters 2 and 4), indicating pathological changes in the circadian timing system occurring from the earliest Braak stages onwards. These findings of rhythmicity, and its changes in a disease condition, validated our methods. Seasonal effects were not seen in our data. Apparently, clock time of death is a more dominant factor for the variations observed in our data from control subjects. In Chapter 4, we observed an early decrease of AVP mRNA levels during the process of AD. We propose that the circadian profile of AVP mRNA may also be disturbed in the early AD stages. However, due to the limited number of cases in each group (9 controls, 8 preclinical AD subjects, 12 clinical AD patients), investigation of a daily rhythm of AVP mRNA during the AD process was not feasible in this study, but this point definitely deserves further study with more matched cases. Some animal studies have shown that melatonin binding (mainly to the MT1 subtype) and MT1 receptor mRNA exhibited significant day/night fluctuations in the rodent SCN 401, 403, 406, 407. However, others did not find a significant daily rhythm of MT1 mRNA in the rat SCN 607. For the first time, we studied the diurnal profile of the MT1 receptor immunoreactivity in the SCN of human subjects. We did not observe a significant day/night difference

151 chapter 8 of MT1 immunoreactivity in the SCN of controls, AD patients or depressed patients (Chapters 6 and 7). Nevertheless, these data are limited in their sample size and de- ﬁnitive conclusions on the daily rhythm of MT1 in the human SCN cannot be drawn from them. Further studies are needed, with more control cases whose time of death are more homogenously distributed over the 24-hour day/night cycle.

Medication Antidepressants Numerous studies have reported that antidepressants affect the synthesis and release of melatonin (reviewed in 589). In human, plasma melatonin levels are 3-fold increased during chronic treatment with clorgyline, a highly selective monoamine oxidase type A (MAOA) inhibitor, while no change was found during MAOB inhibitor treatment 589. MAOA inhibitor is proposed to inhibit the oxidation of serotonin (the precursor of melatonin) and thus increases the availability of serotonin for melatonin synthesis. Ad- ministration of tricyclic antidepressants, such as desipramine, was reported to increase melatonin production in human 608-610. This effect has been attributed to desipramine’s inhibition of norepinephrine (NE) uptake, with a consequent greater availability of NE to simulate pinealocyte beta-adrenoceptors. Moreover, fluvoxamine, a selective serotonin re-uptake inhibitor (SSRI) with antidepressive properties, also increases melatonin production in healthy subjects 611. Another type of SSRI fluexetine, however, leads to measurable depletion of melatonin in human 612. Lithium administration to the rat was associated with a reduced pineal melatonin nocturnal peak 613 and delayed melatonin nocturnal peak 614. However, lithium did not show a significant effect on melatonin production in human subjects with depression 566. Lithium treatment can also influence circadian rhythms in the master clock, the SCN. It directly suppresses neuronal firing of the SCN and increases the circadian period of firing rate rhythms in a dose-dependent manner 585, 586. Moreover, lithium treatment of cells inhibits glycogen synthase kinase 3 (GSK3), which leads to rapid proteasomal degradation of Rev-erbalpha, a negative component of the circadian clock, and results in activation of clock gene Bmal1 615.

Psychiatric medication The typical neuroleptics haloperidol and chlorpromazine increase circulating melatonin concentrations in rats 616, 617. In human studies, no diﬀerence of CSF melatonin concentration was observed between schizophrenic patients with or without chronic administration of chlorpromazine (585 mg/day) 618. Another study reported no altered CSF melatonin but increased serum melatonin levels by chlorpromazine (100-800 mg/day), which was positively correlated with the drug dose 619, 620. Acute evening administration of GABAergic drugs such as diazepam and sodium valproate (VAL)

152 general discussion reduced nocturnal serum melatonin levels in healthy human subjects 621, 622. A recent study, however, showed that VAL had no effect on overall melatonin secretion or dim light melatonin onset 623. On the other hand it significantly decreased the sensitivity of melatonin production to light 623. In addition to the effects on melatonin production, haloperidol also influences mPer1 gene expression in mouse SCN. During the light period, haloperidol increases the amplitude of mPer1 mRNA rhythmicity in mouse SCN. At night it disturbs the rhythmic pattern of mPer1 mRNA 624. Haloperidol may thus cause phase shift effects of circadian rhythms in the SCN. Additionally, GABA and benzodiazepines have been shown to influence SCN activity in rodents 587. As most, if not all, human SCN neurons also contain GABA as neurotransmitters 625, GABAergic drugs might have effects on the human SCN, similar to rodents. This, however, needs further study.

β-adrenergic receptor agonist and antagonist Evening administration of a β-adrenergic receptor agonist (such as propranolol or isoproterenol) or an antagonist (atenolol or acebutolol) strongly affects pineal melatonin production, resulting in increased melatonin levels or decreased melatonin production, respectively, in human 494, 626. However, daytime administration of these medications did not have significant effects on melatonin synthesis in human 495, 496. There are two possible explanations for these differential day/night effects. In the first place, melatonin synthesis may have a clock-controlled response to noradrenergic stimulation in the human pineal gland, possibly regulated by pineal clock genes, as suggested by animal studies 485. The other possibility is that the expression of β1- adrenergic receptors in the pinealocytes shows a day/night rhythm, with lower levels during the day and higher levels at night 260. Therefore the sensitivity of β1-adrenergic receptor to its modulators may show a day/night pattern. In our postmortem human studies we have taken the use of above medications into account, and have tried to exclude these confounding factors as well as possible. In Chapter 7, we attempted to investigate the possible effects of some psychiatric medications, such as lithium, MAOIs, SSRIs and tricyclic antidepressants on the MT1 receptor staining in the SCN. We did not observe any significant effect of these compounds on the MT1 data, which indicates that these medications were unlikely to influence our conclusion.

153 chapter 8

B. Comparison of melatonin synthesis between avian, rodent, ungulate and human: the rhythm-generating enzyme AA-NAT and clock genes in the pineal gland

A feature of vertebrate physiology that has been remarkably preserved is a circadian rhythm in pineal melatonin production, with a high level during the day and a low level at night. This circadian rhythm is present in all vertebrate animals, no matter whether they are day or night active. Thus the moment of the melatonin peak is not linked to activity or lifestyle 252. Melatonin is mainly synthesized in the pineal gland from serotonin via two important enzymes, i.e. serotonin N-acetyltransferease (arylalkylamin N-acetyltransferase, AA-NAT, E.C. 2.3.1.87) and hydroxyindole-o-methyltransferase (HIOMT, E.C. 2.1.1.4) (Fig. 1, melatonin synthesis pathway). Melatonin synthesis is also observed in the retina, which may be involved in the local physiology 253, 482. AA-NAT activity is strongly increased at night in all species studied so far, and is, therefore called the ‘melatonin rhythm-generating enzyme’ 482. The rhythmicity of the AA-NAT activity in the pineal gland is conserved in different species, which is of central importance in the generation of the circadian rhythm of melatonin synthesis. The fundamental structure of AA-NAT protein, especially in the core element, which is presumed to be involved in substrate binding and catalysis, is conserved. For instance, the putative arylalkylamine binding domain, which probably explains the high substrate specificity of AA-NAT, is conserved 252. Moreover, conserved two putative PKA sites, which probably reflect conserved cyclic AMP regulatory mechanisms, are consistent with the importance of cyclic AMP regulation for AA-NAT 252. The magnitude of the nocturnal increase in AA-NAT activity ranges from 7- to 150-fold between the vertebrate species 252. Light exposure at night can cause a rapid decrease of AA-NAT activity, and thus melatonin production 482. On the contrary, HIOMT activity only displays a small day/night variation, which increases with 30% at night in the rat pineal 627, while no day/night difference was observed in the chicken pineal gland 628. Moreover, alterations in the HIOMT activity have little effect on the rhythm of melatonin synthesis 252. Due to the important role of AA-NAT in the regulation of melatonin synthesis, many studies on the mechanism for AANAT activity regulation have been performed, and some remarkable species differences in the molecular mechanism involved in regulation of this enzyme have been revealed. Recently, molecular components of circadian oscillator in vertebrates have been found to involve intracellular transcriptional/translational feedback loops with negative and positive limbs, including a set of clock genes 101, 124. A complete set of clock genes with oscillation has been found in the pineal gland and retina of vertebrates, such as in avian 629, 630, rodents 179, 424, 492, 631, ungulate and human (Chapter 3). In rodent

154 general discussion pineal gland, clock genes are proposed to be involved in the gated melatonin synthesis via regulation in the AA-NAT level 485. Here we discuss the diﬀerences in the molecular mechanisms involved in the regulation of AA-NAT and clock gene oscillation between several species, i.e. avian, rodent, ungulate and human.

AA-NAT In avian (e.g. chicken) The avian pineal gland and retina contain a circadian oscillator, a photoreceptor and melatonin-synthesizing machinery, and they respond directly to environmental lighting 632, 633. The activity of AA-NAT protein and AA-NAT mRNA shows a marked circadian rhythm in the pineal gland and retina in chicken 634. The regulatory mechanism of pineal AA-NAT rhythm in avian is diﬀerent from that in mammals. AA-NAT activity in the mammalian pineal gland is controlled by the sympathetic innervation from the endogenous biological clock, the SCN, while in chickens the endogenous oscillator for AA-NAT activity rhythm is located in the pineal gland itself, which is regulated primarily on the mRNA level of AA-NAT. AA-NAT activity is entrained by light via noncyclic AMP-dependent mechanisms at the protein level, without changes in mRNA levels 635. This posttranscriptional regulation may provide a way to rapidly modulate AA-NAT activity and melatonin production in avian.

In rodents (e.g. rat, mouse and hamster) The rat pineal is the most-studied vertebrate pineal gland. Rat pineal exhibits a substantial daily rhythm in AA-NAT activity and AA-NAT mRNA (a 150 times increase at night) 636, 637. AA-NAT activity is regulated at both transcriptional and posttranscriptional levels in the pineal gland in rodents 252, 454, 637. The rhythmicity of AA-NAT is controlled by the SCN via a well-established multi-synaptic noradrenergic innervation (including the superior cervical ganglion: SCG). The nerve ending gets stimulatory signals from the SCN and releases noradrenalin (NE). NE stimulates both α1- and β1-adrenergic receptors and results in a 100-fold increase in intracellular cAMP levels 252, 636, 638. The rise in cAMP, in turn, stimulates cAMP-dependent protein kinase, which translocates to the nucleus and phosphorylates the transcription factor cAMP response element binding protein (CREB) 637. Phosphorylated CREB binds to cAMP response elements in the promoter region of cAMP-regulated genes (including AA-NAT) and causes activation of transcription 639, 640. This CREB activation of transcription results in a 150-fold increase in the mRNA of AA-NAT 637, and is one of the fastest cellular responses in pinealocytes upon adrenergic stimulation; within minutes 639, 641, 484, 642, 643. In addition to the activator CREB, the transcriptional regulation of the AA-NAT involves another norepinephrine-inducible transcription factors, i.e. the inhibitor

155 chapter 8

ICER (inducible cAMP early repressor) (see review 644). In contrast to CREB, ICER accumulates only after several hours, a time gap resulting from the required de novo protein synthesis upon adrenergic stimulation 644-646. The increase of inhibitory transcription factor ICER ultimately leads to the termination of mRNA accumulation of cAMP-inducible genes, including the gene for the AA-NAT that controls melatonin production. Besides through the transcriptional regulation, adrenergic-cAMP regulation of AA- NAT activity is also mediated by rapid reversible control of selective proteasomal proteolysis of AA-NAT protein 454, 647. Such inhibition of AA-NAT activity by proteasomal proteolysis may be conserved in vertebrates 454. This receptor-regulated proteasomal proteolysis can function as a precise, selective, and very rapid neural switch in response to environmental input. In addition, a variety of neurotransmitters, neuromodulators and post-translational mechanisms are also involved in the regulation of AA-NAT in rodents 648-650. These alternate signaling routes may be important in acute “emergency” situations. Together, combined signaling events in the rodent pineal gland help to generate a stable and reliable hormonal message of darkness for the body. This stability, however, can be rapidly altered upon sudden and unexpected “error” signals.

In ungulate (e.g. sheep, bovine) Sheep AA-NAT activity exhibits a 7-fold rhythm, which is relatively small compared to the rat. In contrast, the nocturnal increase of melatonin and AA-NAT activity in sheep is accompanied by less than a 2-fold increase in AA-NAT mRNA expression 651, 652. The AA-NAT activity in ungulate animals is regulated by cAMP through a mechanism that requires de novo synthesis of protein but not of mRNA 454. The rates at which AA-NAT activity and protein disappear following unexpected light exposure at night are essentially identical. In primary cultures of ovine pinealocytes constitutively high AA-NAT levels are not further increased by stimulation of cAMP by forskolin 653. Moreover, a rapid reverse control of a selective proteasomal degradation pathway may also contribute to the regulation of AA-NAT enzyme activity. Evidence from bovine pinealocytes suggests that the nocturnal increase of AA-NAT activity and melatonin synthesis are due to a noradrenalin-induced inhibition of proteasomal degradation of this enzyme 654, 655. These data suggest a predominance of post-transcriptional regulation in the ovine pineal and explain the rapid onset of AA-NAT activity and melatonin synthesis following darkness at any circadian time in ruminants 656, 657. Moreover, the high level of AA-NAT mRNA in the daytime ovine pineal gland may explain why sheep melatonin levels increase quickly following darkness in the night period and remain elevated for relatively long periods of the night 250, as the availability of mRNA at the start of the night period eliminates the lag time required for synthesis of AA- NAT protein.

156 general discussion

In human and rhesus monkey Knowledge of the mechanisms of melatonin synthesis in primate, including man, is sparse. In rhesus monkey Macaca mulatta, AA-NAT activity shows a 30-fold day/night difference but only a 3-fold day/night difference in AA-NAT mRNA 253. This indicates that, in contrast to rodent, the monkey and sheep may share a similar regulatory mechanism for AA-NAT, i.e. the transcription does not play a dominant role. After lights off, the circulating melatonin levels are also increased immediately in monkey 658, indicating a rapid synthesis of AA-NAT protein on the base of a presence of mRNA at the beginning of the night period. In human, a circadian profile of melatonin levels in CSF, salivary, plasma, or urine samples is reported (see review 202, 203), and it is truly of a circadian nature as it is present even in the absence of a light-dark cycle 590. A diurnal rhythm in AANAT and HIOMT enzyme activities has been shown in postmortem human pineal gland, with the highest levels between 22:00 and 4:00 and the lowest level between 12:00 and 18:00, in accordance with the diurnal profile of the circulating melatonin levels in human 659. Human pineal melatonin synthesis and regulatory pathways were studied for the first time in Chapter 2. Moreover, as melatonin levels are dramatically decreased during aging and even more pronounced in Alzheimer’s disease (AD), actually from the earliest AD neuropathological changes (Braak stage I-II) onwards, we investigated the molecular mechanisms underlying the reduced melatonin during aging and AD by assessing the pineal melatonin synthetic and regulatory pathway in aged control, early AD (Braak stage I-II) and late clinical AD patients (Braak stages V-VI). A significant day/night fluctuation of pineal melatonin levels was observed in human controls, with levels about 5 times higher at night than during the day. β1-adrenergic receptor (β1-ADR) mRNA showed a significant daily fluctuation in controls, with higher level at night than during the day, similar to the daily pattern of melatonin production. Moreover, there was a significant positive correlation between melatonin production and β1-ADR mRNA. As β1-ADR acts as key factor in this cAMP signaling pathway and is directly linked to the noradrenergic stimulus from the SCN 459, 491, these data support a similar regulatory cAMP signaling pathway of pineal melatonin synthesis both in the human and in rodents. However, we did not observe significant daily changes of AA-NAT mRNA, which was constantly highly expressed during 24 hours in controls. This is apparently very different from the rodents, who have dramatic circadian changes in AA-NAT mRNA, with changes that are 150-fold higher at night than during the day 637). Our results suggest that AA-NAT activity is not regulated at the transcriptional levels in the human pineal gland, which is similar to the pineal gland of monkey and sheep. In early and late stage AD patient groups such a daily fluctuation of melatonin was not observed, nor was the daily rhythm of β1-ADR mRNA, accompanied by a dramatic decrease of melatonin levels. No significant changes in

157 chapter 8

AA-NAT mRNA were observed between AD and controls, which further supports the notion that transcriptional control is not dominant for the regulation of AA-NAT activity, and thus melatonin synthesis, in man. In support of this, there is also physiological evidence. Many aspects of the pineal physiology of human are similar to those in monkey and sheep, but diﬀerent to those in rodent. For instance, the rapid increase of melatonin synthesis after lights oﬀ at night, and rapid decrease following exposure to bright light at darkness are seen in human 660-662, monkey 658 and sheep 590, 663. However, in rodents, there is always a delay between the dark or light exposure and the changes in melatonin production 637. Taken together, human AA-NAT activity seems to be regulated primarily at the posttranscriptional level, as is the case in sheep and the rhesus monkey.

Pineal clock genes In avian It is established that avian pineal gland and retina contain an endogenous self-sustained circadian clock 664, which directly links to the melatonin output signals 665-667. Moreover, these endogenous clocks also control the oscillations of clock gene expression in the pineal gland and retina 668. Interestingly, in vitro studies show that chicken clock genes BMAL1/CLOCK heterodimers can bind to the E-box DNA element in the chicken AA-NAT gene to activate the transcription of AA-NAT, and leads to enhanced melatonin production in chicken pineal gland and retina 666. This suggests that there is a functional molecular link between the synthesis of pineal melatonin and clock function in the avian retina and pineal gland.

In rodent In rodents, oscillation of clock genes on both mRNA level and protein level has been described extensively in the pineal gland 423, 424, 484, 631, 669 as well as in the retina 670. The mammalian retina has a self-sustained circadian oscillator. It can respond to light and it can oscillate based on a negative feedback between transcription and translation of clock genes, and control a variety of physiological and behavioral rhythms that often includes rhythmic melatonin production 492, 671, 672. A functional E-box element is identified in the first intron of the rat AA-NAT gene. In the rat retina photoreceptor cells, clock genes BMAL1/CLOCK heterodimers bind to this E-box and activate the transcription of AA-NAT in retina 673. The retinal circadian clock also gates melatonin synthesis in photoreceptor cells via E-box-mediated transcriptional activation of the AC1 gene, which undergoes robust daily fluctuations, that persists under constant illumination 492. In contrast to rodent retina, the rodent pineal gland does not contain a self-sustained circadian oscillator. Expression of clock gene Per1 and Cry2 in the pineal gland of

158 general discussion rodents is controlled by the clock-driven changes in norepinephrine via a cyclic AMP pathway, in a similar manner to the melatonin rhythm-generating enzyme AA-NAT 423, 424, 484. This noradrenergic control of the Per1 oscillation in rat pineal is also confirmed in our study (Chapter 4). Furthermore, we observed a highly positive correlation between Per1 and β1-ADR mRNA in rat pineal, that was lost in the SCN or SCG lesioned rat pineal (both lacking the noradrenergic SCN control of the pineal gland). β1-ADR mRNA acts as a key factor in this cAMP signaling pathway and is directly linked to the noradrenergic stimulation from the SCN 459, 491. The correlation between Per1 and β1-ADR mRNA may thus reflect the SCN regulation of Per1. The expression of clock gene Per2, Per3 and Cry1 displays a daily rhythm not regulated by norepinephrine, suggesting the involvement of another day/night regulated transmitter(s) 423, 424. So far, the functional role of the clock genes in the rodent pineal gland is not yet clear. The E-box-mediated AA-NAT gene expression by clock genes, which is shown in the rat retina 673, is silent in the rat pinealocytes 673. There is no evidence so far for a direct molecular link between clock genes and melatonin synthesis in the rat pineal gland. It is shown that AA-NAT mRNA levels in the pineal gland are increased after adrenergic stimulation only at certain times of day, thereby controlling the gating of melatonin synthesis 485. A potential role of the circadian clock in this gating response of AA-NAT gene expression in mouse pineal gland is suggested 485. However, this still needs to be confirmed.

In ungulate So far, only one study is available on pineal clock genes in ungulate animal, e.g. sheep 490. Johnston et al. reported that, similar to AA-NAT, clock gene Per1 and Cry1 undergo little or no temporal variation over 24-h in the sheep pineal. Moreover, changes in photoperiod modulate Per1 expression, while no corresponding changes were found in AA-NAT or Cry1 490. Compared to rodent pineal, the transcript control seems to be decreased in ovine pineal physiology. More studies are needed on these issues.

In human For obvious reasons, studies on clock genes in human tissues are scarce. Few studies reported that clock gene expression is widely-distributed in the human brain 137. How- ever, an analysis of rhythmic expression has so far only been reported in peripheral tissues such as oral mucosa, skin and peripheral blood mononuclear cells 139, 140. For the first time, we investigated the circadian profile of clock gene expression in the postmortem human pineal gland (Chapter 4). Clock genes hBmal1, hCry1 and hPer1 were rhythmically expressed in the pineal of controls (i.e. subjects devoid of any neurological or psychiatric diseases, and without AD neuropathology: Braak 0), while hClock did not show a significant daily rhythm. Compared to rodent pineal

159 chapter 8 gland, the amplitude of the diurnal rhythm of clock genes in the human pineal gland is much smaller. Intriguingly, in the pineal gland of human control subjects, hPer1 and hβ1-ADR mRNA showed a similar daily expression pattern and they were also positively correlated, just as observed in rat pineal gland. Moreover, functionally active cAMP-responsive elements (CREs) have been found in the promoter region of the hPer1 gene of human 135. It therefore seems likely that the clock gene hPer1 in the human pineal gland is also controlled by the SCN via the β-adrenergic cAMP-signaling pathway. Thus, by activating hPer1, the sympathetic input could potentially trigger the clockwork oscillation in the human pineal gland.

Conclusions In the species discussed above, the rhythmicity of AA-NAT activity in the pineal gland is preserved, which is of central importance in the generation of the circadian rhythm of melatonin synthesis. However, the molecular regulatory mechanisms of AA-NAT are highly diverse among species. For example, AA-NAT activity in rodent pineal gland is regulated on transcriptional as well as post-transcriptional levels. However, in sheep and primate, including human, AA-NAT activity may be dependent on posttranscriptional regulation. Furthermore, studies of the molecular clock in diﬀerent species have indicated that the basic mechanisms underlying their rhythmic gene expression have been evolu- tionarily preserved 499. In this respect, the regulation of clock genes (Bmal1, Clock, Per, Cry) is a diﬀerent story from AA-NAT. However, the studies on pineal clock genes are not as thorough as AA-NAT. In particular, what still needs further investigation is how clock genes are regulated on the protein level.

C. Alterations of the circadian timing system in the process of Alzheimer’s disease and the possible underlying mechanisms

Alterations of the circadian timing system from the earliest AD stages onwards The suprachiasmatic nucleus (SCN) in the anterior hypothalamus is considered to be the biological clock of the brain, which coordinates circadian rhythms throughout the body 53. The rhythm of the pineal hormone melatonin is a reliable output signal of the SCN through a multiple-synaptic pathway 251. Sleep-wake rhythm disruptions and other circadian disturbances are commonly seen in Alzheimer disease (AD) patients 148, 160, 162, 674, which is one of the primary causes of institutionalization of AD patients 675, in fact more often than cognitive impairment 165. The circadian disturbances in AD are accompanied by decreased melatonin levels

160 general discussion and a disrupted circadian melatonin rhythm (reviewed in 202, 254). Strikingly, our recent postmortem studies show that CSF melatonin levels and pineal melatonin content are already decreased from the earliest AD neuropathological stages (Braak stage I-II) onwards 254, 302(Chapter 2). This may indicate a very early affected circadian timing function in the progression of AD. In order to determine the underlying mechanism for the decreased melatonin production in the AD process, we analyzed the melatonin synthesis and regulatory pathway in the pineal gland in different neuropathological stages of AD 254(Chapter 2). We found that the rhythmic adrenergic regulation of the melatonin synthesis pathway, that is controlled by the SCN, is disrupted from Braak stage I-II onwards, which may be responsible for the decreased melatonin levels and the absent melatonin daily rhythm. Moreover, it points to a possible early-affected SCN regulation for the pineal gland in AD progression. Additionally, an up-regulation of MAOA was found in the pineal gland of AD (Chapter 2), which might deplete 5-HT (the precursor of melatonin) and reduce melatonin production. We proposed two possible mechanisms for the increased MAOA in AD pineal: 1) because the pineal melatonin synthesis is decreased in AD, due to a disrupted SCN stimulation, it may result in an accumulation of 5-HT (the precursor of melatonin) and subsequently lead to an increase of MAOA to oxidize 5-HT. 2) it may be a reflection of the AD process in the pineal gland, just like in some other brain regions, such as the prefontal cortex 308. To test the first possibility, we deprived the SCN innervation of the rat pineal gland by SCN-lesion or superior ganglion-ectomy (SCG- ectomy), and investigated the changes in MAOA mRNA levels compared to intact rat pineal. No significant changes of MAOA mRNA were observed in the denervated rat pineal (Fig. 2 MAOA mRNA levels in SCN-X and SCG-X rats), which indicates that the up-regulation of MAOA is not related to the disrupted SCN control in the pineal. It is therefore likely that the increased MAOA expression in the AD pineal may reflect a general MAOA disturbance in AD. A promoter polymorphism of a variable number tandem repeat (VNTR) in the MAOA gene has been shown to affect MAOA transcriptional activity 421, 422, 470. We examined the effects of the MAOA-VNTR genotype on MAOA gene expression, activity, and melatonin production as endophenotypes in the pineal gland of controls and AD patients (Chapter 3). We found that MAOA- VNTR genotype indeed affected MAOA expression in AD, i.e. the long-genotyped AD patients showed higher MAOA activity than the short-genotyped AD patients. However, the long-genotyped AD patients with increased MAOA activity did not show higher pineal melatonin levels compared to short-genotyped AD patients. Our data seem to suggest that the up-regulation of MAOA might not be a major factor for the reduction of melatonin production in AD. In other words, the dysfunction of the SCN control may be more dominant for the melatonin changes in AD.

161 chapter 8

1.2 ZT 8 ls ZT18 ve 1.0 le on 0.8 ssi e pr 0.6 ex ne 0.4 ge OA 0.2 MA

0.0 CON SCN-X SCG-X Fig. 2 Pineal MAOA gene expression levels in control rat, SCN-lesioned (SCN-X) and SCG- lesioned (SCG-X) rats. No diﬀerence between the bars. white bar indicates ZT8 (day) and gray bar represents ZT18 (night).

Previous studies on rodents have shown that the SCN does not only regulate melatonin synthesis, but also clock gene oscillation in the pineal gland through the same β-adrenergic cAMP signaling pathway 108, 423, 424, 484. We found that the diurnal rhythmic expression of hPer1, hCry1, and hBmal1 was lost in both early and late AD stages, which suggests that pineal clock gene oscillation is disrupted very early on in the AD process. Moreover, the positive correlation between hPer1 and hβ1-ADR mRNA, which probably indicates SCN control of hPer1 (as we discussed in the general discussion B, section pineal clock genes), is disturbed in both early and late AD stages (Chapter 4). Our hypothesis, that a disrupted SCN control is responsible for pineal clock gene alteration during AD, is strongly supported by the animal experimental data (Chapter 4). The rat pineal that was deprived of SCN control showed alterations of clock gene expression that were remarkably similar to those we observed in the AD pineal. This goes for the loss of rhythmic clock gene expression, the loss of correlation between Per1 and β1-ADR mRNA, and in particular for the increased Cry1. Our next aim was to study whether the SCN was aﬀected from the early AD stages onwards. Vasopressin (AVP) is a major rhythmic neuropeptide output of the SCN clockwork and regulates the rhythm of activity within the SCN and in other brain regions 94-96, 488. The AVP expression 92 is strictly controlled by the molecular clock in the SCN, as its rhythm had disappeared and its mRNA levels were dramatically decreased in the SCN of Clock-mutated mice 93. By assessing the alterations of AVP mRNA levels in the SCN, we thus also investigated whether the clock function of the SCN is disturbed early on in the AD process (Chapter 4). Remarkably, we found that the AVP mRNA levels in the SCN decreased from the earliest AD neuropathological stages onwards, which suggests that the SCN may have a diminished output and a disrupted clock

162 general discussion function early on in the AD process. Moreover, it supports the possibility that the SCN control of the pineal gland is disturbed very early on in the AD process. A functional disconnection between the SCN and the pineal from the earliest AD stage onwards could therefore account for the pineal clock gene and melatonin changes and underlie the circadian rhythm disturbances in AD. Our series of studies thus suggest that a decreased activity of the SCN, that is already present at the moment of the occurrence of the very ﬁrst tangles in the transentorhinal cortex (i.e. Braak stage I), most probably aﬀects pineal molecular clock synchronization and pineal rhythmic melatonin synthesis, and contributes to the circadian disturbances in the course of AD.

Possible mechanisms for the early-affected SCN in the progression of AD Indeed, neuropathological changes have been found in the SCN in Alzheimer's disease. Typical cytoskeletal alterations are present as pretangles 175, 177 and tangles 174 in the SCN of AD patients. Only diffuse amyloid plaques but no neuritic plaques were noted in this nucleus 174, 177. The loss of circadian rhythmicity may in part reflect a toxic effect of amyloid beta peptide (Aβ) on the SCN, as hypothalamic grafts of cells that produce Aβ disrupt circadian rhythmicity at the system level in wild type rats 676. It is possible that such neuropathological damage may already appear in Braak stages I-II during the progression of AD, which may be an underlying anatomical substrate for the functional disturbances in the earliest AD stages observed in our studies. A study on the SCN investigating this possibility is, thus, warranted. In addition to the occurrence of degeneration of the SCN in AD, the input of the SCN may be attenuated during aging and even more so in AD, which may contribute to a major degree to the de-activation of the SCN (see review 6). Of the inputs received by the SCN, the pathway of environmental light has been most comprehensively described and is considered to be crucial for the entrainment of the SCN. Environmental light- dark information is received by the retina and reaches the SCN via a direct projection through the retinohypothalamic tract (RHT). Several factors have shown to attenuate this projection to the circadian timing system during aging, and these are even more pronounced in AD. First of all, elderly people and, even more so, AD patients expose themselves to significantly less environmental light than younger prople 162, 186, which is related to more night-time awakenings and more daytime sleepiness 187. Secondly, light transmission through the lens, especially for short wavelengths, declines with aging 677, 678. Cataract and maculopathy are more common in the elderly 679, 680, and age-related maculopathy is associated with AD 192. Thirdly, the retina and the optic nerve, which provide direct and indirect light input to the SCN, show degenerative changes in AD 193-197. A neuronal loss of about 36 % of the entire retina was observed in AD, with a significant increase of the ratio of astrocytes to neurons 197, indicating

163 chapter 8 that a process of neurodegeneration was going on. No neurofibrillary tangles, neuritic plaques or amyloid angiopathy were observed in the retina or optic nerves 196, 197, 681. Moreover, visual field defects and/or optic disc cupping compatible with the diagnosis “glaucoma” were found 5 times more frequently in Alzheimer patients than in controls, which may also have implications for the input to the circadian system 198, 199. Thus the light input to the SCN is probably hampered during AD. Whether such alterations in the RHT have already taken place from the earliest AD stages onwards and contribute to functional disturbances of the SCN, needs further studies. Especially the specific RHT tract peptide PACAP cells, fibers, and PACAP receptors which play an important role mediating the light input to the SCN 42, 237, should get attention in this respect. In addition to the photic input system, the SCN also receives input from other hypothalamic nuclei and other brain areas in human, including the raphe nucleus, locus coeruleus, septum and limbic forebrain 19, 38, 41 (Fig. 3, SCN innervation). Although these input pathways have a much less pronounced effect on the circadian rhythms than light 682, they are considered to play an important role in the entrainment of the circadian timing system by stimuli other than light. For instance, the Neuropeptide Y (NPY) input from the intergeniculate leaflet is of crucial importance for the effects of the activity level on the circadian timing system in rodents 683, 684. However, in human, since the SCN has a rather sparce plexus of very fine NPY axons, and the SCN itself contains a large number of NPY neurons, it is not clear whether the intergeniculate leaflet neurons indeed project to the human SCN or whether this projection is very much reduced or even absent in human beings 73, 74. As long-term fitness training improves the circadian rest–activity rhythm in elderly people, whether it is at least partially due to the activation of this pathway deserve further studies 685. The raphe nucleus is one of the major nuclei producing serotonin, which is important for the regulation of circadian rhythms, such as the circadian rhythmicity of ACTH release, food intake, drinking and the sleep–wakefulness cycle 686. Loss of serotoninergic neurons and serotonin transport in the raphe nucleus have been reported in AD patients 687-689, and this loss might be involved in circadian rhythm disorders in AD. Whether these input pathways are affected already in the early stage of the AD process needs further investigation. Furthermore, the SCN also receives cholinergic afferents from both the basal forebrain and brain stem which may mediate cholinergic effects on circadian rhythms and could play a role in the regulating of sleep and EEG 690. Loss, c.q. atrophy, of acetylcholine (ACH)-containing neurons in the basal forebrain is striking in AD 691, 692. Wisor et al. reported that the Tg2576 AD mouse model, which exhibits age- dependent amyloid beta (Abeta) deposition in the brain, showed sleep abnormalities akin to those in AD patients 693. Interestingly, these sleep abnormalities in Tg2576 AD mouse model may be in part due to the cholinergic deficit in these mice 693. All of this suggests a role of cholinergic abnormalities in the pathophysiology of sleep disorders

164 general discussion

Temperature

Pineal Physical activity melatonin IGL

GHT

RGT GABA, NPY RGC RHT Bright light SCN Glu

ACH Forebrain 5-HT 5-HT NA brain stem

Septum DR/MR LC Somatosensory input

Fig. 3 Schematic and simplified overview of the inputs to the suprachiasmatic nucleus, and their interactions. Inputs are in outlined font, structures in bold, tracts in normal font and neurotransmitters and hormones in italics. Abbreviations: 5-HT, 5-hydroxytryptamine (serotonin); ACH: acetylcholine; DR, dorsal raphe nucleus; GABA, y-aminobutyric acid; GHT, geniculohypothalamic tract; Glu, glutamate; IGL, intergeniculate leaflet; LC, locus coeruleus; MR, median raphe nucleus; NA, noradrenaline; NBM, nucleus basalis of Meynert; NPY, neuropptide Y, RGC, retinal ganglion cells; RGT, retinogeniculate tract; RHT, retinohy- potbalamic tract; SCN, suprachiasmatic nucleus; SHT, spinohypothalamic tract (From Van Someren, 1997). in AD, in addition to its well-studied role in AD-related cognitive deficiency. Moreo- ver, the cholinergic nucleus basalis neurons show neurofibillary tangles (detected by thioflavin-S histofluorescence) and pre-tangle tauopathy (identified by AT8, Alz-50) already very early in the course of the continuum that leads from advanced age to mild cognitive impariment (MCI) and AD 694. Our group recently found an increased neuronal metabolic activity in nucleus basalis of meynert (NBM) in cognitive intact subjects in Braak stages I-II compared to those without AD neuropathology (Braak 0), reflected by an increased Golgi apparatus (GA) size of the neurons 695. We hypothesize that this may reflect a compensatory mechanism related to the early neuropathology in this nucleus during AD process. Taken together, whether the cholinergic input pathway of the SCN is degenerated in the early stages of AD and how it contributes to the functional changes of the SCN certainly deserve further study.

165 chapter 8

D. Clinical implications of the present thesis

Decreased melatonin levels: an early sign of Alzheimer’s disease Previous studies from our group have shown that melatonin levels in the cerebrospinal fluid (CSF) are decreased in Alzheimer’s disease, from the earliest Braak stages onwards 302. These observations suggest that decreased CSF melatonin levels may be an early event in the development of AD. However, a possible confounder for these observations is a possible increase in CSF volume or an increased CSF turnover between Braak stage 0 and I-II. There is indeed an age-related enlargement of the ventricular CSF volume in AD 696, which is even more pronounced in late stages of AD patients 453, 697. Fortunately this confounder could be ruled out by our postmortem pineal studies (Chapter 2), as we showed that melatonin levels in CSF were highly correlated with the pineal melatonin content in aged subjects and AD patients 254. Moreover, the pineal melatonin content was significantly decreased in Braak stages I-II and VI, compared to Braak stage 0 254, which confirmed our group’s previous observations in the CSF 302. These results strongly indicate that decreased CSF melatonin levels are due to decreased pineal melatonin production, already from the earliest AD stages onwards. There is also a positive correlation between CSF melatonin levels and serum melatonin levels in healthy human subjects 303. A decrease of melatonin levels in CSF, and probably also in serum, saliva and urine, may be thus considered as an indication of the very first stage of the development of AD, before any cognitive symptoms appear, which cannot be monitored in vivo by any other means. Longitudinal follow-up studies on the salivary or serum melatonin levels in aged controls until the moment the first cognitive symptoms occur have to be performed in order to confirm our hypothesis.

Reactivation of the SCN from the earliest stages onwards in Alzheimer’s disease The circadian timing system, and particularly the SCN, the circadian peacemaker in the brain, is aﬀected during aging and strongly degenerated during Alzheimer’s disease. Our series of studies in this thesis show that the “master clock”, the SCN, is disrupted from the earliest AD stages onwards, which may result in a de-synchronization of circadian oscillations of peripheral clocks, and may be responsible for circadian rhythm disturbances during AD. Many ﬁndings have indicated that a weak expression of circadian rhythms is associated with poor functioning, for which elderly are at risk, such as sleep complaints, cardiovascular problems 698, 699, hypo- and hyperthermia 700, depression 205, 567 and cognitive decline 701, 702. Disturbed circadian rhythms such as the sleep-wake pattern are highly correlated with cognitive decline and functional impairment, particularly in the progression of Alzheimer’s disease 164, and the development of

166 general discussion eﬀective ways to restore the circadian rhythms from the earliest stages of AD onwards is thus of great importance. Our group has proposed a “use it or lose it” theory on the plasticity paradigm in the central nervous system 703, 704, which suggests that (re-)activation of neuronal systems may protect against age-related neurodegenerative changes. Stimulants for the SCN, that area of potential interest in this respect, include bright light, melatonin, temperature, somatosensory input and physical activity.

Light therapy Light is the primary Zeitgeber for the central nervous system. Via the retinohypothalamic tract, bright light reaches the “core” neurons in the SCN and membrane de- polarizes the SCN neurons 53. Increased light intensity enhances the SCN neuronal firing rates 705 and prevents the age-related loss of vasopressin-expressing neurons in the rat suprachiasmatic nucleus 706. Since the light input is affected in aged controls and even more so in AD patients (see review 145), who tend to be exposed to significantly less environmental light than young people 186, a bright light supplementation is considered to be a potential treatment for circadian rhythm disturbances in these conditions. Indeed, many studies in elderly people report that increased light intensity improved a disorganized thermoregulation, a disrupted neuroendocrine system, and poor sleep quality and rest-activity rhythms 26, 298, 707. AD patients, whose sleep-wake and rest-activity rhythms are even more severely disrupted, responded well to bright light treatment in many studies 556-558, 708-714. The positive effects of bright light therapy have once more been confirmed in our group’s recent longitudinal double-blind, multicenter studies in elderly and demented subjects (Riemersma, Van Someren, et al., unpublished observations). These findings indicate that the SCN preserves plasticity through the entire life span, even in Alzheimer’s disease. Stimulating the SCN as early as possible by bright light therapy may be of great value in the treatment of AD related circadian rhythm disturbances. It may not only relieve clinical behavioral symptoms, but could also delay the progression of the disease (Riemersma, Van Someren et al., submitted).

Melatonin supplements in aging and Alzheimer’s disease Melatonin is involved in the regulation of circadian rhythms and appears to feedback to the SCN via two specific, high-affinity G protein-coupled MT1 (also called Mel1a) and MT2 (or Mel1b) melatonin receptors (reviewed in 62). Physiologically, the MT1 receptor mediates the acute inhibitory action of melatonin on the SCN 64, which may be important for defining the sensitivity of the SCN to phase-shifting stimuli, and may contribute to the regulation of sleep. The MT2 receptor is reported to mediate the phase-shift effect of melatonin on the SCN of rodents 65, 66. MT1 receptor mRNA 347, 504 and protein (Chapter 5) have been detected in the post-mortem human SCN,

167 chapter 8 while MT2 subtype was not detected, probably due to a very low expression level in the human hypothalamus 347, 504. The circadian effects of melatonin have led to substantial therapeutic applications for jet lag, shift work, blindness, and some circadian-based sleep disorders 6, 60. Endogenous melatonin levels are decreased during aging and even more predominantly in AD with a flattened circadian rhythm 202, 203. In fact, we found that a decreased melatonin production occurs from the earliest AD neuropathological stages onwards, probably due to a disrupted SCN function 254, 302. Supplementary melatonin does not only directly increase the circulating melatonin levels to enhance the peripheral circadian effects of melatonin, but also exerts a central circadian effect on the SCN via melatonin receptors. It is thus considered to be a potential treatment for circadian rhythm disturbances in aging and AD. Interestingly, although the number of the MT1 receptors in the SCN of aged controls is half of that in young subjects, exogenous melatonin is still effective in aged subjects with insomnia (see review 559). This indicates that the amount of the MT1 receptors in the SCN of aged controls is still sufficient to respond to exogenous melatonin. This may also be the case in preclinical AD subjects who have similar levels of the MT1 receptor in the SCN as elderly controls. However, this possibility should be confirmed in clinical trials. The effectiveness of melatonin in clinical AD patients that was first suggested in some pilot studies and case reports, was not confirmed in more recent well-performed placebo-controlled clinical trials 319, 320. In fact, these negative results may be partly explained by our study, that the numbers of the MT1 receptor in the SCN of late clinical AD patients (Braak stage V-VI) is only 10% of those in age-matched control subjects, which may well be too few to mediate an efficient action of the melatonin supplement. Our study suggests that melatonin administration should be applied as early as possible in the AD process. Whether it than may act favorably on the AD process should be investigated.

The role of monoamine oxidase A (MAOA) polymorphism in Alzheimer with depression Depression, which affects up to 50% of AD patients, is among the most frequent neuropsychiatric comorbidities of AD. Depression in AD is associated with severe negative consequences for patients and caregivers and with high costs of care 476, 477, which are major issues in treatment 715-717. MAOA is one of the key enzymes in the metabolism of monoaminergic neurotransmitters which supposedly plays an important role in the etiology of affective disorders. Moreover, MAOA inhibitors are effective in the pharmacotherapy of depressive syndromes 718. A frequent polymorphism of a variable number tandem repeat (VNTR) in

168 general discussion the promoter region of MAOA has been identiﬁed by Sabol et al. 421, and was shown to aﬀect MAOA transcription, and hence activity 421, 422, 463. The long MAOA-VNTR genotype has been reported to be associated with depressive disorders 478, 479. In Chap- ter 3, we found that the long MAOA-VNTR results in higher MAOA gene expression and activity in AD patients. We hypothesize that the long MAOA-VNTR genotyped AD patients who show elevated MAOA activity may have increased susceptibility to depressive disorders. Moreover, these AD patients may be more sensitive for MAOA inhibitor treatment. In our present study the number of AD cases with clear records of disturbed mood was too small to draw any conclusion on this point. Future studies on this hypothesis are needed, in particular since the diagnosis of depression in AD patients needs careful, prospective studies 416, 480.

E. Future experiments

Future experiment 1 In vivo investigation of central to peripheral clock synchronization during the progression of Alzheimer’s disease-:crossing Per-luciferase transgenic mouse and Alzheimer model mouse In the present thesis we show that pineal clock gene oscillation is disturbed from the earliest stages onwards in Alzheimer’s disease, probably due to a functional disconnection from the “master clock”, the suprachiasmatic nucleus. Thus an early disrupted functional activity of the master clock (the SCN) during the progression of AD may desynchronize peripheral clocks (e.g. the pineal gland) from the environmental light/ dark cycle and may form the molecular basis of the circadian rhythm disturbances observed in AD patients. As our studies are observations on postmortem human material, we would like to confirm our ideas on cause and effect in experimental studies. In order to investigate how Alzheimer’s disease pathology affects the circadian timing system on the molecular level, we propose to make use of mouse models, i.e cross clock gene Per-luciferase transgenic mouse model and an established Alzheimer model mouse. Mice carrying the Per1-luciferase or Per2-luciferase transgene have been widely used to monitor the intrinsic Per1 expression or Per2 expression patterns in different brain areas and other tissues, and their response to changes in the light cycle 107, 481. In Per1-luciferease transgenics, the luciferase signal oscillates for a number of cycles before dampening out. Generally, the in vitro oscillation of luciferase in the SCN remains longer than in other tissues dissociated from the central clock 122. In the Per2-luciferase transgenices, the rhythmicity of luciferase remains for a longer period than in the Per1-luciferase transgenics, for reasons that are as yet not quite understood 481.

169 chapter 8

Several Alzheimer mouse models have been generated that show either amyloid plaques 719, 720, tau tangles 721 or a combination of the two 722. It is generally thought that the latter are most relevant to Alzheimer’s disease, but the other two types of mice provide the opportunity to dissociate tangles and plaques, and study the consequences of either on, for instance, circadian rhythmicity. As a starting point we propose to cross Alzheimer mouse models that show both tangles and plaques with Per1-luciferase transgenics, as these will have the highest load of pathology and the most sensitive reporter for circadian synchronization. Acute in vitro slices of both SCN and pineal in these Alzheimer/luciferase mice will allow correlation of the onset of neuropathology (plaques or tangles) with the loss of synchronization of the SCN and the pineal. Moreover, according to our groups’ hypothesis, paraphrased as “use it or loss it”, we propose that stimulating the SCN may reactivate the central clock, and thus re-synchronize peripheral clocks during aging and AD, an hypothesis that could be tested in these mouse models. The mouse experiments allow us to interfere with the environmental stimulation of the SCN, for instance by providing the mice with higher light intensity or diﬀerent light/dark conditions. These experiments could explain the mixed outcomes of light activation therapy for Alzheimer’s disease.

Future experiment 2 Multiple gene polymorphisms in Alzheimer’s disease with depression Over 80% of Alzheimer’s disease (AD) patients suffer from “noncognitive” neuropsychiatric symptoms, which are a major focus of treatment 715-717. Affecting up to 50% of AD patients, depression is among the most frequent neuropsychiatric comorbidities of AD 476, and is associated with severe negative consequences for patients and caregivers and greater costs of care 476, 477. As we discussed in general discussion section D, MAOA-VNTR gene polymorphism is a candidate gene for association studies for Alzheimer with depression. Apart from the MAOA gene, we will also include some other genes in a future association study. Most of the published genetic association studies on mood disorders have focused on functional polymorphisms in the loci encoding the serotonin transporter (SLC6A4), serotonin 2A receptor (5HTR2A), tyrosine hydroxylase (TH) (the limiting enzyme for dopamine synthesis), tryptophan hydroxylase 1 (TPH1) (serotonin synthesis), and catechol-o-methyltransferase (COMT) (dopamine catabolism). Recent meta-analyses of these studies have shown that serotonin transporter gene-linked polymorphic region (5-HTTLPR) 44-base pair (bp) insertion/deletion polymorphism is significantly associated with in particular mood disorders, suicidal behavior, and/or neuroticism throughout all the available studies (see review 723), strongly suggesting its role in the pathogenesis of depressive disorders. Numerous postmortem studies have shown

170 general discussion evidence of extensive degeneration on the 5-HT system in Alzheimer’s disease 724, which may contribute to depressive disorders in AD. This degeneration of the 5-HT system includes reduction of 5-HT and its metabolites 725, decrease of 5-HT reuptake 726, decreases in cortical 5-HT receptors, with 5-HT2A preferentially aﬀected over 5- HT1A receptors 726, 727, and loss of serotonin transporter 689, 728, 729, which is responsible for the reuptake of 5HT from the synaptic cleft. Interestingly, the low-activity allele of the 5HTTLPR is a risk factor for late-onset AD 730. Moreover, 5-HTTLPR is associated with depression in Parkinson’s disease 731. Taken together, 5-HTTLPR is an excellent candidate gene for an association study on predisposition for depressive disorder in Alzheimer’s disease. The proposed future studies of association of MAOA-VNTR, 5-HTTLPR and other candidate genes discussed above might provide novel ﬁndings, which may be of diagnostic and potentially also of therapeutic and prognostic value for depression in AD patients. Such studies may underlie a gene-by-disease interaction in which an individual response to disease is modulated by his or her genetic makeup.

Future experiment 3 A direct effect of melatonin on the expression of human corticotropin-releasing hormone? As we mentioned above, depression is a major focus of treatment for non-cognitive psychiatric symptoms in AD 476, 477. Yet available studies on the natural course, etiology, and treatment of depression in AD have been few and not equivocal 480 The hypothalamic–pituitary–adrenal (HPA) axis is the ﬁnal common pathway in the stress response. Corticotropin-releasing hormone (CRH), synthesized in the paraventricular nucleus (PVN) of the hypothalamus, plays a key role in the regulation of the HPA axis 520, 732, 733. A characteristic hyperactivity of the HPA axis in depression, resulting from the hyperactive hypothalamic CRH neurons 579, 580, plays a causal role in the development and course of depression (see review 734). Interestingly, recent studies of our group show that the number of CRH-expressing neurons in the human hypothalamic PVN is 3-fold increased in Alzheimer patients with depression compared to Alzheimer patients without depression (p=0.002) (Meynen, et al., in preparation). This observation suggests that a hyperactive HPA axis may be an important neurobiological basis for the depressive comorbidity in Alzheimer’s disease, just as in major depression. Normalization of HPA system regulation is thus thought to be a prerequisite for favorable treatment response in AD patients with depression. The pineal hormone melatonin is dramatically decreased in Alzheimer’s disease patients 203, 254, 301. Previously, melatonin has been proposed in the treatment of Alzhe- imer’s disease because of its neuroprotective role 163, 267, 304, 316, its circadian regulatory eﬀects 163, 316 and its easy penetration through the blood-brain-barrier 251. Moreover,

171 chapter 8 melatonin also exerts a stress-protective role through counteracting the dysregulation of the HPA axis. It attenuates the secretory response of the HPA axis, including the increase of CRH levels, and normalizes most of the symptoms of dysregulated HPA axis in response to acute or chronic stress 509, 510, 735. In humans, there is an inverse relationship between plasma melatonin and cortisol circadian rhythms 736, 737. Decreased melatonin production in AD 203, 254, 301, which is also reported in many studies of major depression 212, 564, may contribute to the hyperactivity of the HPA axis. We propose, therefore, that administration of melatonin may normalize the HPA axis and provide a beneficial response in AD patients with depression. Available studies have indeed showed positive effects of melatonin on mood disorders in elderly with mild cognitive impairment and in AD patients 318, 738, which apparently support our idea, although more systematic studies are certainly needed. Melatonin appears to exert many effects via two different subtypes of G-protein coupled receptors, i.e. the MT1 and MT2 receptors. Our recent study have revealed the strong expression of MT1 receptor in the PVN of human hypothalamus at the protein level by immunocytochemisty 739 with an anti-MT1 antibody 378. Moreover, interestingly, we found that MT1 receptor is co-localized with CRH neurons in the PVN of human hypothalamus 739, which suggests the possibility of a direct action of melatonin on the CRH neurons via MT1 receptor. In vitro studies show that in many responses CRH gene expression is regulated by the cAMP signaling pathway via a cAMP-resonsive element (CRE) in the CRH gene promoter between -248 and -213 bp 740-743. While melatonin acting through the MT1 receptor produces inhibitory responses on the cAMP signal transduction cascade 63, 378, 382-385. Therefore, we hypothesize that melatonin acting through MT1 receptor directly regulates CRH gene expression via the cAMP regulatory element. In this future study we aim to identify a direct regulatory effect of melatonin in the CRH gene expression in the hypothalamic PVN neurons, and its underlying mechanism. Our study will provide the molecular basis for the proposed therapeutic benefits of melatonin in treatment of depressive comorbidity in Alzheimer’s disease, in addition to the neuroprotective and circadian regulatory roles which have been previously proposed for melatonin in Alzheimer’s disease.

172 References

1. Hastings, M. H., Reddy, A. B. & Maywood, E. S. A clockwork web: circadian timing in brain and periphery, in health and disease. Nat Rev Neurosci 4, 649-61 (2003). 2. Klein, D., Moore, R. Y. & Reppert, S. M. Suprachiasmatic Nucleus: The Mind’s Clock (Oxford Univ.Press. , New York, 1991). 3. Moore, R. Y. Circadian rhythms: basic neurobiology and clinical applications. Annu Rev Med 48, 253-66 (1997). 4. Abrahamson, E. E. & Moore, R. Y. Suprachiasmatic nucleus in the mouse: retinal innervation, intrinsic organization and efferent projections. Brain Res 916, 172-91 (2001). 5. Ibata, Y. et al. Functional morphology of the suprachiasmatic nucleus. Front Neuroendocrinol 20, 241-68 (1999). 6. Swaab, D. F. Human hypothalamus : basic and clinical aspects. Part I, Nuclei of the human hypothalamus (eds. Aminoff Michael J. , Franpcois, B. & Swaab, D. F.) (Elsvier, Amsterdam, 2003). 7. Van Esseveldt, K. E., Lehman, M. N. & Boer, G. J. The suprachiasmatic nucleus and the circadian time-keeping system revisited. Brain Res Brain Res Rev 33, 34-77 (2000). 8. Drucker-Colin, R., Aguilar-Roblero, R., Garcia-Hernandez, F., Fernandez-Cancino, F. & Bermu- dez Rattoni, F. Fetal suprachiasmatic nucleus transplants: diurnal rhythm recovery of lesioned rats. Brain Res 311, 353-7 (1984). 9. Aguilar-Roblero, R., Garcia-Hernandez, F., Aguilar, R., Arankowsky-Sandoval, G. & Drucker- Colin, R. Suprachiasmatic nucleus transplants function as an endogenous oscillator only in constant darkness. Neurosci Lett 69, 47-52 (1986). 10. Lehman, M. N. et al. Circadian rhythmicity restored by neural transplant. Immunocytochemi- cal characterization of the graft and its integration with the host brain. J Neurosci 7, 1626-38 (1987). 11. DeCoursey, P. J. & Buggy, J. Circadian rhythmicity after neural transplant to hamster third ventricle: specificity of suprachiasmatic nuclei. Brain Res 500, 263-75 (1989). 12. Saitoh, Y., Nihonmatsu, I. & Kawamura, H. Location of the suprachiasmatic nucleus grafts in rats which restored circadian rhythmicity after transplantation. Neurosci Lett 118, 45-8 (1990). 13. Griffioen, H. A., Duindam, H., Van der Woude, T. P., Rietveld, W. J. & Boer, G. J. Functional development of fetal suprachiasmatic nucleus grafts in suprachiasmatic nucleus-lesioned rats. Brain Res Bull 31, 145-60 (1993). 14. Ralph, M. R., Foster, R. G., Davis, F. C. & Menaker, M. Transplanted suprachiasmatic nucleus determines circadian period. Science 247, 975-8 (1990). 15. Groos, G. & Hendriks, J. Circadian rhythms in electrical discharge of rat suprachiasmatic neurones recorded in vitro. Neurosci Lett 34, 283-8 (1982). 16. Meijer, J. H., Schaap, J., Watanabe, K. & Albus, H. Multiunit activity recordings in the suprachiasmatic nuclei: in vivo versus in vitro models. Brain Res 753, 322-7 (1997). 17. Sato, T. & Kawamura, H. Circadian rhythms in multiple unit activity inside and outside the suprachiasmatic nucleus in the diurnal chipmunk (Eutamias sibiricus). Neurosci Res 1, 45-52 (1984). 18. Schwartz, W. J., Reppert, S. M., Eagan, S. M. & Moore-Ede, M. C. In vivo metabolic activity of the suprachiasmatic nuclei: a comparative study. Brain Res 274, 184-7 (1983). 19. Moore, R. Y. The fourth C.U. Ariens Kappers lecture. The organization of the human circadian timing system. Prog Brain Res 93, 99-115; discussion 115-7 (1992). 20. Hofman, M. A. & Swaab, D. F. Diurnal and seasonal rhythms of neuronal activity in the suprachiasmatic nucleus of humans. J Biol Rhythms 8, 283-95 (1993).

175 references

21. Hofman, M. A. Circadian oscillations of neuropeptide expression in the human biological clock. J Comp Physiol A Neuroethol Sens Neural Behav Physiol 189, 823-831 (2003). 22. Hofman, M. A., Fliers, E., Goudsmit, E. & Swaab, D. F. Morphometric analysis of the suprachiasmatic and paraventricular nuclei in the human brain: sex differences and age-dependent changes. J Anat 160, 127-43 (1988). 23. Swaab, D. F., Fliers, E. & Partiman, T. S. The suprachiasmatic nucleus of the human brain in relation to sex, age and senile dementia. Brain Res 342, 37-44 (1985). 24. Dai, J., Swaab, D. F. & Buijs, R. M. Distribution of vasopressin and vasoactive intestinal polypeptide (VIP) fibers in the human hypothalamus with special emphasis on suprachiasmatic nucleus efferent projections. J Comp Neurol 383, 397-414 (1997). 25. Dai, J., Swaab, D. F., Van der Vliet, J. & Buijs, R. M. Postmortem tracing reveals the organization of hypothalamic projections of the suprachiasmatic nucleus in the human brain. J Comp Neurol 400, 87-102 (1998). 26. Mishima, K., Okawa, M., Hozumi, S. & Hishikawa, Y. Supplementary administration of artifi- cial bright light and melatonin as potent treatment for disorganized circadian rest-activity and dysfunctional autonomic and neuroendocrine systems in institutionalized demented elderly persons. Chronobiol Int 17, 419-32 (2000). 27. Scully, R. E., Mark, E. J. & McNeeley, B. U. Case records of the Massachusetts General Hospi- tal. Weekly clinicopathological exercises. Case 33-1983. A 55-year-old woman with diabetes insipidus and a suprasellar mass. N Engl J Med 309, 418-25 (1983). 28. Schwartz, W. J., Busis, N. A. & Hedley-Whyte, E. T. A discrete lesion of ventral hypothalamus and optic chiasm that disturbed the daily temperature rhythm. J Neurol 233, 1-4 (1986). 29. Cohen, R. A. & Albers, H. E. Disruption of human circadian and cognitive regulation following a discrete hypothalamic lesion: a case study. Neurology 41, 726-9 (1991). 30. Haugh, R. M. & Markesbery, W. R. Hypothalamic astrocytoma. Syndrome of hyperphagia, obesity, and disturbances of behavior and endocrine and autonomic function. Arch Neurol 40, 560-3 (1983). 31. Lockley, S. W. et al. Melatonin administration can entrain the free-running circadian system of blind subjects. J Endocrinol 164, R1-6 (2000). 32. Sack, R. L., Brandes, R. W., Kendall, A. R. & Lewy, A. J. Entrainment of free-running circadian rhythms by melatonin in blind people. N Engl J Med 343, 1070-7 (2000). 33. Klerman, E. B. et al. Nonphotic entrainment of the human circadian pacemaker. Am J Physiol 274, R991-6 (1998). 34. Brainard, G. C. et al. Action spectrum for melatonin regulation in humans: evidence for a novel circadian photoreceptor. J Neurosci 21, 6405-12 (2001). 35. Thapan, K., Arendt, J. & Skene, D. J. An action spectrum for melatonin suppression: evidence for a novel non-rod, non-cone photoreceptor system in humans. J Physiol 535, 261-7 (2001). 36. Lucas, R. J. & Foster, R. G. Neither functional rod photoreceptors nor rod or cone outer segments are required for the photic inhibition of pineal melatonin. Endocrinology 140, 1520-4 (1999). 37. Gooley, J. J., Lu, J., Chou, T. C., Scammell, T. E. & Saper, C. B. Melanopsin in cells of origin of the retinohypothalamic tract. Nat Neurosci 4, 1165 (2001). 38. Moore, R. Y. Entrainment pathways and the functional organization of the circadian system. Prog Brain Res 111, 103-19 (1996). 39. Berson, D. M. Strange vision: ganglion cells as circadian photoreceptors. Trends Neurosci 26, 314-20 (2003). 40. Rollag, M. D., Berson, D. M. & Provencio, I. Melanopsin, ganglion-cell photoreceptors, and mammalian photoentrainment. J Biol Rhythms 18, 227-34 (2003). 41. Moore, R. Y., Speh, J. C. & Card, J. P. The retinohypothalamic tract originates from a distinct subset of retinal ganglion cells. J Comp Neurol 352, 351-66 (1995).

176 references

42. Hannibal, J. et al. Pituitary adenylate cyclase-activating peptide (PACAP) in the retinohypothalamic tract: a potential daytime regulator of the biological clock. J Neurosci 17, 2637-44 (1997). 43. Gooley, J. J., Lu, J., Fischer, D. & Saper, C. B. A broad role for melanopsin in nonvisual photoreception. J Neurosci 23, 7093-106 (2003). 44. Berson, D. M., Dunn, F. A. & Takao, M. Phototransduction by retinal ganglion cells that set the circadian clock. Science 295, 1070-3 (2002). 45. Hannibal, J. & Fahrenkrug, J. Melanopsin: a novel photopigment involved in the photoentrainment of the brain’s biological clock? Ann Med 34, 401-7 (2002). 46. Hattar, S., Liao, H. W., Takao, M., Berson, D. M. & Yau, K. W. Melanopsin-containing retinal ganglion cells: architecture, projections, and intrinsic photosensitivity. Science 295, 1065-70 (2002). 47. Hankins, M. W. & Lucas, R. J. The primary visual pathway in humans is regulated according to long-term light exposure through the action of a nonclassical photopigment. Curr Biol 12, 191-8 (2002). 48. Dai, J., Van der Vliet, J., Swaab, D. F. & Buijs, R. M. Human retinohypothalamic tract as revealed by in vitro postmortem tracing. J Comp Neurol 397, 357-70 (1998). 49. Ding, J. M. et al. Resetting the biological clock: mediation of nocturnal circadian shifts by glutamate and NO. Science 266, 1713-7 (1994). 50. Mikkelsen, J. D. & Serviere, J. Demonstration of a direct projection from the retina to the hypothalamic supraoptic nucleus of the hamster. Neurosci Lett 139, 149-52 (1992). 51. Mikkelsen, J. D. Visualization of efferent retinal projections by immunohistochemical identification of cholera toxin subunit B. Brain Res Bull 28, 619-23 (1992). 52. Meijer, J. H., Groos, G. A. & Rusak, B. Luminance coding in a circadian pacemaker: the suprachiasmatic nucleus of the rat and the hamster. Brain Res 382, 109-18 (1986). 53. Buijs, R. M. & Kalsbeek, A. Hypothalamic integration of central and peripheral clocks. Nat Rev Neurosci 2, 521-6 (2001). 54. Mason, R. & Brooks, A. The electrophysiological effects of melatonin and a putative melatonin antagonist (N-acetyltryptamine) on rat suprachiasmatic neurones in vitro. Neurosci Lett 95, 296-301 (1988). 55. Stehle, J., Vanecek, J. & Vollrath, L. Effects of melatonin on spontaneous electrical activity of neurons in rat suprachiasmatic nuclei: an in vitro iontophoretic study. J Neural Transm 78, 173-7 (1989). 56. Redman, J., Armstrong, S. & Ng, K. T. Free-running activity rhythms in the rat: entrainment by melatonin. Science 219, 1089-91 (1983). 57. Cassone, V. M. Effects of melatonin on vertebrate circadian systems. Trends Neurosci 13, 457-64 (1990). 58. Lewy, A. J., Ahmed, S., Jackson, J. M. & Sack, R. L. Melatonin shifts human circadian rhythms according to a phase-response curve. Chronobiol Int 9, 380-92 (1992). 59. Palm, L., Blennow, G. & Wetterberg, L. Correction of non-24-hour sleep/wake cycle by melatonin in a blind retarded boy. Ann Neurol 29, 336-9 (1991). 60. Arendt, J., Skene, D. J., Middleton, B., Lockley, S. W. & Deacon, S. Efficacy of melatonin treatment in jet lag, shift work, and blindness. J Biol Rhythms 12, 604-17 (1997). 61. Dubocovich, M. L. Melatonin receptors: are there multiple subtypes? Trends Pharmacol Sci 16, 50-6 (1995). 62. Reppert, S. M., Weaver, D. R. & Godson, C. Melatonin receptors step into the light: cloning and classification of subtypes. Trends Pharmacol Sci 17, 100-2 (1996). 63. Reppert, S. M., Weaver, D. R. & Ebisawa, T. Cloning and characterization of a mammalian melatonin receptor that mediates reproductive and circadian responses. Neuron 13, 1177-85 (1994).

177 references

64. Liu, C. et al. Molecular dissection of two distinct actions of melatonin on the suprachiasmatic circadian clock. Neuron 19, 91-102 (1997). 65. Dubocovich, M. L., Yun, K., Al-Ghoul, W. M., Benloucif, S. & Masana, M. I. Selective MT2 melatonin receptor antagonists block melatonin-mediated phase advances of circadian rhythms. Faseb J 12, 1211-20 (1998). 66. Hunt, A. E., Al-Ghoul, W. M., Gillette, M. U. & Dubocovich, M. L. Activation of MT(2) melatonin receptors in rat suprachiasmatic nucleus phase advances the circadian clock. Am J Physiol Cell Physiol 280, C110-8 (2001). 67. Jin, X. et al. Targeted disruption of the mouse Mel(1b) melatonin receptor. Mol Cell Biol 23, 1054-60 (2003). 68. Pickard, G. E., Weber, E. T., Scott, P. A., Riberdy, A. F. & Rea, M. A. 5HT1B receptor agonists inhibit light-induced phase shifts of behavioral circadian rhythms and expression of the immediate-early gene c-fos in the suprachiasmatic nucleus. J Neurosci 16, 8208-20 (1996). 69. Rea, M. A., Glass, J. D. & Colwell, C. S. Serotonin modulates photic responses in the hamster suprachiasmatic nuclei. J Neurosci 14, 3635-42 (1994). 70. Yannielli, P. C. & Harrington, M. E. Neuropeptide Y applied in vitro can block the phase shifts induced by light in vivo. Neuroreport 11, 1587-91 (2000). 71. Ying, S. W. & Rusak, B. Effects of serotonergic agonists on firing rates of photically responsive cells in the hamster suprachiasmatic nucleus. Brain Res 651, 37-46 (1994). 72. Hastings, M. H. et al. Non-photic signalling in the suprachiasmatic nucleus. Biol Cell 89, 495- 503 (1997). 73. Moore, R. Y. The geniculohypothalamic tract in monkey and man. Brain Res 486, 190-4 (1989). 74. Moore, R. Y. & Speh, J. C. A putative retinohypothalamic projection containing substance P in the human. Brain Res 659, 249-53 (1994). 75. Chevassus-au-Louis, N. & Cooper, H. M. Is there a geniculohypothalamic tract in primates? A comparative immunohistochemical study in the circadian system of strepsirhine and haplorhine species. Brain Res 805, 213-9 (1998). 76. Mistlberger, R. E., Bossert, J. M., Holmes, M. M. & Marchant, E. G. Serotonin and feedback effects of behavioral activity on circadian rhythms in mice. Behav Brain Res 96, 93-9 (1998). 77. Marchant, E. G., Watson, N. V. & Mistlberger, R. E. Both neuropeptide Y and serotonin are necessary for entrainment of circadian rhythms in mice by daily treadmill running schedules. J Neurosci 17, 7974-87 (1997). 78. Horikawa, K. et al. Nonphotic entrainment by 5-HT1A/7 receptor agonists accompanied by reduced Per1 and Per2 mRNA levels in the suprachiasmatic nuclei. J Neurosci 20, 5867-73 (2000). 79. Morin, L. P. Serotonin and the regulation of mammalian circadian rhythmicity. Ann Med 31, 12-33 (1999). 80. Dai, J., Van Der Vliet, J., Swaab, D. F. & Buijs, R. M. Postmortem anterograde tracing of intrahypothalamic projections of the human dorsomedial nucleus of the hypothalamus. J Comp Neurol 401, 16-33 (1998). 81. Watts, A. G. & Swanson, L. W. Efferent projections of the suprachiasmatic nucleus: II. Studies using retrograde transport of fluorescent dyes and simultaneous peptide immunohistochemistry in the rat. J Comp Neurol 258, 230-52 (1987). 82. Watts, A. G., Swanson, L. W. & Sanchez-Watts, G. Efferent projections of the suprachiasmatic nucleus: I. Studies using anterograde transport of Phaseolus vulgaris leucoagglutinin in the rat. J Comp Neurol 258, 204-29 (1987). 83. Lu, J. et al. Contrasting effects of ibotenate lesions of the paraventricular nucleus and subparaventricular zone on sleep-wake cycle and temperature regulation. J Neurosci 21, 4864-74 (2001).

178 references

84. Kalsbeek, A. et al. GABA receptors in the region of the dorsomedial hypothalamus of rats are implicated in the control of melatonin and corticosterone release. Neuroendocrinology 63, 69-78 (1996). 85. Reiter, R. J. The melatonin rhythm: both a clock and a calendar. Experientia 49, 654-64 (1993). 86. Macchi, M. M. & Bruce, J. N. Human pineal physiology and functional significance of melatonin. Front Neuroendocrinol 25, 177-95 (2004). 87. Kalsbeek, A., Teclemariam-Mesbah, R. & Pevet, P. Efferent projections of the suprachiasmatic nucleus in the golden hamster (Mesocricetus auratus). J Comp Neurol 332, 293-314 (1993). 88. Cui, L. N., Saeb-Parsy, K. & Dyball, R. E. Neurones in the supraoptic nucleus of the rat are regulated by a projection from the suprachiasmatic nucleus. J Physiol 502 ( Pt 1), 149-59 (1997). 89. Abrahamson, E. E., Leak, R. K. & Moore, R. Y. The suprachiasmatic nucleus projects to posterior hypothalamic arousal systems. Neuroreport 12, 435-40 (2001). 90. Miller, J. D. On the nature of the circadian clock in mammals. Am J Physiol 264, R821-32 (1993). 91. Meijer, J. H. & Rietveld, W. J. Neurophysiology of the suprachiasmatic circadian pacemaker in rodents. Physiol Rev 69, 671-707 (1989). 92. Carter, D. A. & Murphy, D. Nuclear mechanisms mediate rhythmic changes in vasopressin mRNA expression in the rat suprachiasmatic nucleus. Brain Res Mol Brain Res 12, 315-21 (1992). 93. Jin, X. et al. A molecular mechanism regulating rhythmic output from the suprachiasmatic circadian clock. Cell 96, 57-68 (1999). 94. Schwartz, W. J., Lydic, R. & Moore-Ede, M. C. In vivo metabolic activity of the suprachiasmatic nuclei: non-uniform intranuclear distribution of 14C-labeled deoxyglucose uptake. Brain Res 424, 249-57 (1987). 95. Ingram, C. D. et al. Vasopressin neurotransmission and the control of circadian rhythms in the suprachiasmatic nucleus. Prog Brain Res 119, 351-64 (1998). 96. Tousson, E. & Meissl, H. Suprachiasmatic nuclei grafts restore the circadian rhythm in the paraventricular nucleus of the hypothalamus. J Neurosci 24, 2983-8 (2004). 97. Kalsbeek, A. et al. Melatonin sees the light: blocking GABA-ergic transmission in the paraventricular nucleus induces daytime secretion of melatonin. Eur J Neurosci 12, 3146-54 (2000). 98. Perreau-Lenz, S. et al. Suprachiasmatic control of melatonin synthesis in rats: inhibitory and stimulatory mechanisms. Eur J Neurosci 17, 221-8 (2003). 99. Emery, P. & Reppert, S. M. A rhythmic Ror. Neuron 43, 443-6 (2004). 100. Antle, M. C. & Silver, R. Orchestrating time: arrangements of the brain circadian clock. Trends Neurosci 28, 145-51 (2005). 101. Reppert, S. M. & Weaver, D. R. Molecular analysis of mammalian circadian rhythms. Annu Rev Physiol 63, 647-76 (2001). 102. Whitmore, D. et al. A clockwork organ. Biol Chem 381, 793-800 (2000). 103. Lowrey, P. L. & Takahashi, J. S. Genetics of the mammalian circadian system: Photic entrainment, circadian pacemaker mechanisms, and posttranslational regulation. Annu Rev Genet 34, 533-562 (2000). 104. Etchegaray, J. P., Lee, C., Wade, P. A. & Reppert, S. M. Rhythmic histone acetylation underlies transcription in the mammalian circadian clock. Nature 421, 177-82 (2003). 105. Gekakis, N. et al. Role of the CLOCK protein in the mammalian circadian mechanism. Science 280, 1564-9 (1998). 106. Hogenesch, J. B., Gu, Y. Z., Jain, S. & Bradfield, C. A. The basic-helix-loop-helix-PAS orphan MOP3 forms transcriptionally active complexes with circadian and hypoxia factors. Proc Natl Acad Sci U S A 95, 5474-9 (1998).

179 references

107. Yamaguchi, S. et al. The 5’ upstream region of mPer1 gene contains two promoters and is responsible for circadian oscillation. Curr Biol 10, 873-6 (2000). 108. Travnickova-Bendova, Z., Cermakian, N., Reppert, S. M. & Sassone-Corsi, P. Bimodal regulation of mPeriod promoters by CREB-dependent signaling and CLOCK/BMAL1 activity. Proc Natl Acad Sci U S A 99, 7728-33 (2002). 109. Griffin, E. A., Jr., Staknis, D. & Weitz, C. J. Light-independent role of CRY1 and CRY2 in the mammalian circadian clock. Science 286, 768-71 (1999). 110. Kume, K. et al. mCRY1 and mCRY2 are essential components of the negative limb of the circadian clock feedback loop. Cell 98, 193-205 (1999). 111. Preitner, N. et al. The orphan nuclear receptor REV-ERBalpha controls circadian transcription within the positive limb of the mammalian circadian oscillator. Cell 110, 251-60 (2002). 112. Ueda, H. R. et al. A transcription factor response element for gene expression during circadian night. Nature 418, 534-9 (2002). 113. Sato, T. K. et al. A functional genomics strategy reveals Rora as a component of the mammalian circadian clock. Neuron 43, 527-37 (2004). 114. Lee, C., Etchegaray, J. P., Cagampang, F. R., Loudon, A. S. & Reppert, S. M. Posttranslational mechanisms regulate the mammalian circadian clock. Cell 107, 855-67 (2001). 115. Vielhaber, E., Eide, E., Rivers, A., Gao, Z. H. & Virshup, D. M. Nuclear entry of the circadian regulator mPER1 is controlled by mammalian casein kinase I epsilon. Mol Cell Biol 20, 4888-99 (2000). 116. Yagita, K. et al. Nucleocytoplasmic shuttling and mCRY-dependent inhibition of ubiquitylation of the mPER2 clock protein. Embo J 21, 1301-14 (2002). 117. Camacho, F. et al. Human casein kinase Idelta phosphorylation of human circadian clock proteins period 1 and 2. FEBS Lett 489, 159-65 (2001). 118. Eide, E. J., Vielhaber, E. L., Hinz, W. A. & Virshup, D. M. The circadian regulatory proteins BMAL1 and cryptochromes are substrates of casein kinase Iepsilon. J Biol Chem 277, 17248-54 (2002). 119. Akashi, M., Tsuchiya, Y., Yoshino, T. & Nishida, E. Control of intracellular dynamics of mammalian period proteins by casein kinase I epsilon (CKIepsilon) and CKIdelta in cultured cells. Mol Cell Biol 22, 1693-703 (2002). 120. Zylka, M. J., Shearman, L. P., Weaver, D. R. & Reppert, S. M. Three period homologs in mammals: differential light responses in the suprachiasmatic circadian clock and oscillating transcripts outside of brain. Neuron 20, 1103-10 (1998). 121. Balsalobre, A., Damiola, F. & Schibler, U. A serum shock induces circadian gene expression in mammalian tissue culture cells. Cell 93, 929-37 (1998). 122. Yamazaki, S. et al. Resetting central and peripheral circadian oscillators in transgenic rats. Sci- ence 288, 682-5 (2000). 123. Balsalobre, A. Clock genes in mammalian peripheral tissues. Cell Tissue Res 309, 193-9 (2002). 124. Lowrey, P. L. & Takahashi, J. S. Mammalian circadian biology: elucidating genome-wide levels of temporal organization. Annu Rev Genomics Hum Genet 5, 407-41 (2004). 125. Abe, M. et al. Circadian rhythms in isolated brain regions. J Neurosci 22, 350-6 (2002). 126. Bartness, T. J., Song, C. K. & Demas, G. E. SCN efferents to peripheral tissues: implications for biological rhythms. J Biol Rhythms 16, 196-204 (2001). 127. Balsalobre, A. et al. Resetting of circadian time in peripheral tissues by glucocorticoid signaling. Science 289, 2344-7 (2000). 128. Damiola, F. et al. Restricted feeding uncouples circadian oscillators in peripheral tissues from the central pacemaker in the suprachiasmatic nucleus. Genes Dev 14, 2950-61 (2000). 129. Stokkan, K. A., Yamazaki, S., Tei, H., Sakaki, Y. & Menaker, M. Entrainment of the circadian clock in the liver by feeding. Science 291, 490-3 (2001).

180 references

130. Ripperger, J. A., Shearman, L. P., Reppert, S. M. & Schibler, U. CLOCK, an essential pacemaker component, controls expression of the circadian transcription factor DBP. Genes Dev 14, 679- 89 (2000). 131. Lavery, D. J. et al. Circadian expression of the steroid 15 alpha-hydroxylase (Cyp2a4) and cou- marin 7-hydroxylase (Cyp2a5) genes in mouse liver is regulated by the PAR leucine zipper transcription factor DBP. Mol Cell Biol 19, 6488-99 (1999). 132. Mitsui, S., Yamaguchi, S., Matsuo, T., Ishida, Y. & Okamura, H. Antagonistic role of E4BP4 and PAR proteins in the circadian oscillatory mechanism. Genes Dev 15, 995-1006 (2001). 133. Pando, M. P., Morse, D., Cermakian, N. & Sassone-Corsi, P. Phenotypic rescue of a peripheral clock genetic defect via SCN hierarchical dominance. Cell 110, 107-17 (2002). 134. Cermakian, N. & Boivin, D. B. A molecular perspective of human circadian rhythm disorders. Brain Res Brain Res Rev 42, 204-20 (2003). 135. Hida, A. et al. The human and mouse Period1 genes: five well-conserved E-boxes additively contribute to the enhancement of mPer1 transcription. Genomics 65, 224-33 (2000). 136. Kawara, S. et al. Low-dose ultraviolet B rays alter the mRNA expression of the circadian clock genes in cultured human keratinocytes. J Invest Dermatol 119, 1220-3 (2002). 137. Steeves, T. D. et al. Molecular cloning and characterization of the human CLOCK gene: expression in the suprachiasmatic nuclei. Genomics 57, 189-200 (1999). 138. Zanello, S. B., Jackson, D. M. & Holick, M. F. Expression of the circadian clock genes clock and period1 in human skin. J Invest Dermatol 115, 757-60 (2000). 139. Bjarnason, G. A. et al. Circadian expression of clock genes in human oral mucosa and skin: association with specific cell-cycle phases. Am J Pathol 158, 1793-801 (2001). 140. Boivin, D. B. et al. Circadian clock genes oscillate in human peripheral blood mononuclear cells. Blood 102, 4143-4145 (2003). 141. Kusanagi, H. et al. Similar profiles in human period1 gene expression in peripheral mononuclear and polymorphonuclear cells. Neurosci Lett 365, 124-7 (2004). 142. Bjarnason, G. A. & Jordan, R. Rhythms in human gastrointestinal mucosa and skin. Chronobiol Int 19, 129-40 (2002). 143. Takata, M. et al. Daily expression of mRNAs for the mammalian Clock genes Per2 and clock in mouse suprachiasmatic nuclei and liver and human peripheral blood mononuclear cells. Jpn J Pharmacol 90, 263-9 (2002). 144. Oishi, K., Sakamoto, K., Okada, T., Nagase, T. & Ishida, N. Antiphase circadian expression between BMAL1 and period homologue mRNA in the suprachiasmatic nucleus and peripheral tissues of rats. Biochem Biophys Res Commun 253, 199-203 (1998). 145. Van Someren, E. J., Riemersma, R. F. & Swaab, D. F. Functional plasticity of the circadian timing system in old age: light exposure. Prog Brain Res 138, 205-31 (2002). 146. Carrier, J., Paquet, J., Morettini, J. & Touchette, E. Phase advance of sleep and temperature circadian rhythms in the middle years of life in humans. Neurosci Lett 320, 1-4 (2002). 147. Weitzman, E. D., Moline, M. L., Czeisler, C. A. & Zimmerman, J. C. Chronobiology of aging: temperature, sleep-wake rhythms and entrainment. Neurobiol Aging 3, 299-309 (1982). 148. Mirmiran, M. et al. Circadian rhythms and the suprachiasmatic nucleus in perinatal development, aging and Alzheimer’s disease. Prog Brain Res 93, 151-62; discussion 162-3 (1992). 149. Huang, Y. L. et al. Age-associated difference in circadian sleep-wake and rest-activity rhythms. Physiol Behav 76, 597-603 (2002). 150. Myers, B. L. & Badia, P. Changes in circadian rhythms and sleep quality with aging: mechanisms and interventions. Neurosci Biobehav Rev 19, 553-71 (1995). 151. Zhou, J. N., Liu, R. Y., Van Heerikhuize, J., Hofman, M. A. & Swaab, D. F. Alterations in the circadian rhythm of salivary melatonin begin during middle-age. J Pineal Res 34, 11-6 (2003).

181 references

152. Tenover, J. S., Matsumoto, A. M., Clifton, D. K. & Bremner, W. J. Age-related alterations in the circadian rhythms of pulsatile luteinizing hormone and testosterone secretion in healthy men. J Gerontol 43, M163-9 (1988). 153. Olsson, T. et al. Catecholamine excretion in old age. Aging (Milano) 3, 263-8 (1991). 154. Touitou, Y. Effects of ageing on endocrine and neuroendocrine rhythms in humans. Horm Res 43, 12-9 (1995). 155. Magri, F. et al. Changes in endocrine circadian rhythms as markers of physiological and pathological brain aging. Chronobiol Int 14, 385-96 (1997). 156. Forsling, M. L., Montgomery, H., Halpin, D., Windle, R. J. & Treacher, D. F. Daily patterns of secretion of neurohypophysial hormones in man: effect of age. Exp Physiol 83, 409-18 (1998). 157. Monk, T. H., Buysse, D. J., Reynolds, C. F., 3rd, Kupfer, D. J. & Houck, P. R. Circadian temperature rhythms of older people. Exp Gerontol 30, 455-74 (1995). 158. Touitou, Y., Bogdan, A., Haus, E. & Touitou, C. Modifications of circadian and circannual rhythms with aging. Exp Gerontol 32, 603-14 (1997). 159. Duffy, J. F. et al. Peak of circadian melatonin rhythm occurs later within the sleep of older subjects. Am J Physiol Endocrinol Metab 282, E297-303 (2002). 160. Witting, W., Kwa, I. H., Eikelenboom, P., Mirmiran, M. & Swaab, D. F. Alterations in the circadian rest-activity rhythm in aging and Alzheimer’s disease. Biol Psychiatry 27, 563-72 (1990). 161. Bliwise, D. L., Hughes, M., McMahon, P. M. & Kutner, N. Observed sleep/wakefulness and severity of dementia in an Alzheimer’s disease special care unit. J Gerontol A Biol Sci Med Sci 50, M303-6 (1995). 162. Ancoli-Israel, S. et al. Variations in circadian rhythms of activity, sleep, and light exposure related to dementia in nursing-home patients. Sleep 20, 18-23 (1997). 163. Cardinali, D. P., Brusco, L. I., Liberczuk, C. & Furio, A. M. REVIEW. The use of melatonin in Alzheimer’s disease. Neuroendocrinol Lett 23 Suppl 1, 20-3 (2002). 164. Moe, K. E., Vitiello, M. V., Larsen, L. H. & Prinz, P. N. Symposium: Cognitive processes and sleep disturbances: Sleep/wake patterns in Alzheimer’s disease: relationships with cognition and function. J Sleep Res 4, 15-20 (1995). 165. Pollak, C. P. & Perlick, D. Sleep problems and institutionalization of the elderly. J Geriatr Psy- chiatry Neurol 4, 204-10 (1991). 166. Giubilei, F. et al. Altered circadian cortisol secretion in Alzheimer’s disease: clinical and neu- roradiological aspects. J Neurosci Res 66, 262-5 (2001). 167. Mishima, K. et al. Melatonin secretion rhythm disorders in patients with senile dementia of Alzheimer’s type with disturbed sleep-waking. Biol Psychiatry 45, 417-21 (1999). 168. Van Someren, E. J. Circadian and sleep disturbances in the elderly. Exp Gerontol 35, 1229-37 (2000). 169. Hofman, M. A. & Swaab, D. F. Alterations in circadian rhythmicity of the vasopressin-producing neurons of the human suprachiasmatic nucleus (SCN) with aging. Brain Res 651, 134-42 (1994). 170. Hofman, M. A. & Swaab, D. F. Influence of aging on the seasonal rhythm of the vasopressin-expressing neurons in the human suprachiasmatic nucleus. Neurobiol Aging 16, 965-71 (1995). 171. Swaab, D. F. et al. Suprachiasmatic nucleus in aging, Alzheimer’s disease, transsexuality and Prader-Willi syndrome. Prog Brain Res 72, 301-10 (1987). 172. Zhou, J. N. & Swaab, D. F. Activation and degeneration during aging: a morphometric study of the human hypothalamus. Microsc Res Tech 44, 36-48 (1999). 173. Liu, R. Y. et al. Decreased vasopressin gene expression in the biological clock of Alzheimer disease patients with and without depression. J Neuropathol Exp Neurol 59, 314-22 (2000). 174. Stopa, E. G. et al. Pathologic evaluation of the human suprachiasmatic nucleus in severe dementia. J Neuropathol Exp Neurol 58, 29-39 (1999).

182 references

175. Swaab, D. F. et al. Tau and ubiquitin in the human hypothalamus in aging and Alzheimer’s disease. Brain Res 590, 239-49 (1992). 176. Van de Nes, J. A. et al. The monoclonal antibody Alz-50, used to reveal cytoskeletal changes in Alzheimer’s disease, also reacts with a large subpopulation of somatostatin neurons in the normal human hypothalamus and adjoining areas. Brain Res 655, 97-109 (1994). 177. Van de Nes, J. A., Kamphorst, W., Ravid, R. & Swaab, D. F. Comparison of beta-protein/A4 deposits and Alz-50-stained cytoskeletal changes in the hypothalamus and adjoining areas of Alzheimer’s disease patients: amorphic plaques and cytoskeletal changes occur independently. Acta Neuropathol (Berl) 96, 129-38 (1998). 178. Nygard, M. & Palomba, M. The GABAergic network in the suprachiasmatic nucleus as a key regulator of the biological clock: does it change during senescence? Chronobiol Int 23, 427-35 (2006). 179. Asai, M. et al. Circadian profile of Per gene mRNA expression in the suprachiasmatic nucleus, paraventricular nucleus, and pineal body of aged rats. J Neurosci Res 66, 1133-9 (2001). 180. Kolker, D. E. et al. Aging alters circadian and light-induced expression of clock genes in golden hamsters. J Biol Rhythms 18, 159-69 (2003). 181. Yamazaki, S. et al. Effects of aging on central and peripheral mammalian clocks. Proc Natl Acad Sci U S A 99, 10801-6 (2002). 182. Weinert, H., Weinert, D., Schurov, I., Maywood, E. S. & Hastings, M. H. Impaired expression of the mPer2 circadian clock gene in the suprachiasmatic nuclei of aging mice. Chronobiol Int 18, 559-65 (2001). 183. Oster, H., Baeriswyl, S., Van Der Horst, G. T. & Albrecht, U. Loss of circadian rhythmicity in aging mPer1-/-mCry2-/- mutant mice. Genes Dev 17, 1366-79 (2003). 184. Kolker, D. E., Vitaterna, M. H., Fruechte, E. M., Takahashi, J. S. & Turek, F. W. Effects of age on circadian rhythms are similar in wild-type and heterozygous Clock mutant mice. Neurobiol Aging 25, 517-23 (2004). 185. Kunieda, T. et al. Cellular senescence impairs circadian expression of clock genes in vitro and in vivo. Circ Res 98, 532-9 (2006). 186. Campbell, S. S., Kripke, D. F., Gillin, J. C. & Hrubovcak, J. C. Exposure to light in healthy elderly subjects and Alzheimer’s patients. Physiol Behav 42, 141-4 (1988). 187. Shochat, T., Martin, J., Marler, M. & Ancoli-Israel, S. Illumination levels in nursing home patients: effects on sleep and activity rhythms. J Sleep Res 9, 373-9 (2000). 188. Mishima, K., Okawa, M., Shimizu, T. & Hishikawa, Y. Diminished melatonin secretion in the elderly caused by insufficient environmental illumination. J Clin Endocrinol Metab 86, 129-34 (2001). 189. Van Someren, E. J. et al. Circadian rest-activity rhythm disturbances in Alzheimer’s disease. Biol Psychiatry 40, 259-70 (1996). 190. Sperduto, R. D., Hiller, R. & Seigel, D. Lens opacities and senile maculopathy. Arch Ophthalmol 99, 1004-8 (1981). 191. Hirvela, H. & Laatikainen, L. Visual acuity in a population aged 70 years or older; prevalence and causes of visual impairment. Acta Ophthalmol Scand 73, 99-104 (1995). 192. Klaver, C. C. et al. Is age-related maculopathy associated with Alzheimer’s Disease? The Rot- terdam Study. Am J Epidemiol 150, 963-8 (1999). 193. Hinton, D. R., Sadun, A. A., Blanks, J. C. & Miller, C. A. Optic-nerve degeneration in Alzheimer’s disease. N Engl J Med 315, 485-7 (1986). 194. Katz, B., Rimmer, S., Iragui, V. & Katzman, R. Abnormal pattern electroretinogram in Alzhe- imer’s disease: evidence for retinal ganglion cell degeneration? Ann Neurol 26, 221-5 (1989). 195. Trick, G. L., Barris, M. C. & Bickler-Bluth, M. Abnormal pattern electroretinograms in patients with senile dementia of the Alzheimer type. Ann Neurol 26, 226-31 (1989).

183 references

196. Blanks, J. C. et al. Retinal pathology in Alzheimer’s disease. II. Regional neuron loss and glial changes in GCL. Neurobiol Aging 17, 385-95 (1996). 197. Blanks, J. C., Torigoe, Y., Hinton, D. R. & Blanks, R. H. Retinal pathology in Alzheimer’s disease. I. Ganglion cell loss in foveal/parafoveal retina. Neurobiol Aging 17, 377-84 (1996). 198. Bayer, A. U. & Ferrari, F. Severe progression of glaucomatous optic neuropathy in patients with Alzheimer’s disease. Eye 16, 209-12 (2002). 199. Bayer, A. U., Ferrari, F. & Erb, C. High occurrence rate of glaucoma among patients with Alzhe- imer’s disease. Eur Neurol 47, 165-8 (2002). 200. Reppert, S. M., Weaver, D. R., Rivkees, S. A. & Stopa, E. G. Putative melatonin receptors in a human biological clock. Science 242, 78-81 (1988). 201. Von Gall, C., Stehle, J. H. & Weaver, D. R. Mammalian melatonin receptors: molecular biology and signal transduction. Cell Tissue Res 309, 151-62 (2002). 202. Skene, D. J. & Swaab, D. F. Melatonin rhythmicity: effect of age and Alzheimer’s disease. Exp Gerontol 38, 199-206 (2003). 203. Wu, Y. H. & Swaab, D. F. The human pineal gland and melatonin in aging and Alzheimer’s disease. J Pineal Res 38, 145-52 (2005). 204. Neylan, T. C. Treatment of sleep disturbances in depressed patients. J Clin Psychiatry 56 Suppl 2, 56-61 (1995). 205. Souetre, E. et al. Circadian rhythms in depression and recovery: evidence for blunted amplitude as the main chronobiological abnormality. Psychiatry Res 28, 263-78 (1989). 206. Wirz-Justice, A. in The Fourth Generation of Progress (ed. Bloom FE, K. D., eds) 999-1017 (Raven Press, New York, NY, 1995). 207. Harmatz, M. G. et al. Seasonal variation of depression and other moods: a longitudinal approach. J Biol Rhythms 15, 344-50 (2000). 208. Madden, P. A., Heath, A. C., Rosenthal, N. E. & Martin, N. G. Seasonal changes in mood and behavior. The role of genetic factors. Arch Gen Psychiatry 53, 47-55 (1996). 209. Weber, B. et al. Increased diurnal plasma concentrations of cortisone in depressed patients. J Clin Endocrinol Metab 85, 1133-6 (2000). 210. Zhou, J. N. et al. Alterations in arginine vasopressin neurons in the suprachiasmatic nucleus in depression. Arch Gen Psychiatry 58, 655-62 (2001). 211. Bernstein, H. G. et al. Further immunohistochemical evidence for impaired NO signaling in the hypothalamus of depressed patients. Ann N Y Acad Sci 973, 91-3 (2002). 212. Pacchierotti, C., Iapichino, S., Bossini, L., Pieraccini, F. & Castrogiovanni, P. Melatonin in psychiatric disorders: a review on the melatonin involvement in psychiatry. Front Neuroendocrinol 22, 18-32 (2001). 213. Fountoulakis, K. N. et al. Morning and evening plasma melatonin and dexamethasone suppression test in patients with nonseasonal major depressive disorder from northern Greece (latitude 40-41.5 degrees ). Neuropsychobiology 44, 113-7 (2001). 214. deVries, M. W. & Peeters, F. P. Melatonin as a therapeutic agent in the treatment of sleep disturbance in depression. J Nerv Ment Dis 185, 201-2 (1997). 215. Partonen, T. One pacemaker in seasonal affective disorder. Med Hypotheses 51, 297-8 (1998). 216. Shiino, Y., Nakajima, S., Ozeki, Y., Isono, T. & Yamada, N. Mutation screening of the human period 2 gene in bipolar disorder. Neurosci Lett 338, 82-4 (2003). 217. Desan, P. H. et al. Genetic polymorphism at the CLOCK gene locus and major depression. Am J Med Genet 96, 418-21 (2000). 218. Benedetti, F. et al. Influence of CLOCK gene polymorphism on circadian mood fluctuation and illness recurrence in bipolar depression. Am J Med Genet B Neuropsychiatr Genet 123, 23-6 (2003).

184 references

219. Johansson, C. et al. Circadian clock-related polymorphisms in seasonal aﬀective disorder and their relevance to diurnal preference. Neuropsychopharmacology 28, 734-9 (2003). 220. Nievergelt, C. M. et al. Suggestive evidence for association of the circadian genes PERIOD3 and ARNTL with bipolar disorder. Am J Med Genet B Neuropsychiatr Genet 141, 234-41 (2006). 221. Wagner, D. R. Disorders of the circadian sleep-wake cycle. Neurol Clin 14, 651-70 (1996). 222. Jones, C. R. et al. Familial advanced sleep-phase syndrome: A short-period circadian rhythm variant in humans. Nat Med 5, 1062-5 (1999). 223. Toh, K. L. et al. An hPer2 phosphorylation site mutation in familial advanced sleep phase syndrome. Science 291, 1040-3 (2001). 224. Zheng, B. et al. The mPer2 gene encodes a functional component of the mammalian circadian clock. Nature 400, 169-73 (1999). 225. Lowrey, P. L. et al. Positional syntenic cloning and functional characterization of the mammalian circadian mutation tau. Science 288, 483-92 (2000). 226. Xu, Y. et al. Functional consequences of a CKIdelta mutation causing familial advanced sleep phase syndrome. Nature 434, 640-4 (2005). 227. Carpen, J. D., Archer, S. N., Skene, D. J., Smits, M. & Von Schantz, M. A single-nucleotide polymorphism in the 5’-untranslated region of the hPER2 gene is associated with diurnal preference. J Sleep Res 14, 293-7 (2005). 228. Satoh, K., Mishima, K., Inoue, Y., Ebisawa, T. & Shimizu, T. Two pedigrees of familial advanced sleep phase syndrome in Japan. Sleep 26, 416-7 (2003). 229. Ebisawa, T. et al. Association of structural polymorphisms in the human period3 gene with delayed sleep phase syndrome. EMBO Rep 2, 342-6 (2001). 230. Archer, S. N. et al. A length polymorphism in the circadian clock gene Per3 is linked to delayed sleep phase syndrome and extreme diurnal preference. Sleep 26, 413-5 (2003). 231. Iwase, T. et al. Mutation screening of the human Clock gene in circadian rhythm sleep disorders. Psychiatry Res 109, 121-8 (2002). 232. Kennaway, D. J. & Van Dorp, C. F. Free-running rhythms of melatonin, cortisol, electrolytes, and sleep in humans in Antarctica. Am J Physiol 260, R1137-44 (1991). 233. Boivin, D. B., James, F. O., Santo, J. B., Caliyurt, O. & Chalk, C. Non-24-hour sleep-wake syndrome following a car accident. Neurology 60, 1841-3 (2003). 234. Boivin, D. B., Caliyurt, O., James, F. O. & Chalk, C. Association between delayed sleep phase and hypernyctohemeral syndromes: a case study. Sleep 27, 417-21 (2004). 235. Ebisawa, T. et al. Alleic variants of human melatonin 1a receptor: function and prevalence in subjects with circadian rhythm sleep disorders. Biochem Biophys Res Commun 262, 832-7 (1999). 236. Ebisawa, T. et al. Genetic polymorphisms of human melatonin 1b receptor gene in circadian rhythm sleep disorders and controls. Neurosci Lett 280, 29-32 (2000). 237. Hannibal, J. et al. Dissociation between light-induced phase shift of the circadian rhythm and clock gene expression in mice lacking the pituitary adenylate cyclase activating polypeptide type 1 receptor. J Neurosci 21, 4883-90 (2001). 238. Furukawa, T., Morrow, E. M., Li, T., Davis, F. C. & Cepko, C. L. Retinopathy and attenuated circadian entrainment in Crx-deﬁcient mice. Nat Genet 23, 466-70 (1999). 239. Panda, S. et al. Melanopsin (Opn4) requirement for normal light-induced circadian phase shifting. Science 298, 2213-6 (2002). 240. Panda, S. et al. Melanopsin is required for non-image-forming photic responses in blind mice. Science 301, 525-7 (2003). 241. Harmar, A. J. et al. The VPAC(2) receptor is essential for circadian function in the mouse suprachiasmatic nuclei. Cell 109, 497-508 (2002). 242. Shen, H. et al. Role of neural cell adhesion molecule and polysialic acid in mouse circadian clock function. J Neurosci 17, 5221-9 (1997).

185 references

243. Shima, T. et al. Changes in circadian period and morphology of the hypothalamic suprachiasmatic nucleus in fyn kinase-deficient mice. Brain Res 870, 36-43 (2000). 244. Frederic, F., Chianale, C., Oliver, C. & Mariani, J. Enhanced endocrine response to novel environment stress and lack of corticosterone circadian rhythm in staggerer (Rora sg/sg) mutant mice. J Neurosci Res 83, 1525-32 (2006). 245. Allen, G. C., Qu, X. & Earnest, D. J. TrkB-deficient mice show diminished phase shifts of the circadian activity rhythm in response to light. Neurosci Lett 378, 150-5 (2005). 246. Kapfhamer, D. et al. Mutations in Rab3a alter circadian period and homeostatic response to sleep loss in the mouse. Nat Genet 32, 290-5 (2002). 247. Reppert, S. M. & Weaver, D. R. Coordination of circadian timing in mammals. Nature 418, 935-41 (2002). 248. Hasegawa, A., Ohtsubo, K. & Mori, W. Pineal gland in old age; quantitative and qualitative morphological study of 168 human autopsy cases. Brain Res 409, 343-9 (1987). 249. Swaab, D. F. The human hypothalamus basic and clinical aspects (Elsevier, Amsterdam, 2003). 250. Arendt, J. Melatonin and the mammalian pineal gland (Chapman&Hall, London, 1995). 251. Reiter, R. J. Pineal melatonin: cell biology of its synthesis and of its physiological interactions. Endocr Rev 12, 151-80 (1991). 252. Klein, D. C. et al. The melatonin rhythm-generating enzyme: molecular regulation of serotonin N-acetyltransferase in the pineal gland. Recent Prog Horm Res 52, 307-57; discussion 357-8 (1997). 253. Coon, S. L., Del Olmo, E., Young, W. S., 3rd & Klein, D. C. Melatonin synthesis enzymes in Macaca mulatta: focus on arylalkylamine N-acetyltransferase (EC 2.3.1.87). J Clin Endocrinol Metab 87, 4699-706 (2002). 254. Wu, Y. H. et al. Molecular changes underlying reduced pineal melatonin levels in Alzheimer disease: alterations in preclinical and clinical stages. J Clin Endocrinol Metab 88, 5898-906 (2003). 255. Teclemariam-Mesbah, R., Ter Horst, G. J., Postema, F., Wortel, J. & Buijs, R. M. Anatomical demonstration of the suprachiasmatic nucleus-pineal pathway. J Comp Neurol 406, 171-82 (1999). 256. Lockley, S. W. et al. Relationship between napping and melatonin in the blind. J Biol Rhythms 12, 16-25 (1997). 257. Skene, D. J., Lockley, S. W., James, K. & Arendt, J. Correlation between urinary cortisol and 6-sulphatoxymelatonin rhythms in field studies of blind subjects. Clin Endocrinol (Oxf) 50, 715-9 (1999). 258. Karasek, M. et al. Melatonin concentrations in patients with large goiter before and after surgery. Neuroendocrinol Lett 21, 437-439 (2000). 259. Bruce, J., Tamarkin, L., Riedel, C., Markey, S. & Oldfield, E. Sequential cerebrospinal fluid and plasma sampling in humans: 24-hour melatonin measurements in normal subjects and after peripheral sympathectomy. J Clin Endocrinol Metab 72, 819-23 (1991). 260. Oxenkrug, G. F., Anderson, G. F., Dragovic, L., Blaivas, M. & Riederer, P. Circadian rhythms of human pineal melatonin, related indoles, and beta adrenoreceptors: post-mortem evaluation. J Pineal Res 9, 1-11 (1990). 261. Mayeda, A. et al. Effects of indirect light and propranolol on melatonin levels in normal human subjects. Psychiatry Res 81, 9-17 (1998). 262. Stoschitzky, K. et al. Influence of beta-blockers on melatonin release. Eur J Clin Pharmacol 55, 111-5 (1999). 263. Skene, D. J. Optimization of light and melatonin to phase-shift human circadian rhythms. J Neuroendocrinol 15, 438-41 (2003).

186 references

264. Fischer, S., Smolnik, R., Herms, M., Born, J. & Fehm, H. L. Melatonin acutely improves the neuroendocrine architecture of sleep in blind individuals. J Clin Endocrinol Metab 88, 5315-20 (2003). 265. Allegra, M. et al. The chemistry of melatonin’s interaction with reactive species. J Pineal Res 34, 1-10 (2003). 266. Rodriguez, C. et al. Regulation of antioxidant enzymes: a significant role for melatonin. J Pineal Res 36, 1-9 (2004). 267. Pappolla, M. A. et al. An assessment of the antioxidant and the antiamyloidogenic properties of melatonin: implications for Alzheimer’s disease. J Neural Transm 107, 203-31 (2000). 268. Reiter, R. J., Tan, D. X., Manchester, L. C. & El-Sawi, M. R. Melatonin reduces oxidant damage and promotes mitochondrial respiration: implications for aging. Ann N Y Acad Sci 959, 238-50 (2002). 269. Czeisler, C. A. et al. Stability, precision, and near-24-hour period of the human circadian pacemaker. Science 284, 2177-81 (1999). 270. Mailloux, A., Benstaali, C., Bogdan, A., Auzeby, A. & Touitou, Y. Body temperature and locomotor activity as marker rhythms of aging of the circadian system in rodents. Exp Gerontol 34, 733-40 (1999). 271. Duffy, J. F., Dijk, D. J., Klerman, E. B. & Czeisler, C. A. Later endogenous circadian temperature nadir relative to an earlier wake time in older people. Am J Physiol 275, R1478-87 (1998). 272. Dijk, D. J., Duffy, J. F. & Czeisler, C. A. Contribution of circadian physiology and sleep homeostasis to age-related changes in human sleep. Chronobiol Int 17, 285-311 (2000). 273. Luboshitzky, R. et al. Daily and seasonal variations in the concentration of melatonin in the human pineal gland. Brain Res Bull 47, 271-6 (1998). 274. Zeitzer, J. M. et al. Do plasma melatonin concentrations decline with age? Am J Med 107, 432-6 (1999). 275. Fourtillan, J. B. et al. Melatonin secretion occurs at a constant rate in both young and older men and women. Am J Physiol Endocrinol Metab 280, E11-22 (2001). 276. Kennaway, D. J. et al. Urinary 6-sulfatoxymelatonin excretion and aging: new results and a critical review of the literature. J Pineal Res 27, 210-20 (1999). 277. Zhao, Z. Y., Xie, Y., Fu, Y. R., Bogdan, A. & Touitou, Y. Aging and the circadian rhythm of melatonin: a cross-sectional study of Chinese subjects 30-110 yr of age. Chronobiol Int 19, 1171-82 (2002). 278. Bergiannaki, J. D., Soldatos, C. R., Paparrigopoulos, T. J., Syrengelas, M. & Stefanis, C. N. Low and high melatonin excretors among healthy individuals. J Pineal Res 18, 159-64 (1995). 279. Kunz, D. et al. A new concept for melatonin deficit: on pineal calcification and melatonin excretion. Neuropsychopharmacology 21, 765-72 (1999). 280. Gusek, W. [Histology of the pineal gland in the elderly human]. Aktuelle Gerontol 13, 111-4 (1983). 281. Kunz, D., Bes, F., Schlattmann, P. & Herrmann, W. M. On pineal calcification and its relation to subjective sleep perception: a hypothesis-driven pilot study. Psychiatry Res 82, 187-91 (1998). 282. Schmid, H. A. Decreased melatonin biosynthesis, calcium flux, pineal gland calcification and aging: a hypothetical framework. Gerontology 39, 189-99 (1993). 283. Bojkowski, C. J. & Arendt, J. Factors influencing urinary 6-sulphatoxymelatonin, a major melatonin metabolite, in normal human subjects. Clin Endocrinol (Oxf) 33, 435-44 (1990). 284. Otani, T., Gyorkey, F. & Farrell, G. Enzymes of the human pineal body. J Clin Endocrinol Metab 28, 349-54 (1968). 285. Tapp, E. & Huxley, M. The histological appearance of the human pineal gland from puberty to old age. J Pathol 108, 137-44 (1972).

187 references

286. Pardo, C. A., Martin, L. J., Troncoso, J. C. & Price, D. L. The human pineal gland in aging and Alzheimer’s disease: patterns of cytoskeletal antigen immunoreactivity. Acta Neuropathol (Berl) 80, 535-40 (1990). 287. Arieti, S. The pineal gland in old age. J Neuropathol Exp Neurol 13, 482-91 (1954). 288. Jengeleski, C. A., Powers, R. E., O’Connor, D. T. & Price, D. L. Noradrenergic innervation of human pineal gland: abnormalities in aging and Alzheimer’s disease. Brain Res 481, 378-82 (1989). 289. Hofman, M. A. The human circadian clock and aging. Chronobiol Int 17, 245-59 (2000). 290. Charman, W. N. Age, lens transmittance, and the possible effects of light on melatonin suppression. Ophthalmic Physiol Opt 23, 181-7 (2003). 291. Klerman, E. B., Duffy, J. F., Dijk, D. J. & Czeisler, C. A. Circadian phase resetting in older people by ocular bright light exposure. J Investig Med 49, 30-40 (2001). 292. Werth, E. et al. Decline in long-term circadian rest-activity cycle organization in a patient with dementia. J Geriatr Psychiatry Neurol 15, 55-9 (2002). 293. Martin, J., Marler, M., Shochat, T. & Ancoli-Israel, S. Circadian rhythms of agitation in institutionalized patients with Alzheimer’s disease. Chronobiol Int 17, 405-18 (2000). 294. Van Someren, E. J., Mirmiran, M. & Swaab, D. F. Non-pharmacological treatment of sleep and wake disturbances in aging and Alzheimer’s disease: chronobiological perspectives. Behav Brain Res 57, 235-53 (1993). 295. Burney-Puckett, M. Sundown syndrome: etiology and management. J Psychosoc Nurs Ment Health Serv 34, 40-3 (1996). 296. Wirz-Justice, A. et al. Haloperidol disrupts, clozapine reinstates the circadian rest-activity cycle in a patient with early-onset Alzheimer disease. Alzheimer Dis Assoc Disord 14, 212-5 (2000). 297. Ferrari, E. et al. Pineal and pituitary-adrenocortical function in physiological aging and in senile dementia. Exp Gerontol 35, 1239-50 (2000). 298. Ohashi, Y. et al. Daily rhythm of serum melatonin levels and effect of light exposure in patients with dementia of the Alzheimer’s type. Biol Psychiatry 45, 1646-52 (1999). 299. Skene, D. J. et al. Daily variation in the concentration of melatonin and 5-methoxytryptophol in the human pineal gland: effect of age and Alzheimer’s disease. Brain Res 528, 170-4 (1990). 300. Uchida, K., Okamoto, N., Ohara, K. & Morita, Y. Daily rhythm of serum melatonin in patients with dementia of the degenerate type. Brain Res 717, 154-9 (1996). 301. Liu, R. Y., Zhou, J. N., Van Heerikhuize, J., Hofman, M. A. & Swaab, D. F. Decreased melatonin levels in postmortem cerebrospinal fluid in relation to aging, Alzheimer’s disease, and apolipoprotein E-epsilon4/4 genotype. J Clin Endocrinol Metab 84, 323-7 (1999). 302. Zhou, J. N., Liu, R. Y., Kamphorst, W., Hofman, M. A. & Swaab, D. F. Early neuropathological Alzheimer’s changes in aged individuals are accompanied by decreased cerebrospinal fluid melatonin levels. J Pineal Res 35, 125-130 (2003). 303. Rousseau, A., Petren, S., Plannthin, J., Eklundh, T. & Nordin, C. Serum and cerebrospinal fluid concentrations of melatonin: a pilot study in healthy male volunteers. J Neural Transm 106, 883-8 (1999). 304. Matsubara, E. et al. Melatonin increases survival and inhibits oxidative and amyloid pathology in a transgenic model of Alzheimer’s disease. J Neurochem 85, 1101-1108 (2003). 305. Friedland, R. P., Luxenberg, J. S. & Koss, E. A quantitative study of intracranial calcification in dementia of the Alzheimer type. Int Psychogeriatr 2, 36-43 (1990). 306. Sparks, D. L., Woeltz, V. M. & Markesbery, W. R. Alterations in brain monoamine oxidase activity in aging, Alzheimer’s disease, and Pick’s disease. Arch Neurol 48, 718-21 (1991). 307. Sherif, F., Gottfries, C. G., Alafuzoff, I. & Oreland, L. Brain gamma-aminobutyrate aminotrans- ferase (GABA-T) and monoamine oxidase (MAO) in patients with Alzheimer’s disease. J Neural Transm Park Dis Dement Sect 4, 227-40 (1992).

188 references

308. Emilsson, L. et al. Increased monoamine oxidase messenger RNA expression levels in frontal cortex of Alzheimer’s disease patients. Neurosci Lett 326, 56-60 (2002). 309. Takehashi, M., Tanaka, S., Masliah, E. & Ueda, K. Association of monoamine oxidase A gene polymorphism with Alzheimer’s disease and Lewy body variant. Neurosci Lett 327, 79-82 (2002). 310. Zhou, J. N., Hofman, M. A. & Swaab, D. F. VIP neurons in the human SCN in relation to sex, age, and Alzheimer’s disease. Neurobiol Aging 16, 571-6 (1995). 311. Van de Nes, J. A., Kamphorst, W., Ravid, R. & Swaab, D. F. The distribution of Alz-50 immunoreactivity in the hypothalamus and adjoining areas of Alzheimer’s disease patients. Brain 116 (Pt 1), 103-15 (1993). 312. Curcio, C. A. & Drucker, D. N. Retinal ganglion cells in Alzheimer’s disease and aging. Ann Neurol 33, 248-57 (1993). 313. Davies, D. C., McCoubrie, P., McDonald, B. & Jobst, K. A. Myelinated axon number in the optic nerve is unaffected by Alzheimer’s disease. Br J Ophthalmol 79, 596-600 (1995). 314. Justino, L. et al. Neuroretinal function is normal in early dementia of the Alzheimer type. Neurobiol Aging 22, 691-5 (2001). 315. Cohen-Mansfield, J., Garfinkel, D. & Lipson, S. Melatonin for treatment of sundowning in elderly persons with dementia - a preliminary study. Arch Gerontol Geriatr 31, 65-76 (2000). 316. Brusco, L. I., Marquez, M. & Cardinali, D. P. Melatonin treatment stabilizes chronobiologic and cognitive symptoms in Alzheimer’s disease. Neuroendocrinol Lett 21, 39-42 (2000). 317. Brusco, L. I., Marquez, M. & Cardinali, D. P. Monozygotic twins with Alzheimer’s disease treated with melatonin: Case report. J Pineal Res 25, 260-3 (1998). 318. Jean-Louis, G., Von Gizycki, H. & Zizi, F. Melatonin effects on sleep, mood, and cognition in elderly with mild cognitive impairment. J Pineal Res 25, 177-83 (1998). 319. Singer, C. et al. A multicenter, placebo-controlled trial of melatonin for sleep disturbance in Alzheimer’s disease. Sleep 26, 893-901 (2003). 320. Serfaty, M., Kennell-Webb, S., Warner, J., Blizard, R. & Raven, P. Double blind randomised placebo controlled trial of low dose melatonin for sleep disorders in dementia. Int J Geriatr Psychiatry 17, 1120-7 (2002). 321. Kobayashi, R. et al. Effects of morning bright light on sleep in healthy elderly women. Psychiatry Clin Neurosci 53, 237-8 (1999). 322. Van Someren, E. J., Kessler, A., Mirmiran, M. & Swaab, D. F. Indirect bright light improves circadian rest-activity rhythm disturbances in demented patients. Biol Psychiatry 41, 955-63 (1997). 323. Yamadera, H. et al. Effects of bright light on cognitive and sleep-wake (circadian) rhythm disturbances in Alzheimer-type dementia. Psychiatry Clin Neurosci 54, 352-3 (2000). 324. Graf, A. et al. The effects of light therapy on mini-mental state examination scores in demented patients. Biol Psychiatry 50, 725-7 (2001). 325. Mishima, K., Hishikawa, Y. & Okawa, M. Randomized, dim light controlled, crossover test of morning bright light therapy for rest-activity rhythm disorders in patients with vascular dementia and dementia of Alzheimer’s type. Chronobiol Int 15, 647-54 (1998). 326. Van Someren, E. J., Lijzenga, C., Mirmiran, M. & Swaab, D. F. Long-term fitness training improves the circadian rest-activity rhythm in healthy elderly males. J Biol Rhythms 12, 146-56 (1997). 327. Van Someren, E. J., Scherder, E. J. & Swaab, D. F. Transcutaneous electrical nerve stimulation (TENS) improves circadian rhythm disturbances in Alzheimer disease. Alzheimer Dis Assoc Disord 12, 114-8 (1998). 328. Scherder, E. J., Van Someren, E. J. & Swaab, D. F. Transcutaneous electrical nerve stimulation (TENS) improves the rest-activity rhythm in midstage Alzheimer’s disease. Behav Brain Res 101, 105-7 (1999).

189 references

329. Malpaux, B., Migaud, M., Tricoire, H. & Chemineau, P. Biology of mammalian photoperiodism and the critical role of the pineal gland and melatonin. J Biol Rhythms 16, 336-47 (2001). 330. Reiter, R. J. Functional pleiotropy of the neurohormone melatonin: antioxidant protection and neuroendocrine regulation. Front Neuroendocrinol 16, 383-415 (1995). 331. Pang, S. F. et al. Neuroendocrinology of melatonin in reproduction: recent developments. J Chem Neuroanat 14, 157-66 (1998). 332. Wetterberg, L. Melatonin and clinical application. Reprod Nutr Dev 39, 367-82 (1999). 333. Krauchi, K. & Wirz-Justice, A. Circadian clues to sleep onset mechanisms. Neuropsychophar- macology 25, S92-6 (2001). 334. Satoh, K. & Mishima, K. Hypothermic action of exogenously administered melatonin is dose- dependent in humans. Clin Neuropharmacol 24, 334-40 (2001). 335. Vakkuri, O., Lamsa, E., Rahkamaa, E., Ruotsalainen, H. & Leppaluoto, J. Iodinated melatonin: preparation and characterization of the molecular structure by mass and 1H NMR spectroscopy. Anal Biochem 142, 284-9 (1984). 336. Vakkuri, O., Leppaluoto, J. & Vuolteenaho, O. Development and validation of a melatonin radioimmunoassay using radioiodinated melatonin as tracer. Acta Endocrinol (Copenh) 106, 152-7 (1984). 337. Vanecek, J., Pavlik, A. & Illnerova, H. Hypothalamic melatonin receptor sites revealed by autoradiography. Brain Res 435, 359-62 (1987). 338. Weaver, D. R., Namboodiri, M. A. & Reppert, S. M. Iodinated melatonin mimics melatonin action and reveals discrete binding sites in fetal brain. FEBS Lett 228, 123-7 (1988). 339. Weaver, D. R., Rivkees, S. A. & Reppert, S. M. Localization and characterization of melatonin receptors in rodent brain by in vitro autoradiography. J Neurosci 9, 2581-90 (1989). 340. Duncan, M. J., Takahashi, J. S. & Dubocovich, M. L. Characteristics and autoradiographic localization of 2-[125I]iodomelatonin binding sites in Djungarian hamster brain. Endocrinology 125, 1011-8 (1989). 341. Stankov, B., Fraschini, F. & Reiter, R. J. Melatonin binding sites in the central nervous system. Brain Res Brain Res Rev 16, 245-56 (1991). 342. Weaver, D. R., Stehle, J. H., Stopa, E. G. & Reppert, S. M. Melatonin receptors in human hypothalamus and pituitary: implications for circadian and reproductive responses to melatonin. J Clin Endocrinol Metab 76, 295-301 (1993). 343. Morgan, P. J., Barrett, P., Howell, H. E. & Helliwell, R. Melatonin receptors: localization, molecular pharmacology and physiological signiﬁcance. Neurochem Int 24, 101-46 (1994). 344. Morgan, P. J. & Williams, L. M. The pars tuberalis of the pituitary: a gateway for neuroendocrine output. Rev Reprod 1, 153-61 (1996). 345. Ross, A. W. & Morgan, P. J. The pars tuberalis as a target of the central clock. Cell Tissue Res 309, 163-71 (2002). 346. Yuan, H., Lu, Y. & Pang, S. F. Binding characteristics and regional distribution of [125I]iodomelatonin binding sites in the brain of the human fetus. Neurosci Lett 130, 229-32 (1991). 347. Thomas, L., Purvis, C. C., Drew, J. E., Abramovich, D. R. & Williams, L. M. Melatonin receptors in human fetal brain: 2-[(125)I]iodomelatonin binding and MT1 gene expression. J Pineal Res 33, 218-24 (2002). 348. Lopez-Gonzalez, M. A., Calvo, J. R., Rubio, A., Goberna, R. & Guerrero, J. M. Characterization of melatonin binding sites in the harderian gland and median eminence of the rat. Life Sci 48, 1165-71 (1991). 349. Coto-Montes, A. M., Rodriguez-Colunga, M. J., Tolivia, D. & Menendez-Pelaez, A. Immuno- cytochemical localization of melatonin in the harderian gland of Syrian hamster. Anat Rec 245, 13-6 (1996).

190 references

350. Song, Y., Poon, A. M., Brown, G. M. & Pang, S. F. Ontogeny of 2-[125I]iodomelatonin binding sites in the chicken (Gallus domesticus) kidney and spleen. Can J Physiol Pharmacol 73, 685-92 (1995). 351. Pang, C. S., Tsang, K. F., Brown, G. M. & Pang, S. F. Specific 2-[125I]iodomelatonin binding sites in the duck adrenal gland. Neurosci Lett 165, 55-8 (1994). 352. Song, Y., Tam, P. C., Poon, A. M., Brown, G. M. & Pang, S. F. 2-[125I]iodomelatonin-binding sites in the human kidney and the effect of guanosine 5’-O-(3-thiotriphosphate). J Clin Endo- crinol Metab 80, 1560-5 (1995). 353. Viswanathan, M., Laitinen, J. T. & Saavedra, J. M. Expression of melatonin receptors in arteries involved in thermoregulation. Proc Natl Acad Sci U S A 87, 6200-3 (1990). 354. Van Vuuren, R. J., Pitout, M. J., Van Aswegen, C. H. & Theron, J. J. Putative melatonin receptor in human spermatozoa. Clin Biochem 25, 125-7 (1992). 355. Lopez-Gonzalez, M. A., Calvo, J. R., Osuna, C., Rubio, A. & Guerrero, J. M. Synergistic action of melatonin and vasoactive intestinal peptide in stimulating cyclic AMP production in human lymphocytes. J Pineal Res 12, 174-80 (1992). 356. Maestroni, G. J. T-helper-2 lymphocytes as a peripheral target of melatonin. J Pineal Res 18, 84-9 (1995). 357. Dubocovich, M. L. Pharmacology and function of melatonin receptors. Faseb J 2, 2765-73 (1988). 358. Reppert, S. M. et al. Molecular characterization of a second melatonin receptor expressed in human retina and brain: the Mel1b melatonin receptor. Proc Natl Acad Sci U S A 92, 8734-8 (1995). 359. Reppert, S. M., Weaver, D. R., Cassone, V. M., Godson, C. & Kolakowski, L. F., Jr. Melatonin receptors are for the birds: molecular analysis of two receptor subtypes differentially expressed in chick brain. Neuron 15, 1003-15 (1995). 360. Dubocovich, M. et al. in The IUPHAR compendium of receptor characterisation and classification. 187-193 (IUPHAR Media, London, 1998). 361. Nosjean, O. et al. Identification of the melatonin-binding site MT3 as the quinone reductase 2. J Biol Chem 275, 31311-7 (2000). 362. Acuna-Castroviejo, D., Reiter, R. J., Menendez-Pelaez, A., Pablos, M. I. & Burgos, A. Charac- terization of high-affinity melatonin binding sites in purified cell nuclei of rat liver. J Pineal Res 16, 100-12 (1994). 363. Steinhilber, D. et al. The nuclear receptor for melatonin represses 5-lipoxygenase gene expression in human B lymphocytes. J Biol Chem 270, 7037-40 (1995). 364. Becker-Andre, M. et al. Pineal gland hormone melatonin binds and activates an orphan of the nuclear receptor superfamily. J Biol Chem 269, 28531-4 (1994). 365. Reppert, S. M., Weaver, D. R., Ebisawa, T., Mahle, C. D. & Kolakowski, L. F., Jr. Cloning of a melatonin-related receptor from human pituitary. FEBS Lett 386, 219-24 (1996). 366. Drew, J. E. et al. Localization of the melatonin-related receptor in the rodent brain and peripheral tissues. J Neuroendocrinol 13, 453-8 (2001). 367. Poirel, V. J. et al. MT1 melatonin receptor mRNA tissular localization by PCR amplification. Neuroendocrinol Lett 24, 33-8 (2003). 368. Sallinen, P., Saarela, S., Ilves, M., Vakkuri, O. & Leppaluoto, J. The expression of MT1 and MT2 melatonin receptor mRNA in several rat tissues. Life Sci 76, 1123-34 (2005). 369. Mazzucchelli, C. et al. The melatonin receptor in the human brain: cloning experiments and distribution studies. Brain Res Mol Brain Res 39, 117-26 (1996). 370. Von Gall, C. et al. Rhythmic gene expression in pituitary depends on heterologous sensitization by the neurohormone melatonin. Nat Neurosci 5, 234-8 (2002). 371. Jilg, A. et al. Rhythms in clock proteins in the mouse pars tuberalis depend on MT1 melatonin receptor signalling. Eur J Neurosci 22, 2845-2854 (2005).

191 references

372. Naji, L., Carrillo-Vico, A., Guerrero, J. M. & Calvo, J. R. Expression of membrane and nuclear melatonin receptors in mouse peripheral organs. Life Sci 74, 2227-36 (2004). 373. Dillon, D. C. et al. Differential expression of high-affinity melatonin receptors (MT1) in normal and malignant human breast tissue. Am J Clin Pathol 118, 451-8 (2002). 374. Meyer, P. et al. Melatonin MT-1-receptor immunoreactivity in the human eye. Br J Ophthalmol 86, 1053-7 (2002). 375. Scher, J., Wankiewicz, E., Brown, G. M. & Fujieda, H. AII amacrine cells express the MT1 melatonin receptor in human and macaque retina. Exp Eye Res 77, 375-82 (2003). 376. Kobayashi, H. et al. A role of melatonin in neuroectodermal-mesodermal interactions: the hair follicle synthesizes melatonin and expresses functional melatonin receptors. Faseb J 19, 1710-2 (2005). 377. Ekmekcioglu, C. et al. 24h variation in the expression of the mt1 melatonin receptor subtype in coronary arteries derived from patients with coronary heart disease. Chronobiol Int 18, 973-85 (2001). 378. Brydon, L. et al. Dual signaling of human Mel1a melatonin receptors via G(i2), G(i3), and G(q/11) proteins. Mol Endocrinol 13, 2025-38 (1999). 379. Chan, A. S. et al. Melatonin mt1 and MT2 receptors stimulate c-Jun N-terminal kinase via pertussis toxin-sensitive and -insensitive G proteins. Cell Signal 14, 249-57 (2002). 380. Ho, M. K. et al. Galpha(14) links a variety of G(i)- and G(s)-coupled receptors to the stimulation of phospholipase C. Br J Pharmacol 132, 1431-40 (2001). 381. Morgan, P. J. et al. Intracellular signalling in the ovine pars tuberalis: an investigation using aluminium fluoride and melatonin. J Mol Endocrinol 7, 137-44 (1991). 382. Niles, L. P. & Hashemi, F. Picomolar-affinity binding and inhibition of adenylate cyclase activity by melatonin in Syrian hamster hypothalamus. Cell Mol Neurobiol 10, 553-8 (1990). 383. Witt-Enderby, P. A., Masana, M. I. & Dubocovich, M. L. Physiological exposure to melatonin su- persensitizes the cyclic adenosine 3’,5’-monophosphate-dependent signal transduction cascade in Chinese hamster ovary cells expressing the human mt1 melatonin receptor. Endocrinology 139, 3064-71 (1998). 384. Morgan, P. J., Barrett, P., Davidson, G., Lawson, W. & Hazlerigg, D. p72, a marker protein for melatonin action in ovine pars tuberalis cells: its regulation by protein kinase A and protein kinase C and differential secretion relative to prolactin. Neuroendocrinology 59, 325-35 (1994). 385. McNulty, S., Ross, A. W., Barrett, P., Hastings, M. H. & Morgan, P. J. Melatonin regulates the phosphorylation of CREB in ovine pars tuberalis. J Neuroendocrinol 6, 523-32 (1994). 386. Godson, C. & Reppert, S. M. The Mel1a melatonin receptor is coupled to parallel signal transduction pathways. Endocrinology 138, 397-404 (1997). 387. Roka, F. et al. Tight association of the human Mel(1a)-melatonin receptor and G(i): precoupling and constitutive activity. Mol Pharmacol 56, 1014-24 (1999). 388. Witt-Enderby, P. A. et al. Melatonin induction of filamentous structures in non-neuronal cells that is dependent on expression of the human mt1 melatonin receptor. Cell Motil Cytoskeleton 46, 28-42 (2000). 389. Wan, Q. et al. Differential modulation of GABAA receptor function by Mel1a and Mel1b receptors. Nat Neurosci 2, 401-3 (1999). 390. Scher, J., Wankiewicz, E., Brown, G. M. & Fujieda, H. MT(1) melatonin receptor in the human retina: expression and localization. Invest Ophthalmol Vis Sci 43, 889-97 (2002). 391. Savaskan, E. et al. Reduced hippocampal MT2 melatonin receptor expression in Alzheimer’s disease. J Pineal Res 38, 10-6 (2005). 392. Ekmekcioglu, C. et al. The melatonin receptor subtype MT2 is present in the human cardiovascular system. J Pineal Res 35, 40-4 (2003).

192 references

393. Dubocovich, M. L., Masana, M. I., Iacob, S. & Sauri, D. M. Melatonin receptor antagonists that differentiate between the human Mel1a and Mel1b recombinant subtypes are used to assess the pharmacological profile of the rabbit retina ML1 presynaptic heteroreceptor. Naunyn Schmiedebergs Arch Pharmacol 355, 365-75 (1997). 394. Doolen, S., Krause, D. N., Dubocovich, M. L. & Duckles, S. P. Melatonin mediates two distinct responses in vascular smooth muscle. Eur J Pharmacol 345, 67-9 (1998). 395. Lotufo, C. M., Lopes, C., Dubocovich, M. L., Farsky, S. H. & Markus, R. P. Melatonin and N- acetylserotonin inhibit leukocyte rolling and adhesion to rat microcirculation. Eur J Pharmacol 430, 351-7 (2001). 396. Wang, L. M., Suthana, N. A., Chaudhury, D., Weaver, D. R. & Colwell, C. S. Melatonin inhibits hippocampal long-term potentiation. Eur J Neurosci 22, 2231-7 (2005). 397. Nosjean, O. et al. Comparative pharmacological studies of melatonin receptors: MT1, MT2 and MT3/QR2. Tissue distribution of MT3/QR2. Biochem Pharmacol 61, 1369-79 (2001). 398. Pintor, J., Martin, L., Pelaez, T., Hoyle, C. H. & Peral, A. Involvement of melatonin MT(3) receptors in the regulation of intraocular pressure in rabbits. Eur J Pharmacol 416, 251-4 (2001). 399. Masana, M. I. & Dubocovich, M. L. Melatonin receptor signaling: finding the path through the dark. Sci STKE 2001, PE39 (2001). 400. Gauer, F., Masson-Pevet, M. & Pevet, P. Melatonin receptor density is regulated in rat pars tuberalis and suprachiasmatic nuclei by melatonin itself. Brain Res 602, 153-6 (1993). 401. Gauer, F., Masson-Pevet, M., Stehle, J. & Pevet, P. Daily variations in melatonin receptor density of rat pars tuberalis and suprachiasmatic nuclei are distinctly regulated. Brain Res 641, 92-8 (1994). 402. Gauer, F., Masson-Pevet, M., Skene, D. J., Vivien-Roels, B. & Pevet, P. Daily rhythms of melatonin binding sites in the rat pars tuberalis and suprachiasmatic nuclei; evidence for a regulation of melatonin receptors by melatonin itself. Neuroendocrinology 57, 120-6 (1993). 403. Schuster, C. et al. The circadian clock, light/dark cycle and melatonin are differentially involved in the expression of daily and photoperiodic variations in mt(1) melatonin receptors in the Siberian and Syrian hamsters. Neuroendocrinology 74, 55-68 (2001). 404. Gillespie, J. M., Chan, B. P., Roy, D., Cai, F. & Belsham, D. D. Expression of circadian rhythm genes in gonadotropin-releasing hormone-secreting GT1-7 neurons. Endocrinology 144, 5285- 92 (2003). 405. Barrett, P., MacLean, A., Davidson, G. & Morgan, P. J. Regulation of the Mel 1a melatonin receptor mRNA and protein levels in the ovine pars tuberalis: evidence for a cyclic adenosine 3’,5’-monophosphate-independent Mel 1a receptor coupling and an autoregulatory mechanism of expression. Mol Endocrinol 10, 892-902 (1996). 406. Masana, M. I., Benloucif, S. & Dubocovich, M. L. Circadian rhythm of mt1 melatonin receptor expression in the suprachiasmatic nucleus of the C3H/HeN mouse. J Pineal Res 28, 185-92 (2000). 407. Poirel, V. J., Masson-Pevet, M., Pevet, P. & Gauer, F. MT1 melatonin receptor mRNA expression exhibits a circadian variation in the rat suprachiasmatic nuclei. Brain Res 946, 64-71 (2002). 408. Barrett, P., Morris, M., Choi, W. S., Ross, A. & Morgan, P. J. Melatonin receptors and signal transduction mechanisms. Biol Signals Recept 8, 6-14 (1999). 409. Roca, A. L., Godson, C., Weaver, D. R. & Reppert, S. M. Structure, characterization, and expression of the gene encoding the mouse Mel1a melatonin receptor. Endocrinology 137, 3469-77 (1996). 410. Clemens, J. W., Jarzynka, M. J. & Witt-Enderby, P. A. Down-regulation of mt1 melatonin receptors in rat ovary following estrogen exposure. Life Sci 69, 27-35 (2001). 411. Masana, M. I., Soares, J. M., Jr. & Dubocovich, M. L. 17Beta-estradiol modulates hMT1 melatonin receptor function. Neuroendocrinology 81, 87-95 (2005).

193 references

412. Seltzer, A., Viswanathan, M. & Saavedra, J. M. Melatonin-binding sites in brain and caudal arteries of the female rat during the estrous cycle and after estrogen administration. Endocrinology 130, 1896-902 (1992). 413. Heideman, P. D., Bierl, C. K. & Sylvester, C. J. Photoresponsive Fischer 344 Rats are reproduc- tively inhibited by melatonin and differ in 2-[125I] lodomelatonin binding from nonphotore- sponsive Sprague-Dawley rats. J Neuroendocrinol 13, 223-32 (2001). 414. Viswanathan, M., Laitinen, J. T. & Saavedra, J. M. Vascular melatonin receptors. Biol Signals 2, 221-7 (1993). 415. Karasek, M. Melatonin, human aging, and age-related diseases. Exp Gerontol 39, 1723-9 (2004). 416. Hoogendijk, W. J. et al. Increased activity of surviving locus ceruleus neurons in Alzheimer’s disease. Ann Neurol 45, 82-91 (1999). 417. Burke, W. J. et al. Accumulation of 3,4-dihydroxyphenylglycolaldehyde, the neurotoxic monoamine oxidase A metabolite of norepinephrine, in locus ceruleus cell bodies in Alzheimer’s disease: mechanism of neuron death. Brain Res 816, 633-7 (1999). 418. Gottfries, C. G. et al. Biochemical changes in dementia disorders of Alzheimer type (AD/SDAT). Neurobiol Aging 4, 261-71 (1983). 419. Storga, D., Vrecko, K., Birkmayer, J. G. & Reibnegger, G. Monoaminergic neurotransmitters, their precursors and metabolites in brains of Alzheimer patients. Neurosci Lett 203, 29-32 (1996). 420. Kennedy, B. P. et al. Early and persistent alterations in prefrontal cortex MAO A and B in Alzheimer’s disease. J Neural Transm 110, 789-801 (2003). 421. Sabol, S. Z., Hu, S. & Hamer, D. A functional polymorphism in the monoamine oxidase A gene promoter. Hum Genet 103, 273-9 (1998). 422. Denney, R. M., Koch, H. & Craig, I. W. Association between monoamine oxidase A activity in human male skin fibroblasts and genotype of the MAOA promoter-associated variable number tandem repeat. Hum Genet 105, 542-51 (1999). 423. Takekida, S., Yan, L., Maywood, E. S., Hastings, M. H. & Okamura, H. Differential adrenergic regulation of the circadian expression of the clock genes Period1 and Period2 in the rat pineal gland. Eur J Neurosci 12, 4557-61 (2000). 424. Simonneaux, V. et al. Daily rhythm and regulation of clock gene expression in the rat pineal gland. Brain Res Mol Brain Res 120, 164-72 (2004). 425. Lebert, F., Pasquier, F. & Petit, H. [Behavior assessment in Alzheimer type dementia using the disruptive behavior questionnaire]. Presse Med 25, 665-7 (1996). 426. Cassone, V. M., Chesworth, M. J. & Armstrong, S. M. Entrainment of rat circadian rhythms by daily injection of melatonin depends upon the hypothalamic suprachiasmatic nuclei. Physiol Behav 36, 1111-21 (1986). 427. Vanecek, J. & Watanabe, K. Mechanisms of melatonin action in the pituitary and SCN. Adv Exp Med Biol 460, 191-8 (1999). 428. Reiter, R. J. Melatonin: Lowering the High Price of Free Radicals. News Physiol Sci 15, 246-250 (2000). 429. Neuwelt, E. A. & Lewy, A. J. Disappearance of plasma melatonin after removal of a neoplastic pineal gland. N Engl J Med 308, 1132-5 (1983). 430. Cardinali, D. P. Melatonin. A mammalian pineal hormone. Endocr Rev 2, 327-46 (1981). 431. Brzezinski, A. Melatonin in humans. N Engl J Med 336, 186-95 (1997). 432. Borjigin, J., Li, X. & Snyder, S. H. The pineal gland and melatonin: molecular and pharmacologic regulation. Annu Rev Pharmacol Toxicol 39, 53-65 (1999). 433. Braak, H. & Braak, E. Neuropathological stageing of Alzheimer-related changes. Acta Neu- ropathol (Berl) 82, 239-59 (1991).

194 references

434. Braak, H. & Braak, E. Evolution of the neuropathology of Alzheimer’s disease. Acta Neurol Scand Suppl 165, 3-12 (1996). 435. Rub, U., Del Tredici, K., Del Turco, D. & Braak, H. The intralaminar nuclei assigned to the medial pain system and other components of this system are early and progressively affected by the Alzheimer’s disease-related cytoskeletal pathology. J Chem Neuroanat 23, 279-90 (2002). 436. McKhann, G. et al. Clinical diagnosis of Alzheimer’s disease: report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer’s Disease. Neurology 34, 939-44 (1984). 437. Murphy, D. L., Tamarkin, L., Sunderland, T., Garrick, N. A. & Cohen, R. M. Human plasma melatonin is elevated during treatment with the monoamine oxidase inhibitors clorgyline and tranylcypromine but not deprenyl. Psychiatry Res 17, 119-27 (1986). 438. King, T. S., Richardson, B. A. & Reiter, R. J. Regulation of rat pineal melatonin synthesis: effect of monoamine oxidase inhibition. Mol Cell Endocrinol 25, 327-38 (1982). 439. Ravid, R., Van Zwieten, E. J. & Swaab, D. F. Brain banking and the human hypothalamus--factors to match for, pitfalls and potentials. Prog Brain Res 93, 83-95 (1992). 440. Lerchl, A. & Partsch, C. J. Reliable analysis of individual human melatonin profiles by complex cosinor analysis. J Pineal Res 16, 85-90 (1994). 441. Hofman, M. A., Skene, D. J. & Swaab, D. F. Effect of photoperiod on the diurnal melatonin and 5-methoxytryptophol rhythms in the human pineal gland. Brain Res 671, 254-60 (1995). 442. Bradford, M. M. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72, 248-54 (1976). 443. Stanley, M. & Brown, G. M. Melatonin levels are reduced in the pineal glands of suicide victims. Psychopharmacol Bull 24, 484-8 (1988). 444. Maywood, E. S. et al. Circadian and daily rhythms of melatonin in the blood and pineal gland of free-running and entrained Syrian hamsters. J Endocrinol 136, 65-73 (1993). 445. Feenstra, M. G., Botterblom, M. H. & Van Uum, J. F. Local activation of metabotropic glutamate receptors inhibits the handling-induced increased release of dopamine in the nucleus accumbens but not that of dopamine or noradrenaline in the prefrontal cortex: comparison with inhibition of ionotropic receptors. J Neurochem 70, 1104-13 (1998). 446. Ihalainen, J. A., Riekkinen, P., Jr. & Feenstra, M. G. Comparison of dopamine and noradrenaline release in mouse prefrontal cortex, striatum and hippocampus using microdialysis. Neurosci Lett 277, 71-4 (1999). 447. Sambrook, J. & Russell, D. W. Molecular cloning : a laboratory manual (Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y., 2001). 448. Vatsis, K. P. & Weber, W. W. Structural heterogeneity of Caucasian N-acetyltransferase at the NAT1 gene locus. Arch Biochem Biophys 301, 71-6 (1993). 449. Coon, S. L. et al. The human serotonin N-acetyltransferase (EC 2.3.1.87) gene (AANAT): structure, chromosomal localization, and tissue expression. Genomics 34, 76-84 (1996). 450. Debiec-Rychter, M., Land, S. J. & King, C. M. Histological localization of acetyltransferases in human tissue. Cancer Lett 143, 99-102 (1999). 451. Fowler, J. et al. Monoamine oxidase: radiotracer development and human studies. Methods 27, 263 (2002). 452. Warrington, J. A., Nair, A., Mahadevappa, M. & Tsyganskaya, M. Comparison of human adult and fetal expression and identification of 535 housekeeping/maintenance genes. Physiol Ge- nomics 2, 143-7 (2000). 453. Tanna, N. K. et al. Analysis of brain and cerebrospinal fluid volumes with MR imaging: impact on PET data correction for atrophy. Part II. Aging and Alzheimer dementia. Radiology 178, 123-30 (1991).

195 references

454. Gastel, J. A., Roseboom, P. H., Rinaldi, P. A., Weller, J. L. & Klein, D. C. Melatonin production: proteasomal proteolysis in serotonin N-acetyltransferase regulation. Science 279, 1358-60 (1998). 455. Nazarali, A. J. & Reynolds, G. P. Monoamine neurotransmitters and their metabolites in brain regions in Alzheimer’s disease: a postmortem study. Cell Mol Neurobiol 12, 581-7 (1992). 456. Boschi, G., Scherman, D. & Rips, R. Predominant cytosolic distribution of serotonin in rat pineal gland in contrast to biogenic monoamine localization in midbrain and adrenal gland. J Pineal Res 7, 23-9 (1989). 457. Oxenkrug, G. F., Balon, R., Jain, A. K. & McIntyre, I. M. Melatonin plasma response to MAO inhibitor: influence on human pineal activity? Acta Psychiatr Scand 77, 160-2 (1988). 458. Oxenkrug, G. F., McIntyre, I. M., Requintina, P. J. & Duffy, J. D. The response of the pineal melatonin biosynthesis to the selective MAO-A inhibitor, clorgyline, in young and middle-aged rats. Prog Neuropsychopharmacol Biol Psychiatry 15, 895-902 (1991). 459. Pfeffer, M., Kuhn, R., Krug, L., Korf, H. W. & Stehle, J. H. Rhythmic variation in beta1-adrenergic receptor mRNA levels in the rat pineal gland: circadian and developmental regulation. Eur J Neurosci 10, 2896-904 (1998). 460. Moller, M., Phansuwan-Pujito, P., Morgan, K. C. & Badiu, C. Localization and diurnal expression of mRNA encoding the beta1-adrenoceptor in the rat pineal gland: an in situ hybridization study. Cell Tissue Res 288, 279-84 (1997). 461. Hynd, M. R., Lewohl, J. M., Scott, H. L. & Dodd, P. R. Biochemical and molecular studies using human autopsy brain tissue. J Neurochem 85, 543-62 (2003). 462. Shih, J. C., Chen, K. & Ridd, M. J. Monoamine oxidase: from genes to behavior. Annu Rev Neurosci 22, 197-217 (1999). 463. Deckert, J. et al. Excess of high activity monoamine oxidase A gene promoter alleles in female patients with panic disorder. Hum Mol Genet 8, 621-4 (1999). 464. Theuns, J. & Van Broeckhoven, C. Transcriptional regulation of Alzheimer’s disease genes: implications for susceptibility. Hum Mol Genet 9, 2383-94 (2000). 465. Leboyer, M. et al. Psychiatric genetics: search for phenotypes. Trends Neurosci 21, 102-5 (1998). 466. Corder, E. H. et al. Gene dose of apolipoprotein E type 4 allele and the risk of Alzheimer’s disease in late onset families. Science 261, 921-3 (1993). 467. Wenham, P. R., Price, W. H. & Blandell, G. Apolipoprotein E genotyping by one-stage PCR. Lancet 337, 1158-9 (1991). 468. Preisig, M. et al. Association between bipolar disorder and monoamine oxidase A gene polymorphisms: results of a multicenter study. Am J Psychiatry 157, 948-55 (2000). 469. Huang, Y. Y. et al. An association between a functional polymorphism in the monoamine oxidase a gene promoter, impulsive traits and early abuse experiences. Neuropsychopharmacology 29, 1498-505 (2004). 470. Balciuniene, J., Emilsson, L., Oreland, L., Pettersson, U. & Jazin, E. Investigation of the functional effect of monoamine oxidase polymorphisms in human brain. Hum Genet 110, 1-7 (2002). 471. Halliwell, B. Oxidative stress in cell culture: an under-appreciated problem? FEBS Lett 540, 3-6 (2003). 472. Benzi, G. & Moretti, A. Are reactive oxygen species involved in Alzheimer’s disease? Neurobiol Aging 16, 661-74 (1995). 473. Behl, C. & Moosmann, B. Oxidative nerve cell death in Alzheimer’s disease and stroke: anti- oxidants as neuroprotective compounds. Biol Chem 383, 521-36 (2002). 474. Yu, Y. W. et al. Association study of a monoamine oxidase a gene promoter polymorphism with major depressive disorder and antidepressant response. Neuropsychopharmacology 30, 1719-23 (2005).

196 references

475. Cohen, I. L. et al. Association of autism severity with a monoamine oxidase A functional polymorphism. Clin Genet 64, 190-7 (2003). 476. Olin, J. T., Katz, I. R., Meyers, B. S., Schneider, L. S. & Lebowitz, B. D. Provisional diagnostic criteria for depression of Alzheimer disease: rationale and background. Am J Geriatr Psychiatry 10, 129-41 (2002). 477. Holtzer, R. et al. Depressive symptoms in Alzheimer’s disease: natural course and temporal relation to function and cognitive status. J Am Geriatr Soc 53, 2083-9 (2005). 478. Schulze, T. G. et al. Association between a functional polymorphism in the monoamine oxidase A gene promoter and major depressive disorder. Am J Med Genet 96, 801-3 (2000). 479. Du, L., Bakish, D., Ravindran, A. & Hrdina, P. D. MAO-A gene polymorphisms are associated with major depression and sleep disturbance in males. Neuroreport 15, 2097-101 (2004). 480. Hoogendijk, W. J. et al. Lack of association between depression and loss of neurons in the locus coeruleus in Alzheimer disease. Arch Gen Psychiatry 56, 45-51 (1999). 481. Yoo, S. H. et al. PERIOD2::LUCIFERASE real-time reporting of circadian dynamics reveals persistent circadian oscillations in mouse peripheral tissues. Proc Natl Acad Sci U S A 101, 5339-46 (2004). 482. Klein, D. C. Photoneural regulation of the mammalian pineal gland. Ciba Found Symp 117, 38-56 (1985). 483. Fukuhara, C., Dirden, J. C. & Tosini, G. Circadian expression of period 1, period 2, and arylalkylamine N-acetyltransferase mRNA in the rat pineal gland under diﬀerent light conditions. Neurosci Lett 286, 167-70 (2000). 484. Von Gall, C. et al. Clock gene protein mPER1 is rhythmically synthesized and under cAMP control in the mouse pineal organ. J Neuroendocrinol 13, 313-6 (2001). 485. Fukuhara, C., Yamazaki, S. & Liang, J. Pineal circadian clocks gate arylalkylamine N-acetyltransferase gene expression in the mouse pineal gland. J Neurochem 93, 156-62 (2005). 486. Perreau-Lenz, S., Kalsbeek, A., Pevet, P. & Buijs, R. M. Glutamatergic clock output stimulates melatonin synthesis at night. Eur J Neurosci 19, 318-24 (2004). 487. Garidou, M. L., Bartol, I., Calgari, C., Pevet, P. & Simonneaux, V. In vivo observation of a non- noradrenergic regulation of arylalkylamine N-acetyltransferase gene expression in the rat pineal complex. Neuroscience 105, 721-9 (2001). 488. Kalsbeek, A., Van Heerikhuize, J. J., Wortel, J. & Buijs, R. M. A diurnal rhythm of stimulatory input to the hypothalamo-pituitary-adrenal system as revealed by timed intrahypothalamic administration of the vasopressin V1 antagonist. J Neurosci 16, 5555-65 (1996). 489. Engel, L., Lorenzkowski, V., Langer, C., Rohleder, N. & Spessert, R. The photoperiod entrains the molecular clock of the rat pineal. Eur J Neurosci 21, 2297-304 (2005). 490. Johnston, J. D. et al. Rhythmic melatonin secretion does not correlate with the expression of arylalkylamine N-acetyltransferase, inducible cyclic amp early repressor, period1 or crypto- chrome1 mRNA in the sheep pineal. Neuroscience 124, 789-95 (2004). 491. Carter, D. A. Up-regulation of beta 1-adrenoceptor messenger ribonucleic acid in the rat pineal gland: nocturnally, through a beta-adrenoceptor-linked mechanism, and in vitro, through a novel posttranscriptional mechanism activated by speciﬁc protein synthesis inhibitors. Endo- crinology 133, 2263-8 (1993). 492. Fukuhara, C. et al. Gating of the cAMP signaling cascade and melatonin synthesis by the circadian clock in mammalian retina. J Neurosci 24, 1803-11 (2004). 493. Coon, S. L. et al. cAmp regulation of arylalkylamine N-acetyltransferase (AANAT, EC 2.3.1.87): a new cell line (1E7) provides evidence of intracellular AANAT activation. J Biol Chem 276, 24097-107 (2001). 494. Arendt, J., Bojkowski, C., Franey, C., Wright, J. & Marks, V. Immunoassay of 6-hydroxymelatonin sulfate in human plasma and urine: abolition of the urinary 24-hour rhythm with atenolol. J Clin Endocrinol Metab 60, 1166-73 (1985).

197 references

495. Vaughan, G. M. et al. Nocturnal elevation of plasma melatonin and urinary 5-hydroxyindoleacetic acid in young men: attempts at modification by brief changes in environmental lighting and sleep and by autonomic drugs. J Clin Endocrinol Metab 42, 752-64 (1976). 496. Berlin, I., Touitou, Y., Guillemant, S., Danjou, P. & Puech, A. J. Beta-adrenoceptor agonists do not stimulate daytime melatonin secretion in healthy subjects. A double blind placebo controlled study. Life Sci 56, PL325-31 (1995). 497. Grady, C. L., Furey, M. L., Pietrini, P., Horwitz, B. & Rapoport, S. I. Altered brain functional connectivity and impaired short-term memory in Alzheimer’s disease. Brain 124, 739-56 (2001). 498. Selkoe, D. J. & Podlisny, M. B. Deciphering the genetic basis of Alzheimer’s disease. Annu Rev Genomics Hum Genet 3, 67-99 (2002). 499. Bell-Pedersen, D. et al. Circadian rhythms from multiple oscillators: lessons from diverse organ- isms. Nat Rev Genet 6, 544-56 (2005). 500. Burke, W. J. et al. Effect of pre- and postmortem variables on specific mRNA levels in human brain. Brain Res Mol Brain Res 11, 37-41 (1991). 501. Barton, A. J., Pearson, R. C., Najlerahim, A. & Harrison, P. J. Pre- and postmortem influences on brain RNA. J Neurochem 61, 1-11 (1993). 502. Arendt, J. & Skene, D. J. Melatonin as a chronobiotic. Sleep Med Rev 9, 25-39 (2005). 503. Witt-Enderby, P. A., Bennett, J., Jarzynka, M. J., Firestine, S. & Melan, M. A. Melatonin receptors and their regulation: biochemical and structural mechanisms. Life Sci 72, 2183-98 (2003). 504. Weaver, D. R. & Reppert, S. M. The Mel1a melatonin receptor gene is expressed in human suprachiasmatic nuclei. Neuroreport 8, 109-12 (1996). 505. Yasin, S. A. et al. Melatonin and its analogs inhibit the basal and stimulated release of hypothalamic vasopressin and oxytocin in vitro. Endocrinology 132, 1329-36 (1993). 506. Juszczak, M. The hypothalamo-neurohypophysial response to melatonin. Neuro Endocrinol Lett 22, 169-74 (2001). 507. Forsling, M. L., Wheeler, M. J. & Williams, A. J. The effect of melatonin administration on pituitary hormone secretion in man. Clin Endocrinol (Oxf) 51, 637-42 (1999). 508. Forsling, M. L. & Williams, A. J. The effect of exogenous melatonin on stimulated neurohypophysial hormone release in man. Clin Endocrinol (Oxf) 57, 615-20 (2002). 509. Konakchieva, R., Mitev, Y., Almeida, O. F. & Patchev, V. K. Chronic melatonin treatment and the hypothalamo-pituitary-adrenal axis in the rat: attenuation of the secretory response to stress and effects on hypothalamic neuropeptide content and release Biol Cell 89, 587-96 (1997). 510. Konakchieva, R., Mitev, Y., Almeida, O. F. & Patchev, V. K. Chronic melatonin treatment coun- teracts glucocorticoid-induced dysregulation of the hypothalamic-pituitary-adrenal axis in the rat. Neuroendocrinology 67, 171-80 (1998). 511. Savaskan, E. et al. Cerebrovascular melatonin MT1-receptor alterations in patients with Alzhe- imer’s disease. Neurosci Lett 308, 9-12 (2001). 512. Savaskan, E. et al. Increased melatonin 1a-receptor immunoreactivity in the hippocampus of Alzheimer’s disease patients. J Pineal Res 32, 59-62 (2002). 513. Van der Sluis, P. J., Pool, C. W. & Sluiter, A. A. Immunochemical detection of peptides and proteins on press-blots after direct tissue gel isoelectric focusing. Electrophoresis 9, 654-61 (1988). 514. Kruijver, F. P., Fernandez-Guasti, A., Fodor, M., Kraan, E. M. & Swaab, D. F. Sex differences in androgen receptors of the human mamillary bodies are related to endocrine status rather than to sexual orientation or transsexuality. J Clin Endocrinol Metab 86, 818-27 (2001). 515. Adams, J. C. Biotin amplification of biotin and horseradish peroxidase signals in histochemical stains. J Histochem Cytochem 40, 1457-63 (1992). 516. Hou-Yu, A. et al. A monoclonal antibody to vasopressin: preparation, characterization, and application in immunocytochemistry. J Histochem Cytochem 30, 1249-60 (1982).

198 references

517. de Bree, F. M. & Burbach, J. P. Heterologous biosynthesis and processing of preprovasopressin in Neuro2A neuroblastoma cells. Biochimie 76, 315-9 (1994). 518. Hou-Yu, A., Lamme, A. T., Zimmerman, E. A. & Silverman, A. J. Comparative distribution of vasopressin and oxytocin neurons in the rat brain using a double-label procedure. Neuroen- docrinology 44, 235-46 (1986). 519. Van Oers, J. W., Tilders, J. H. & Berkenbosch, F. Characterization and biological activity of a rat monoclonal antibody to rat/human corticotropin-releasing factor. Endocrinology 124, 1239-46 (1989). 520. Raadsheer, F. C., Sluiter, A. A., Ravid, R., Tilders, F. J. & Swaab, D. F. Localization of corticotropin-releasing hormone (CRH) neurons in the paraventricular nucleus of the human hypothalamus; age-dependent colocalization with vasopressin. Brain Res 615, 50-62 (1993). 521. Kruijver, F. P., Balesar, R., Espila, A. M., Unmehopa, U. A. & Swaab, D. F. Estrogen receptor- alpha distribution in the human hypothalamus in relation to sex and endocrine status. J Comp Neurol 454, 115-39 (2002). 522. Weaver, D. R. & Capodice, C. E. Postmortem stability of melatonin receptor binding and clock- relevant mRNAs in mouse suprachiasmatic nucleus. J Biol Rhythms 16, 216-23 (2001). 523. Laudon, M., Nir, I. & Zisapel, N. Melatonin receptors in discrete brain areas of the male rat. Impact of aging on density and on circadian rhythmicity. Neuroendocrinology 48, 577-83 (1988). 524. Benloucif, S., Masana, M. I. & Dubocovich, M. L. Responsiveness to melatonin and its receptor expression in the aging circadian clock of mice. Am J Physiol 273, R1855-60 (1997). 525. Weaver, D. R., Liu, C. & Reppert, S. M. Nature’s knockout: the Mel1b receptor is not necessary for reproductive and circadian responses to melatonin in Siberian hamsters. Mol Endocrinol 10, 1478-87 (1996). 526. Thomas, L., Drew, J. E., Abramovich, D. R. & Williams, L. M. The role of melatonin in the human fetus (review). Int J Mol Med 1, 539-43 (1998). 527. Song, C. K., Bartness, T. J., Petersen, S. L. & Bittman, E. L. Co-expression of melatonin (MEL1a) receptor and arginine vasopressin mRNAs in the Siberian hamster suprachiasmatic nucleus. J Neuroendocrinol 12, 627-34 (2000). 528. Watanabe, K., Yamaoka, S. & Vanecek, J. Melatonin inhibits spontaneous and VIP-induced vasopressin release from suprachiasmatic neurons. Brain Res 801, 216-9 (1998). 529. Hawthorn, J., Ang, V. T. & Jenkins, J. S. Effects of lesions in the hypothalamic paraventricular, supraoptic and suprachiasmatic nuclei on vasopressin and oxytocin in rat brain and spinal cord. Brain Res 346, 51-7 (1985). 530. Pasqualetti, P., Festuccia, V., Collacciani, A., Acitelli, P. & Casale, R. Circadian rhythm of arginine vasopressin in hepatorenal syndrome. Nephron 78, 33-7 (1998). 531. Pierpaoli, W. & Maestroni, G. J. Melatonin: a principal neuroimmunoregulatory and anti-stress hormone: its anti-aging effects. Immunol Lett 16, 355-61 (1987). 532. Aoyama, H., Mori, N. & Mori, W. Anti-glucocorticoid effects of melatonin on adult rats. Acta Pathol Jpn 37, 1143-8 (1987). 533. Claes, S. J. Corticotropin-releasing hormone (CRH) in psychiatry: from stress to psychopathol- ogy. Ann Med 36, 50-61 (2004). 534. Baxter, M. G. & Chiba, A. A. Cognitive functions of the basal forebrain. Curr Opin Neurobiol 9, 178-83 (1999). 535. Gibbs, R. B. & Aggarwal, P. Estrogen and basal forebrain cholinergic neurons: implications for brain aging and Alzheimer’s disease-related cognitive decline. Horm Behav 34, 98-111 (1998). 536. Wada, H., Inagaki, N., Yamatodani, A. & Watanabe, T. Is the histaminergic neuron system a regulatory center for whole-brain activity? Trends Neurosci 14, 415-8 (1991).

199 references

537. Huston, J. P., Wagner, U. & Hasenohrl, R. U. The tuberomammillary nucleus projections in the control of learning, memory and reinforcement processes: evidence for an inhibitory role. Behav Brain Res 83, 97-105 (1997). 538. Raz, N., Torres, I. J. & Acker, J. D. Age-related shrinkage of the mamillary bodies: in vivo MRI evidence. Neuroreport 3, 713-6 (1992). 539. Loesch, D. V., Gilman, S., Del Dotto, J. & Rosenblum, M. L. Cavernous malformation of the mammillary bodies: neuropsychological implications. Case report. J Neurosurg 83, 354-8 (1995). 540. Parker, A., Eacott, M. J. & Gaffan, D. The recognition memory deficit caused by mediodorsal thalamic lesion in non-human primates: a comparison with rhinal cortex lesion. Eur J Neurosci 9, 2423-31 (1997). 541. Reiter, R. J. Oxidative damage in the central nervous system: protection by melatonin. Prog Neurobiol 56, 359-84 (1998). 542. Tan, D. X. et al. Melatonin suppresses autoxidation and hydrogen peroxide-induced lipid per- oxidation in monkey brain homogenate. Neuroendocrinol Lett 21, 361-365 (2000). 543. Husson, I. et al. Melatoninergic neuroprotection of the murine periventricular white matter against neonatal excitotoxic challenge. Ann Neurol 51, 82-92 (2002). 544. Mayo, J. C. et al. Melatonin regulation of antioxidant enzyme gene expression. Cell Mol Life Sci 59, 1706-13 (2002). 545. Pappolla, M. A. et al. The neuroprotective activities of melatonin against the Alzheimer beta- protein are not mediated by melatonin membrane receptors. J Pineal Res 32, 135-42 (2002). 546. Braak, H. & Braak, E. Demonstration of amyloid deposits and neurofibrillary changes in whole brain sections. Brain Pathol 1, 213-6 (1991). 547. Brusco, L. I., Fainstein, I., Marquez, M. & Cardinali, D. P. Effect of melatonin in selected populations of sleep-disturbed patients. Biol Signals Recept 8, 126-31 (1999). 548. Monti, J. M. & Cardinali, D. P. A critical assessment of the melatonin effect on sleep in humans. Biol Signals Recept 9, 328-39 (2000). 549. Balik, A., Kretschmannova, K., Mazna, P., Svobodova, I. & Zemkova, H. Melatonin action in neonatal gonadotrophs. Physiol Res 53 Suppl 1, S153-66 (2004). 550. Williams, L. M. et al. Melatonin receptors in the brain and pituitary gland of hypothalamo- pituitary disconnected Soay rams. J Neuroendocrinol 9, 639-43 (1997). 551. Dardente, H., Klosen, P., Pevet, P. & Masson-Pevet, M. MT1 melatonin receptor mRNA expressing cells in the pars tuberalis of the European hamster: effect of photoperiod. J Neuroendocrinol 15, 778-86 (2003). 552. Vanecek, J. The melatonin receptors in rat ontogenesis. Neuroendocrinology 48, 201-3 (1988). 553. Williams, L. M., Martinoli, M. G., Titchener, L. T. & Pelletier, G. The ontogeny of central melatonin binding sites in the rat. Endocrinology 128, 2083-90 (1991). 554. Whealin, J. M., Burwell, R. D. & Gallagher, M. The effects of aging on diurnal water intake and melatonin binding in the suprachiasmatic nucleus. Neurosci Lett 154, 149-52 (1993). 555. Romijn, H. J., Van Uum, J. F., Emmering, J., Goncharuk, V. & Buijs, R. M. Colocalization of VIP with AVP in neurons of the human paraventricular, supraoptic and suprachiasmatic nucleus. Brain Res 832, 47-53 (1999). 556. Okawa, M. et al. Circadian rhythm disorders in sleep-waking and body temperature in elderly patients with dementia and their treatment. Sleep 14, 478-85 (1991). 557. Satlin, A., Volicer, L., Ross, V., Herz, L. & Campbell, S. Bright light treatment of behavioral and sleep disturbances in patients with Alzheimer’s disease. Am J Psychiatry 149, 1028-32 (1992). 558. Lovell, B. B., Ancoli-Israel, S. & Gevirtz, R. Effect of bright light treatment on agitated behavior in institutionalized elderly subjects. Psychiatry Res 57, 7-12 (1995).

200 references

559. Zisapel, N. The use of melatonin for the treatment of insomnia. Biol Signals Recept 8, 84-9 (1999). 560. Dubocovich, M. L., Rivera-Bermudez, M. A., Gerdin, M. J. & Masana, M. I. Molecular pharmacology, regulation and function of mammalian melatonin receptors. Front Biosci 8, d1093-108 (2003). 561. Beck-Friis, J. et al. Serum melatonin in relation to clinical variables in patients with major depressive disorder and a hypothesis of a low melatonin syndrome. Acta Psychiatr Scand 71, 319-30 (1985). 562. Claustrat, B., Chazot, G., Brun, J., Jordan, D. & Sassolas, G. A chronobiological study of melatonin and cortisol secretion in depressed subjects: plasma melatonin, a biochemical marker in major depression. Biol Psychiatry 19, 1215-28 (1984). 563. Brown, R. et al. Differences in nocturnal melatonin secretion between melancholic depressed patients and control subjects. Am J Psychiatry 142, 811-6 (1985). 564. Kennedy, S. H., Garfinkel, P. E., Parienti, V., Costa, D. & Brown, G. M. Changes in melatonin levels but not cortisol levels are associated with depression in patients with eating disorders. Arch Gen Psychiatry 46, 73-8 (1989). 565. McIntyre, I. M., Judd, F. K., Marriott, P. M., Burrows, G. D. & Norman, T. R. Plasma melatonin levels in affective states. Int J Clin Pharmacol Res 9, 159-64 (1989). 566. Nurnberger, J. I., Jr. et al. Melatonin suppression by light in euthymic bipolar and unipolar patients. Arch Gen Psychiatry 57, 572-9 (2000). 567. Brown, R. P. et al. Depressed mood and reality disturbance correlate with decreased nocturnal melatonin in depressed patients. Acta Psychiatr Scand 76, 272-5 (1987). 568. Lewy, A. J., Bauer, V. K., Cutler, N. L. & Sack, R. L. Melatonin treatment of winter depression: a pilot study. Psychiatry Res 77, 57-61 (1998). 569. Wirz-Justice, A. et al. Morning or night-time melatonin is ineffective in seasonal affective disorder. J Psychiatr Res 24, 129-37 (1990). 570. Dolberg, O. T., Hirschmann, S. & Grunhaus, L. Melatonin for the treatment of sleep disturbances in major depressive disorder. Am J Psychiatry 155, 1119-21 (1998). 571. Loo, H., Dalery, J., Macher, J. P. & Payen, A. [Pilot study comparing in blind the therapeutic effect of two doses of agomelatine, melatoninergic agonist and selective 5HT2C receptors antagonist, in the treatment of major depressive disorders]. Encephale 28, 356-62 (2002). 572. Leppamaki, S. et al. Effect of controlled-release melatonin on sleep quality, mood, and quality of life in subjects with seasonal or weather-associated changes in mood and behaviour. Eur Neuropsychopharmacol 13, 137-45 (2003). 573. Bellipanni, G., Bianchi, P., Pierpaoli, W., Bulian, D. & Ilyia, E. Effects of melatonin in perimeno- pausal and menopausal women: a randomized and placebo controlled study. Exp Gerontol 36, 297-310 (2001). 574. Guerrero, H. Y., Gauer, F., Schuster, C., Pevet, P. & Masson-Pevet, M. Melatonin regulates the mRNA expression of the mt(1) melatonin receptor in the rat Pars tuberalis. Neuroendocrinol- ogy 71, 163-9 (2000). 575. Holsboer, F. & Barden, N. Antidepressants and hypothalamic-pituitary-adrenocortical regulation. Endocr Rev 17, 187-205 (1996). 576. Swaab, D. F., Fliers, E., Hoogendijk, W. J., Veltman, D. J. & Zhou, J. N. Interaction of prefrontal cortical and hypothalamic systems in the pathogenesis of depression. Prog Brain Res 126, 369-96 (2000). 577. Kellner, M. et al. Corticotropin-releasing hormone inhibits melatonin secretion in healthy volunteers--a potential link to low-melatonin syndrome in depression? Neuroendocrinology 65, 284-90 (1997). 578. Raadsheer, F. C., Oorschot, D. E., Verwer, R. W., Tilders, F. J. & Swaab, D. F. Age-related increase in the total number of corticotropin-releasing hormone neurons in the human paraventricular nucleus in controls and Alzheimer’s disease: comparison of the disector with an unfolding method. J Comp Neurol 339, 447-57 (1994).

201 references

579. Raadsheer, F. C., Hoogendijk, W. J., Stam, F. C., Tilders, F. J. & Swaab, D. F. Increased numbers of corticotropin-releasing hormone expressing neurons in the hypothalamic paraventricular nucleus of depressed patients. Neuroendocrinology 60, 436-44 (1994). 580. Raadsheer, F. C. et al. Corticotropin-releasing hormone mRNA levels in the paraventricular nucleus of patients with Alzheimer’s disease and depression. Am J Psychiatry 152, 1372-6 (1995). 581. Bao, A. M., Hestiantoro, A., Van Someren, E. J., Swaab, D. F. & Zhou, J. N. Colocalization of corticotropin-releasing hormone and oestrogen receptor-alpha in the paraventricular nucleus of the hypothalamus in mood disorders. Brain 128, 1301-13 (2005). 582. Eastwood, S. L. & Harrison, P. J. Synaptic pathology in the anterior cingulate cortex in schizophrenia and mood disorders. A review and a Western blot study of synaptophysin, GAP-43 and the complexins. Brain Res Bull 55, 569-78 (2001). 583. Eastwood, S. L. & Harrison, P. J. Hippocampal synaptic pathology in schizophrenia, bipolar disorder and major depression: a study of complexin mRNAs. Mol Psychiatry 5, 425-32 (2000). 584. Beasley, C. L. et al. Reductions in cholesterol and synaptic markers in association cortex in mood disorders. Bipolar Disord 7, 449-55 (2005). 585. Abe, M., Herzog, E. D. & Block, G. D. Lithium lengthens the circadian period of individual suprachiasmatic nucleus neurons. Neuroreport 11, 3261-4 (2000). 586. Mason, R. & Biello, S. M. A neurophysiological study of a lithium-sensitive phosphoinositide system in the hamster suprachiasmatic (SCN) biological clock in vitro. Neurosci Lett 144, 135-8 (1992). 587. Mason, R., Biello, S. M. & Harrington, M. E. The effects of GABA and benzodiazepines on neurones in the suprachiasmatic nucleus (SCN) of Syrian hamsters. Brain Res 552, 53-7 (1991). 588. Oxenkrug, G. F., Requintina, P. J., McIntyre, I. M. & White, K. Chronic effect of the irrevers- ible and reversible selective MAO-A inhibitors on rat pineal melatonin biosynthesis. J Neural Transm Suppl 41, 381-4 (1994). 589. Murphy, D. L., Garrick, N. A., Tamarkin, L., Taylor, P. L. & Markey, S. P. Effects of antidepressants and other psychotropic drugs on melatonin release and pineal gland function. J Neural Transm Suppl 21, 291-309 (1986). 590. Arendt, J. Mammalian Pineal Rhythms (Alan R. Liss, Inc, 1985). 591. Harrison, P. J. et al. The relative importance of premortem acidosis and postmortem interval for human brain gene expression studies: selective mRNA vulnerability and comparison with their encoded proteins. Neurosci Lett 200, 151-4 (1995). 592. Johnston, N. L., Cervenak, J., Shore, A. D., Torrey, E. F. & Yolken, R. H. Multivariate analysis of RNA levels from postmortem human brains as measured by three different methods of RT- PCR. Stanley Neuropathology Consortium. J Neurosci Methods 77, 83-92 (1997). 593. Yasojima, K., McGeer, E. G. & McGeer, P. L. High stability of mRNAs postmortem and protocols for their assessment by RT-PCR. Brain Res Brain Res Protoc 8, 212-8 (2001). 594. Dodd, P. R., Hambley, J. W., Cowburn, R. F. & Hardy, J. A. A comparison of methodologies for the study of functional transmitter neurochemistry in human brain. J Neurochem 50, 1333-45 (1988). 595. Kontur, P. J., al-Tikriti, M., Innis, R. B. & Roth, R. H. Postmortem stability of monoamines, their metabolites, and receptor binding in rat brain regions. J Neurochem 62, 282-90 (1994). 596. Catts, V. S. et al. A microarray study of post-mortem mRNA degradation in mouse brain tissue. Brain Res Mol Brain Res 138, 164-77 (2005). 597. Preece, P. & Cairns, N. J. Quantifying mRNA in postmortem human brain: influence of gender, age at death, postmortem interval, brain pH, agonal state and inter-lobe mRNA variance. Brain Res Mol Brain Res 118, 60-71 (2003). 598. Schramm, M. et al. Stability of RNA transcripts in post-mortem psychiatric brains. J Neural Transm 106, 329-35 (1999).

202 references

599. Perrett, C. W., Marchbanks, R. M. & Whatley, S. A. Characterisation of messenger RNA extracted post-mortem from the brains of schizophrenic, depressed and control subjects. J Neurol Neurosurg Psychiatry 51, 325-31 (1988). 600. Swift, G. H., Peyton, M. J. & MacDonald, R. J. Assessment of RNA quality by semi-quantitative RT-PCR of multiple regions of a long ubiquitous mRNA. Biotechniques 28, 524, 526, 528, 530-1 (2000). 601. Johnson, S. A., Morgan, D. G. & Finch, C. E. Extensive postmortem stability of RNA from rat and human brain. J Neurosci Res 16, 267-80 (1986). 602. Penev, P. D. & Zee, P. C. Melatonin: a clinical perspective. Ann Neurol 42, 545-53 (1997). 603. Waldhauser, F. et al. Alterations in nocturnal serum melatonin levels in humans with growth and aging. J Clin Endocrinol Metab 66, 648-52 (1988). 604. Luboshitzky, R., Lavi, S. & Lavie, P. The association between melatonin and sleep stages in normal adults and hypogonadal men. Sleep 22, 867-74 (1999). 605. Swaab, D. F. & Hofman, M. A. Age, sex and light: variability in the human suprachiasmatic nucleus in relation to its functions. Prog Brain Res 100, 261-5 (1994). 606. Molina-Carballo, A. et al. 5-Methoxytryptophol and melatonin in children: differences due to age and sex. J Pineal Res 21, 73-9 (1996). 607. Sugden, D., McArthur, A. J., Ajpru, S., Duniec, K. & Piggins, H. D. Expression of mt(1) melatonin receptor subtype mRNA in the entrained rat suprachiasmatic nucleus: a quantitative RT-PCR study across the diurnal cycle. Brain Res Mol Brain Res 72, 176-82 (1999). 608. Brown, R. P. et al. Nocturnal serum melatonin in major depressive disorder before and after desmethylimipramine treatment. Psychopharmacol Bull 21, 579-81 (1985). 609. Thompson, C. et al. The effect of desipramine upon melatonin and cortisol secretion in -de pressed and normal subjects. Br J Psychiatry 147, 389-93 (1985). 610. Golden, R. N. et al. Antidepressants reduce whole-body norepinephrine turnover while enhanc- ing 6-hydroxymelatonin output. Arch Gen Psychiatry 45, 150-4 (1988). 611. Von Bahr, C. et al. Fluvoxamine but not citalopram increases serum melatonin in healthy subjects-- an indication that cytochrome P450 CYP1A2 and CYP2C19 hydroxylate melatonin. Eur J Clin Pharmacol 56, 123-7 (2000). 612. Childs, P. A. et al. Effect of fluoxetine on melatonin in patients with seasonal affective disorder and matched controls. Br J Psychiatry 166, 196-8 (1995). 613. Yocca, F. D., de Paul Lynch, V. & Friedman, E. Effect of chronic lithium treatment on rat pineal rhythms: N-acetyltransferase, N-acetylserotonin and melatonin. J Pharmacol Exp Ther 226, 733-7 (1983). 614. Seggie, J., Werstiuk, E., Grota, L. & Brown, G. M. Chronic lithium treatment and twenty-four hour rhythm of serum prolactin, growth hormone and melatonin in rats. Prog Neuropsychop- harmacol Biol Psychiatry 7, 827-30 (1983). 615. Yin, L., Wang, J., Klein, P. S. & Lazar, M. A. Nuclear receptor Rev-erbalpha is a critical lithium- sensitive component of the circadian clock. Science 311, 1002-5 (2006). 616. Gaffori, O., Geffard, M. & Van Ree, J. M. des-Tyr1-gamma-endorphin and haloperidol increase pineal gland melatonin levels in rats. Peptides 4, 393-5 (1983). 617. Wurtman, R. J. & Axelrod, J. Effect of chlorpromazine and other drugs on the disposition of circulating melatonin. Nature 212, 312 (1966). 618. Beckmann, H., Wetterberg, L. & Gattaz, W. F. Melatonin immunoreactivity in cerebrospinal fluid of schizophrenic patients and healthy controls. Psychiatry Res 11, 107-10 (1984). 619. Smith, J. A., Mee, T. J. & Barnes, J. L. Elevated melatonin serum concentrations in psychiatric patients treated with chlorpromazine [proceedings]. J Pharm Pharmacol 29 Suppl, 30P (1977). 620. Smith, J. A., Barnes, J. L. & Mee, T. J. The effect of neuroleptic drugs on serum and cerebrospinal fluid melatonin concentrations in psychiatric subjects. J Pharm Pharmacol 31, 246-8 (1979).

203 references

621. Monteleone, P., Forziati, D., Orazzo, C. & Maj, M. Preliminary observations on the suppression of nocturnal plasma melatonin levels by short-term administration of diazepam in humans. J Pineal Res 6, 253-8 (1989). 622. Monteleone, P. et al. Suppression of nocturnal plasma melatonin levels by evening administration of sodium valproate in healthy humans. Biol Psychiatry 41, 336-41 (1997). 623. Hallam, K. T., Olver, J. S. & Norman, T. R. Effect of sodium valproate on nocturnal melatonin sensitivity to light in healthy volunteers. Neuropsychopharmacology 30, 1400-4 (2005). 624. Viyoch, J. et al. Effect of haloperidol on mPer1 gene expression in mouse suprachiasmatic nuclei. J Biol Chem 280, 6309-15 (2005). 625. Gao, B. & Moore, R. Y. Glutamic acid decarboxylase message isoforms in human suprachiasmatic nucleus. J Biol Rhythms 11, 172-9 (1996). 626. Cowen, P. J., Fraser, S., Sammons, R. & Green, A. R. Atenolol reduces plasma melatonin concentration in man. Br J Clin Pharmacol 15, 579-81 (1983). 627. Ribelayga, C., Gauer, F., Calgari, C., Pevet, P. & Simonneaux, V. Photoneural regulation of rat pineal hydroxyindole-O-methyltransferase (HIOMT) messenger ribonucleic acid expression: an analysis of its complex relationship with HIOMT activity. Endocrinology 140, 1375-84 (1999). 628. Bernard, M. et al. Transcriptional regulation of hydroxyindole O-methyltransferase in the chicken pineal gland: day/night changes and long-term effects of light and darkness. Biochem J 290 ( Pt 3), 661-4 (1993). 629. Yoshimura, T. et al. Molecular analysis of avian circadian clock genes. Brain Res Mol Brain Res 78, 207-15 (2000). 630. Fukada, Y. & Okano, T. Circadian clock system in the pineal gland. Mol Neurobiol 25, 19-30 (2002). 631. Karolczak, M., Burbach, G. J., Sties, G., Korf, H. W. & Stehle, J. H. Clock gene mRNA and protein rhythms in the pineal gland of mice. Eur J Neurosci 19, 3382-8 (2004). 632. Deguchi, T. Circadian rhythm of serotonin N-acetyltransferase activity in organ culture of chicken pineal gland. Science 203, 1245-7 (1979). 633. Takahashi, J. S., Hamm, H. & Menaker, M. Circadian rhythms of melatonin release from individual superfused chicken pineal glands in vitro. Proc Natl Acad Sci U S A 77, 2319-22 (1980). 634. Bernard, M. et al. Melatonin synthesis pathway: circadian regulation of the genes encoding the key enzymes in the chicken pineal gland and retina. Reprod Nutr Dev 39, 325-34 (1999). 635. Bernard, M., Klein, D. C. & Zatz, M. Chick pineal clock regulates serotonin N-acetyltransferase mRNA rhythm in culture. Proc Natl Acad Sci U S A 94, 304-9 (1997). 636. Klein, D. C. & Weller, J. L. Indole metabolism in the pineal gland: a circadian rhythm in N- acetyltransferase. Science 169, 1093-5 (1970). 637. Roseboom, P. H. et al. Melatonin synthesis: analysis of the more than 150-fold nocturnal increase in serotonin N-acetyltransferase messenger ribonucleic acid in the rat pineal gland. Endocrinology 137, 3033-45 (1996). 638. Klein, D. C. et al. Lesions of the paraventricular nucleus area of the hypothalamus disrupt the suprachiasmatic leads to spinal cord circuit in the melatonin rhythm generating system. Brain Res Bull 10, 647-52 (1983). 639. Roseboom, P. H. & Klein, D. C. Norepinephrine stimulation of pineal cyclic AMP response element-binding protein phosphorylation: primary role of a beta-adrenergic receptor/cyclic AMP mechanism. Mol Pharmacol 47, 439-49 (1995). 640. Baler, R., Covington, S. & Klein, D. C. The rat arylalkylamine N-acetyltransferase gene promoter. cAMP activation via a cAMP-responsive element-CCAAT complex. J Biol Chem 272, 6979-85 (1997).

204 references

641. Tamotsu, S., Schomerus, C., Stehle, J. H., Roseboom, P. H. & Korf, H. W. Norepinephrine-induced phosphorylation of the transcription factor CREB in isolated rat pinealocytes: an immunocytochemical study. Cell Tissue Res 282, 219-26 (1995). 642. Schomerus, C., Maronde, E., Laedtke, E. & Korf, H. W. Vasoactive intestinal peptide (VIP) and pituitary adenylate cyclase-activating polypeptide (PACAP) induce phosphorylation of the transcription factor CREB in subpopulations of rat pinealocytes: immunocytochemical and immunochemical evidence. Cell Tissue Res 286, 305-13 (1996). 643. Maronde, E. et al. CREB phosphorylation and melatonin biosynthesis in the rat pineal gland: involvement of cyclic AMP dependent protein kinase type II. J Pineal Res 27, 170-82 (1999). 644. Stehle, J. H., Von Gall, C. & Korf, H. W. Analysis of cell signalling in the rodent pineal gland deci- phers regulators of dynamic transcription in neural/endocrine cells. Eur J Neurosci 14, 1-9 (2001). 645. Foulkes, N. S., Borjigin, J., Snyder, S. H. & Sassone-Corsi, P. Transcriptional control of circadian hormone synthesis via the CREM feedback loop. Proc Natl Acad Sci U S A 93, 14140-5 (1996). 646. Pfeffer, M., Maronde, E., Molina, C. A., Korf, H. W. & Stehle, J. H. Inducible cyclic AMP early repressor protein in rat pinealocytes: a highly sensitive natural reporter for regulated gene transcription. Mol Pharmacol 56, 279-89 (1999). 647. Terriff, D. L., Chik, C. L., Price, D. M. & Ho, A. K. Proteasomal proteolysis in the adrenergic induction of arylalkylamine-N-acetyltransferase in rat pinealocytes. Endocrinology 146, 4795- 803 (2005). 648. Ebadi, M. & Govitrapong, P. Neural pathways and neurotransmitters affecting melatonin synthesis. J Neural Transm Suppl 21, 125-55 (1986). 649. Laitinen, J. T., Laitinen, K. S. & Kokkola, T. Cholinergic signaling in the rat pineal gland. Cell Mol Neurobiol 15, 177-92 (1995). 650. Simonneaux, V. & Ribelayga, C. Generation of the melatonin endocrine message in mammals: a review of the complex regulation of melatonin synthesis by norepinephrine, peptides, and other pineal transmitters. Pharmacol Rev 55, 325-95 (2003). 651. Coon, S. L. et al. Pineal serotonin N-acetyltransferase: expression cloning and molecular analysis. Science 270, 1681-3 (1995). 652. Privat, K., Ravault, J. P., Chesneau, D. & Fevre-Montange, M. Day/night variation of tryptophan hydroxylase and serotonin N-acetyltransferase mRNA levels in the ovine pineal gland and retina. J Pineal Res 26, 193-203 (1999). 653. Fleming, J. V., Barrett, P., Coon, S. L., Klein, D. C. & Morgan, P. J. Ovine arylalkylamine N- acetyltransferase in the pineal and pituitary glands: differences in function and regulation. Endocrinology 140, 972-8 (1999). 654. Ganguly, S. et al. Role of a pineal cAMP-operated arylalkylamine N-acetyltransferase/14-3-3- binding switch in melatonin synthesis. Proc Natl Acad Sci U S A 98, 8083-8 (2001). 655. Schomerus, C., Korf, H. W., Laedtke, E., Weller, J. L. & Klein, D. C. Selective adrenergic/cyclic AMP-dependent switch-off of proteasomal proteolysis alone switches on neural signal transduction: an example from the pineal gland. J Neurochem 75, 2123-32 (2000). 656. Namboodiri, M. Regulation of melatonin synthesis in the ovine pineal gland. adv exp med biol 294, 137-48 (1991). 657. Eloranta, E. et al. Seasonal and daily patterns in melatonin secretion in female reindeer and their calves. Endocrinology 130, 1645-52 (1992). 658. Reppert, S. M., Perlow, M. J., Tamarkin, L. & Klein, D. C. A diurnal melatonin rhythm in primate cerebrospinal fluid. Endocrinology 104, 295-301 (1979). 659. Smith, J. A., Padwick, D., Mee, T. J., Minneman, K. P. & Bird, E. D. Synchronous nyctohemeral rhythms in human blood melatonin and in human post-mortem pineal enzyme. Clin Endo- crinol (Oxf) 6, 219-25 (1977).

205 references

660. Zeitzer, J. M., Dijk, D. J., Kronauer, R., Brown, E. & Czeisler, C. Sensitivity of the human circadian pacemaker to nocturnal light: melatonin phase resetting and suppression. J Physiol 526 Pt 3, 695-702 (2000). 661. Lewy, A. J. The dim light melatonin onset, melatonin assays and biological rhythm research in humans. Biol Signals Recept 8, 79-83 (1999). 662. Smith, K. A., Schoen, M. W. & Czeisler, C. A. Adaptation of human pineal melatonin suppression by recent photic history. J Clin Endocrinol Metab 89, 3610-4 (2004). 663. Rollag, M. D. & Niswender, G. D. Radioimmunoassay of serum concentrations of melatonin in sheep exposed to different lighting regimens. Endocrinology 98, 482-9 (1976). 664. Besharse, J. C. & Luvone, P. M. Circadian clock in Xenopus eye controlling retinal serotonin N-acetyltransferase. Nature 305, 133-135 (1983). 665. Tosini, G. & Menaker, M. Circadian rhythms in cultured mammalian retina. Science 272, 419- 21 (1996). 666. Chong, N. W., Bernard, M. & Klein, D. C. Characterization of the chicken serotonin N-acetyltransferase gene. Activation via clock gene heterodimer/E box interaction. J Biol Chem 275, 32991-8 (2000). 667. Chong, N. W., Chaurasia, S. S., Haque, R., Klein, D. C. & Iuvone, P. M. Temporal-spatial characterization of chicken clock genes: circadian expression in retina, pineal gland, and peripheral tissues. J Neurochem 85, 851-60 (2003). 668. Bailey, M. J. et al. Transcriptional profiling of circadian patterns of mRNA expression in the chick retina. J Biol Chem 279, 52247-54 (2004). 669. Namihira, M. et al. Daily variation and light responsiveness of mammalian clock gene, Clock and BMAL1, transcripts in the pineal body and different areas of brain in rats. Neurosci Lett 267, 69-72 (1999). 670. Namihira, M. et al. Circadian pattern, light responsiveness and localization of rPer1 and rPer2 gene expression in the rat retina. Neuroreport 12, 471-5 (2001). 671. Sakamoto, K. et al. Multitissue circadian expression of rat period homolog (rPer2) mRNA is governed by the mammalian circadian clock, the suprachiasmatic nucleus in the brain. J Biol Chem 273, 27039-42 (1998). 672. Fukuhara, C., Dirden, J. C. & Tosini, G. Photic regulation of melatonin in rat retina and the role of proteasomal proteolysis. Neuroreport 12, 3833-7 (2001). 673. Chen, W. & Baler, R. The rat arylalkylamine N-acetyltransferase E-box: differential use in a master vs. a slave oscillator. Brain Res Mol Brain Res 81, 43-50 (2000). 674. Bliwise, D. L. et al. Disruptive nocturnal behavior in Parkinson’s disease and Alzheimer’s disease. J Geriatr Psychiatry Neurol 8, 107-10 (1995). 675. Steele, C., Rovner, B., Chase, G. A. & Folstein, M. Psychiatric symptoms and nursing home placement of patients with Alzheimer’s disease. Am J Psychiatry 147, 1049-51 (1990). 676. Tate, B. et al. Disruption of circadian regulation by brain grafts that overexpress Alzheimer beta/A4 amyloid. Proc Natl Acad Sci U S A 89, 7090-4 (1992). 677. Bron, A. J., Vrensen, G. F., Koretz, J., Maraini, G. & Harding, J. J. The ageing lens. Ophthalmo- logica 214, 86-104 (2000). 678. Sample, P. A., Esterson, F. D., Weinreb, R. N. & Boynton, R. M. The aging lens: in vivo assessment of light absorption in 84 human eyes. Invest Ophthalmol Vis Sci 29, 1306-11 (1988). 679. Swann, P. G. & Lovie-Kitchin, J. E. Age-related maculopathy. I: A review of its morphology and effects on visual function. Ophthalmic Physiol Opt 10, 149-58 (1990). 680. Algvere, P. V. & Seregard, S. Age-related maculopathy: pathogenetic features and new treatment modalities. Acta Ophthalmol Scand 80, 136-43 (2002). 681. Blanks, J. C., Hinton, D. R., Sadun, A. A. & Miller, C. A. Retinal ganglion cell degeneration in Alzheimer’s disease. Brain Res 501, 364-72 (1989).

206 references

682. Rietveld, W. J. & Groos, G. A. The role of suprachiasmatic nucleus afferents in the central regulation of circadian rhythms. Prog Clin Biol Res 59C, 205-11 (1981). 683. Janik, D. & Mrosovsky, N. Intergeniculate leaflet lesions and behaviorally-induced shifts of circadian rhythms. Brain Res 651, 174-82 (1994). 684. Wickland, C. & Turek, F. W. Lesions of the thalamic intergeniculate leaflet block activity-induced phase shifts in the circadian activity rhythm of the golden hamster. Brain Res 660, 293-300 (1994). 685. Van Someren, E. J. Actigraphic monitoring of movement and rest-activity rhythms in aging, Alzheimer’s disease, and Parkinson’s disease. IEEE Trans Rehabil Eng 5, 394-8 (1997). 686. Balestreri, R. et al. [Availability of serotonin in the central nervous system and negative feedback mechanism of corticosteroids in humans]. Arch Maragliano Patol Clin 34, 83-90 (1978). 687. Arai, H., Kosaka, K. & Iizuka, R. Changes of biogenic amines and their metabolites in postmortem brains from patients with Alzheimer-type dementia. J Neurochem 43, 388-93 (1984). 688. Ebinger, G. et al. Distribution of biogenic amines and their catabolites in brains from patients with Alzheimer’s disease. J Neurol Sci 77, 267-83 (1987). 689. Tejani-Butt, S. M., Yang, J. & Pawlyk, A. C. Altered serotonin transporter sites in Alzheimer’s disease raphe and hippocampus. Neuroreport 6, 1207-10 (1995). 690. Bina, K. G., Rusak, B. & Semba, K. Localization of cholinergic neurons in the forebrain and brainstem that project to the suprachiasmatic nucleus of the hypothalamus in rat. J Comp Neurol 335, 295-307 (1993). 691. Whitehouse, P. J. et al. Alzheimer’s disease and senile dementia: loss of neurons in the basal forebrain. Science 215, 1237-9 (1982). 692. Francis, P. T., Palmer, A. M., Snape, M. & Wilcock, G. K. The cholinergic hypothesis of Alzhe- imer’s disease: a review of progress. J Neurol Neurosurg Psychiatry 66, 137-47 (1999). 693. Wisor, J. P. et al. Sleep and circadian abnormalities in a transgenic mouse model of Alzheimer’s disease: a role for cholinergic transmission. Neuroscience 131, 375-85 (2005). 694. Mesulam, M., Shaw, P., Mash, D. & Weintraub, S. Cholinergic nucleus basalis tauopathy emerges early in the aging-MCI-AD continuum. Ann Neurol 55, 815-28 (2004). 695. Dubelaar, E. J. et al. ApoE epsilon4 genotype is accompanied by lower metabolic activity in nucleus basalis of Meynert neurons in Alzheimer patients and controls as indicated by the size of the Golgi apparatus. J Neuropathol Exp Neurol 63, 159-69 (2004). 696. Murphy, D. G., DeCarli, C., Schapiro, M. B., Rapoport, S. I. & Horwitz, B. Age-related differences in volumes of subcortical nuclei, brain matter, and cerebrospinal fluid in healthy men as measured with magnetic resonance imaging. Arch Neurol 49, 839-45 (1992). 697. Wahlund, L. O. et al. Cognitive functions and brain structures: a quantitative study of CSF volumes on Alzheimer patients and healthy control subjects. Magn Reson Imaging 11, 169-74 (1993). 698. Dumont, M., Montplaisir, J. & Infante-Rivard, C. Sleep Quality of Former Night-shift Workers. Int J Occup Environ Health 3, S10-S14 (1997). 699. Knutsson, A., Akerstedt, T., Jonsson, B. G. & Orth-Gomer, K. Increased risk of ischaemic heart disease in shift workers. Lancet 2, 89-92 (1986). 700. Van Someren, E. J. More than a marker: interaction between the circadian regulation of temperature and sleep, age-related changes, and treatment possibilities. Chronobiol Int 17, 313-54 (2000). 701. Meguro, K. et al. Disturbance in daily sleep/wake patterns in patients with cognitive impairment and decreased daily activity. J Am Geriatr Soc 38, 1176-82 (1990). 702. Meguro, K. et al. Sleep disturbance in elderly patients with cognitive impairment, decreased daily activity and periventricular white matter lesions. Sleep 18, 109-14 (1995). 703. Swaab, D. F. Brain aging and Alzheimer’s disease, “wear and tear” versus “use it or lose it”. Neurobiol Aging 12, 317-24 (1991).

207 references

704. Swaab, D. F. et al. Brain aging and Alzheimer’s disease; use it or lose it. Prog Brain Res 138, 343-73 (2002). 705. Rea, M. A. Photic entrainment of circadian rhythms in rodents. Chronobiol Int 15, 395-423 (1998). 706. Lucassen, P. J., Hofman, M. A. & Swaab, D. F. Increased light intensity prevents the age related loss of vasopressin-expressing neurons in the rat suprachiasmatic nucleus. Brain Res 693, 261-6 (1995). 707. Kohsaka, M. et al. Effects of moderately bright light on subjective evaluations in healthy elderly women. Psychiatry Clin Neurosci 53, 239-41 (1999). 708. Okawa, M. et al. Sleep disorders in elderly patients with dementia and attempts of bright light therapy in senile dementia of Alzheimer’s type and multi-infarct dementia. (eds. Hiroshige, T. & Honma, K.) (Hokkaido University Press, Sapporo, , 1994). 709. Mishima, K. et al. Morning bright light therapy for sleep and behavior disorders in elderly patients with dementia. Acta Psychiatr Scand 89, 1-7 (1994). 710. Okumoto, Y., Koyama, E., Matsubara, H., Nakano, T. & Nakamura, R. Sleep improvement by light in a demented aged individual. Psychiatry Clin Neurosci 52, 194-6 (1998). 711. Van Someren, E. J. in Medical Faculty of the University of Amsterdam 303 (University of Am- sterdam, Amsterdam, 1997). 712. Lyketsos, C. G., Lindell Veiel, L., Baker, A. & Steele, C. A randomized, controlled trial of bright light therapy for agitated behaviors in dementia patients residing in long-term care. Int J Geriatr Psychiatry 14, 520-5 (1999). 713. Dowling, G. A. et al. Effect of morning bright light treatment for rest-activity disruption in institutionalized patients with severe Alzheimer’s disease. Int Psychogeriatr 17, 221-36 (2005). 714. Dowling, G. A., Mastick, J., Hubbard, E. M., Luxenberg, J. S. & Burr, R. L. Effect of timed bright light treatment for rest-activity disruption in institutionalized patients with Alzheimer’s disease. Int J Geriatr Psychiatry 20, 738-43 (2005). 715. Drevets, W. C. & Rubin, E. H. Psychotic symptoms and the longitudinal course of senile dementia of the Alzheimer type. Biol Psychiatry 25, 39-48 (1989). 716. Mega, M. S., Cummings, J. L., Fiorello, T. & Gornbein, J. The spectrum of behavioral changes in Alzheimer’s disease. Neurology 46, 130-5 (1996). 717. Ryu, S. H., Katona, C., Rive, B. & Livingston, G. Persistence of and changes in neuropsychiatric symptoms in Alzheimer disease over 6 months: the LASER-AD study. Am J Geriatr Psychiatry 13, 976-83 (2005). 718. Chouinard, G. et al. A Canadian multicentre placebo-controlled study of a fixed dose of bro- faromine, a reversible selective MAO-A inhibitor, in the treatment of major depression. J Affect Disord 32, 105-14 (1994). 719. Hsiao, K. K. et al. Age-related CNS disorder and early death in transgenic FVB/N mice over- expressing Alzheimer amyloid precursor proteins. Neuron 15, 1203-18 (1995). 720. Smith, M. A. et al. Amyloid-beta deposition in Alzheimer transgenic mice is associated with oxidative stress. J Neurochem 70, 2212-5 (1998). 721. Lewis, J. et al. Neurofibrillary tangles, amyotrophy and progressive motor disturbance in mice expressing mutant (P301L) tau protein. Nat Genet 25, 402-5 (2000). 722. Oddo, S. et al. Triple-transgenic model of Alzheimer’s disease with plaques and tangles: intracellular Abeta and synaptic dysfunction. Neuron 39, 409-21 (2003). 723. Levinson, D. F. The Genetics of Depression: A Review. Biol Psychiatry (2005). 724. Meltzer, C. C. et al. Serotonin in aging, late-life depression, and Alzheimer’s disease: the emerg- ing role of functional imaging. Neuropsychopharmacology 18, 407-30 (1998). 725. Reinikainen, K. J., Soininen, H. & Riekkinen, P. J. Neurotransmitter changes in Alzheimer’s disease: implications to diagnostics and therapy. J Neurosci Res 27, 576-86 (1990).

208 references

726. Cross, A. J. Serotonin in Alzheimer-type dementia and other dementing illnesses. Ann N Y Acad Sci 600, 405-15; discussion 415-7 (1990). 727. Cheng, A. V. et al. Cortical serotonin-S2 receptor binding in Lewy body dementia, Alzheimer’s and Parkinson’s diseases. J Neurol Sci 106, 50-5 (1991). 728. Marcusson, J. O. et al. 5-Hydroxytryptamine-sensitive [3H]imipramine binding of protein nature in the human brain. II. Effect of normal aging and dementia disorders. Brain Res 425, 137-45 (1987). 729. Chen, C. P. et al. Presynaptic serotonergic markers in community-acquired cases of Alzheimer’s disease: correlations with depression and neuroleptic medication. J Neurochem 66, 1592-8 (1996). 730. Mossner, R. et al. The serotonin transporter in Alzheimer’s and Parkinson’s disease. J Neural Transm Suppl, 345-50 (2000). 731. Mossner, R. et al. Allelic variation of serotonin transporter expression is associated with depression in Parkinson’s disease. Mol Psychiatry 6, 350-2 (2001). 732. Vale, W., Spiess, J., Rivier, C. & Rivier, J. Characterization of a 41-residue ovine hypothalamic peptide that stimulates secretion of corticotropin and beta-endorphin. Science 213, 1394-7 (1981). 733. Chrousos, G. P. & Gold, P. W. The concepts of stress and stress system disorders. Overview of physical and behavioral homeostasis. Jama 267, 1244-52 (1992). 734. Keck, M. E. & Holsboer, F. Hyperactivity of CRH neuronal circuits as a target for therapeutic interventions in affective disorders. Peptides 22, 835-44 (2001). 735. Raghavendra, V., Kaur, G. & Kulkarni, S. K. Anti-depressant action of melatonin in chronic forced swimming-induced behavioral despair in mice, role of peripheral benzodiazepine receptor modulation. Eur Neuropsychopharmacol 10, 473-81 (2000). 736. Weibel, L. & Brandenberger, G. The start of the quiescent period of cortisol remains phase locked to the melatonin onset despite circadian phase alterations in humans working the night schedule. Neurosci Lett 318, 89-92 (2002). 737. Leproult, R., Colecchia, E. F., L’Hermite-Baleriaux, M. & Van Cauter, E. Transition from dim to bright light in the morning induces an immediate elevation of cortisol levels. J Clin Endocrinol Metab 86, 151-7 (2001). 738. Asayama, K. et al. Double blind study of melatonin effects on the sleep-wake rhythm, cognitive and non-cognitive functions in Alzheimer type dementia. J Nippon Med Sch 70 334-41 (2003). 739. Wu, Y. H. et al. Distribution of MT1 melatonin receptor immunoreactivity in the human hypothalamus and pituitary gland: Colocalization of MT1 with vasopressin, oxytocin, and corticotropin-releasing hormone. J Comp Neurol 499, 897-910 (2006). 740. Spengler, D., Rupprecht, R., Van, L. P. & Holsboer, F. Identification and characterization of a 3’,5’-cyclic adenosine monophosphate-responsive element in the human corticotropin-releasing hormone gene promoter. Mol Endocrinol 6, 1931-41 (1992). 741. Cheng, Y. H. et al. Glucocorticoid stimulation of corticotropin-releasing hormone gene expression requires a cyclic adenosine 3’,5’-monophosphate regulatory element in human primary placental cytotrophoblast cells. J Clin Endocrinol Metab 85, 1937-45 (2000). 742. Ni, X. et al. Estrogen receptor-mediated down-regulation of corticotropin-releasing hormone gene expression is dependent on a cyclic adenosine 3’,5’-monophosphate regulatory element in human placental syncytiotrophoblast cells. J Clin Endocrinol Metab 89, 2312-8 (2004). 743. Shepard, J. D., Liu, Y., Sassone-Corsi, P. & Aguilera, G. Role of glucocorticoids and cAMP- mediated repression in limiting corticotropin-releasing hormone transcription during stress. J Neurosci 25, 4073-81 (2005).

209 Summary

Circadian rhythms are an important aspect of human biology. Many of our behaviors and physiological and endocrine functions, such as the sleep-wake cycle, food intake, locomotor activity and melatonin and corticosterone release, have a daily rhythm, which is generated and synchronized to the 24 hour day/night cycle by the central biological clock in the brain, the suprachiasmatic nucleus (SCN). The rhythm of the pineal hormone melatonin is considered to be a precise and stable output signal of the SCN. It does not only provide the rest of the body with a precise perception of daily and seasonal time, but also feeds back on the SCN via melatonin receptors and is involved in the regulation of the SCN rhythmicity. Strong evidence indicates that aging is characterized by a progressive deterioration of circadian timekeeping. Circadian rhythm disruptions are frequently associated with poor sleep, reduced daytime alertness, and decreased cognitive performance. In Alzheimer’s disease these age-related disturbances in circadian rhythms are even more severe, and are the primary cause of institutionalization. Moreover, depression is also closely tied to circadian rhythm disorders because almost all depressive patients have sleep problems and feel worse at a particular time of day as well as in the winter. There is a great deal of evidence that circulating melatonin levels are decreased during aging, depression, and even more so in AD. Strikingly, our group recently revealed that melatonin levels in the cerebrospinal ﬂuid (CSF) are decreased from the earliest AD neuropathological stages (Braak stages) (i.e. Braak stage I-II) onwards, when subjects are still cognitively intact. This observation was conﬁrmed in our postmortem human pineal studies in the present thesis. We found that pineal melatonin content was indeed decreased already in Braak stages I-II compared to controls (Braak stage 0). Moreover, the diurnal rhythm of pineal melatonin content had disappeared in Braak stages I-II and VI (Chapter 2). Furthermore, pineal melatonin content appeared to be highly correlated with the CSF melatonin levels supporting our earlier studies (Chapter 2). These studies therefore suggest that decreased melatonin production is an early sign of AD. Since CSF melatonin levels are also highly correlated with peripheral melatonin levels such as in serum, it seems worthwhile to study in the future whether monitoring circulating melatonin levels may provide us with the earliest information of the AD neuropathology, which cannot be detected by any other clinical means. In Chapter 2 we also investigated the molecular mechanisms underlying the early reduced pineal melatonin productions during the AD process. Melatonin is mainly synthesized in the pineal gland from serotonin via two important enzymes, i.e. serotonin N-acetyltransferease (AA-NAT), which is the rate-limiting-enzyme, and hydroxyindole-o-methyltransferase (HIOMT). Melatonin synthesis is controlled by the SCN

210 via a multisynaptic pathway, which at night stimulates the release of noradrenalin (NA) acting on β1-adrenergic receptor (β1-ADR) of the pinealocyte to activate the enzyme AA-NAT and thus melatonin synthesis. We measured monoamines and mRNA levels of enzymes and β1-ADR in the melatonin synthesis and its noradrenergic regulatory pathway in pineal glands from controls (Braak 0), preclinical AD subjects (Braak I-II), and late clinical AD patients (Braak VI). We found that the daily rhythm of β1-ADR mRNA had disappeared in preclinical and clinical AD patients, which indicated a disrupted noradrenergic regulation of the melatonin synthesis by the SCN. Moreover, we have found an up-regulation of monoamine oxidase A (MAOA) expression and activity in the pineal gland of AD patients, which might deplete 5-HT, i.e. the precursor of melatonin. These changes may contribute to the reduced melatonin production and disrupted melatonin rhythm during the AD process. Increased MAOA is not only observed in the pineal gland of AD, but also found in many other brain regions and thus seems to be a general phenomenon in AD. Recently, a frequently occurring polymorphism of a variable number tandem repeats (VNTR) in the promoter region of MAOA was shown to affect MAOA transcriptional activity. In Chapter 3 we assessed the effects of the MAOA-VNTR polymorphism on MAOA transcriptional levels and MAOA activity in the pineal gland as endophenotype, as well as its involvement in the reduced melatonin production in AD. AD patients carrying a long MAOA-VNTR genotype showed higher MAOA gene expression and activity than the short-genotyped AD patients, suggesting that the MAOA-VNTR polymorphism is associated with the up-regulation of MAOA in AD patients. Interestingly, long-genotyped AD patients who had higher MAOA mRNA levels and activity did not show decreased melatonin production, compared to short-genotyped AD patients. This indicated that the up-regulation of MAOA is not a major factor for the reduction of melatonin production in AD. In other words, the dysfunction of the SCN-control may be more dominant for the melatonin changes in AD. Recently, the molecular components of the clock have been revealed to be a series of clock genes forming transcriptional-translational feedback loops. Previous studies on rodents have shown clock gene oscillation in the pineal gland. Moreover, the pineal clock genes Per1 and Cry2 are directly controlled by the SCN through a β-adrenergic cAMP signaling pathway, the same pathway that controls melatonin synthesis. In Chapter 4 we investigated whether and how the pineal molecular clock is affected in the AD process. We found a rhythmic expression of a number of clock genes (hBmal1, hCry1 and hPer1) in the human pineal gland of controls (Braak 0). Moreover, hPer1 and hβ1-ADR mRNA showed a similar daily expression pattern and were positively correlated. Interestingly, the diurnal rhythmic expression of hPer1, hCry1, and hBmal1 was lost in both preclinical (Braak I-II) and clinical AD (Braak V-VI), which suggests that pineal clock gene oscillation is disrupted very early on in the AD process. Moreo-

211 ver, the positive correlation between hPer1 and hβ1-ADR mRNA, which probably indicates SCN control of hPer1, was absent in both preclinical and clinical AD. Our hypothesis that a disrupted SCN control is responsible for pineal clock gene alterations during AD was strongly supported by our animal experimental data. The rat pineal that was deprived of SCN control showed alterations of clock gene expression that were remarkably similar to the changes we observed in the AD pineal. This holds for the loss of rhythmic clock gene expression, the loss of correlation between Per1 and β1-ADR mRNA, and in particular for the increased Cry1 mRNA levels. Our next aim was to study whether the activity of the SCN was decreased from the early AD stages onwards. AVP is a major rhythmic neuropeptide output of the SCN clockwork and regulates the rhythm of activity within the SCN and in other brain regions. AVP expression is strictly controlled by the molecular clock in the SCN, as its rhythm had disappeared and its mRNA levels were dramatically decreased in the SCN of Clock mutated mice. In Chapter 4 we therefore investigated whether the clock function of the SCN is disturbed early on in the AD process, by assessing the alterations of AVP mRNA levels in the SCN. Remarkably, we found that the AVP mRNA levels in the SCN were decreased in Braak stages I-II and Braak stages V-VI compared to those in Braak stage 0. These decreased AVP mRNA levels suggest that the SCN has a diminished output and a disrupted clock function from the earliest AD stages onwards. Moreover, this supports the possibility that the SCN control of the pineal gland is disturbed very early on in the AD process. Thus, a functional disconnection between the SCN and the pineal gland from the earliest AD stage onwards could account for the pineal clock gene and melatonin changes, and underlie the circadian rhythm disturbances in AD. Signaling through melatonin receptors, melatonin plays a role in numerous physiological processes, including circadian rhythms, sleep, stress and reproduction, many of which are based on the actions on the hypothalamus and pituitary. In Chapter 5 we described the distribution of the melatonin receptor MT1 in the human hypothalamus and pituitary by immunocytochemistry. MT1 immunoreactivity showed a widespread pattern in the hypothalamus, including the SCN, paraventricular nucleus (PVN), periventricular nucleus, supraoptic nucleus (SON), sexually dimorphic nucleus, the diagonal band of Broca, the nucleus basalis of Meynert, infundibular nucleus, ventromedial and dorsomedial nucleus, tuberomamillary nucleus, mamillary body and paraventricular thalamic nucleus. No staining was observed in the nucleus tuberalis lateralis and bed nucleus of the stria terminalis. The MT1 receptor was found to be colocalized with vasopressin in the SCN neurons, in parvocellular and magnocellular neurons of the PVN and SON, and with corticotropin-releasing hormone (CRH) in parvocellular neurons of the PVN. In the pituitary, strong MT1 expression was observed in the pars tuberalis, while a weak staining was found in the posterior and an-

212 terior pituitary. These findings provide a neurobiological basis for the participation of melatonin in the regulation of various hypothalamic processes and pituitary functions. The colocalization of MT1 and CRH suggests that melatonin might directly modulate the hypothalamus-pituitary-adrenal axis in the PVN, which may have implications for stress conditions such as depression. Melatonin is involved in the regulation of circadian rhythms both on peripheral levels and on the central level. The latter effect occurs by feeding back on the SCN via melatonin receptors. Melatonin levels are decreased in aged subjects, depressed patients and even more dramatically in Alzheimer’s disease (AD) patients. Supple- mentary melatonin is considered to be a potential treatment for aging, depression and AD-related circadian disorders. We determined the alterations of the MT1 melatonin receptor, compared to those of the neuropeptides vasopressin (AVP) and vasoactive intestinal peptide (VIP), in the SCN during aging and AD (Chapter 6), as well as in depression (Chapter 7) by immunocytochemistry. In Chapter 6 we found that the number and density of neurons in the SCN expressing neuropeptides such as vasopressin (AVP) and/or vasoactive intestinal peptide (VIP) did not change, but that the number and density of MT1-expressing neurons in the SCN were decreased in aged controls compared to young controls. This suggests that MT1 receptor expression is disrupted earlier than the expression of the neuropeptides AVP and VIP in the SCN during aging. Furthermore, both MT1-expressing neurons and AVP/VIP-expressing neurons were strongly diminished in the last neuropathological stages of AD (Braak stages V-VI), but not in the earliest stages (Braak stages I-II), compared to aged controls (Braak stage 0). Our study suggests that the MT1-mediated effects of melatonin on the SCN are disturbed during aging and even more so in late stage AD, which may contribute to the clinical circadian disorders and to the efficacy of therapeutic melatonin administration under these conditions. In Chapter 7 we found that MT1-immunoreactive neuron density and numbers in the central SCN of depressed patients were more than 1.8 times higher than those in controls. Moreover, the MT1- expressing neuron density and numbers were negatively correlated with the age at onset, while they were positively correlated with the disease duration in depressed patients. These results suggest an enhanced MT1 receptor availability in the SCN of depressed patients, which may be accumulatively increased during the course of disease. An enhanced efficacy of melatonin administration in depression might thus be expected. The present thesis suggests that a decreased activity of the SCN —already present at the moment of the occurrence of the very first tangles in the entorhinal cortex (Braak stage I)— most probably affects pineal molecular clock synchronization and pineal rhythmic melatonin synthesis. The loss of functional connectivity between the master clock (the SCN) and the peripheral slave clock (e.g. the pineal gland) may underlie

213 the circadian disturbances that occur in the course of AD. Apparently, the circadian system is extraordinarily vulnerable to AD pathogenesis. We propose that the circulating melatonin levels and their daily rhythmicity may thus provide information about the very ﬁrst AD stages that cannot be monitored in any other clinical way at this moment. This possibility should be tested in the future. The surprisingly early- stage decrease of SCN activity during the AD process demands further study of its underlying mechanisms. Moreover, it is also suggested that melatonin administration should be applied as early as possible in aged subjects complaining about sleep and cognition, and in depression considering the available number of melatonin receptors in the SCN in these conditions.

214 Samenvatting

Circadiane ritmes spelen een cruciale rol in ons functioneren en worden gegenereerd en gesynchroniseerd door de suprachiasmatic nucleus (SCN), de centrale biologische klok in de hersenen. Niet alleen ons gedrag, maar ook fysiologische lichaamsfuncties, zoals het slaap/waak ritme, voedselinname, en melatonine en corticosteron afgifte, vertonen een dagelijks ritme. Het ritme van het epifysehormoon melatonine wordt beschouwd als een nauwkeurig en stabiel output-signaal van de SCN, die het lichaam informatie verschaft over een precies tijdsbesef. Bovendien moduleert melatonine zelf het ritme van de SCN via de daar aanwezige melatonine receptoren. Vele studies hebben aangetoond dat veroudering onder meer gekenmerkt wordt door een sterke vermindering van het circadiane tijdsbesef. Verstoring van het circadiane ritme gaat samen met gebrekkige slaap, beperkte alertheid overdag, en verminderde cognitieve prestaties. Tijdens de ziekte van Alzheimer (AD) zijn deze leeftijd-gerelateerde circadiane ritmestoornissen nog sterker, en vormen de belangrijkste oorzaak voor ziekenhuisopname. Hiernaast gaat depressie vaak gepaard met circadiane ritmestoornissen; het merendeel van de depressieve patiënten lijdt aan slaapproblemen en ﬂuctuerende stemmingen gedurende de dag en per seizoen. Tevens is aangetoond dat de circulerende melatoninespiegels gereduceerd zijn bij veroudering en depressie, en zelfs sterk verminderd zijn bij AD. Bovendien heeft onderzoek van onze groep onlangs laten zien dat de melatonineniveaus in de cerebros- pinale vloeistof (CSF) al sterk verminderd zijn vanaf het vroegste neuropathologische AD stadium (in Braak stadium I-II), ondanks het feit dat deze personen cognitief nog normaal functioneren. Deze observatie werd bevestigd door onze postmortem studies van de menselijke epifyse (dit proefschrift) waarbij werd vastgesteld dat de epifyse inderdaad minder melatonine bevatte bij Braak stadium I-II patiënten dan die van de controles zonder enige AD neuropathologie (Braak stadium 0). Tevens bleek dat er bij Braak stadia I-II en VI een diurnaal ritme van de hoeveelheid melatonine in de epifyse afwezig was (Hoofdstuk 2). Onze resultaten suggereren dat een verminderde melatonineproductie een teken is van beginnende AD. Ook bleek uit ons onderzoek dat de melatoninespiegels in het CSF en in de epifyse sterk gecorreleerd zijn. Aangezien de CSF melatoninespiegels eveneens sterk correleren met perifere melatoninespiegels zoals in serum, is het belangrijk om in toekomstige studies na te gaan of circulerende melatonineniveaus een vroege indicatie geven voor neuropathologische veranderingen tijdens AD, die tot op heden niet op een andere wijze kunnen worden opgespoord. In Hoofdstuk 2 werden de moleculaire mechanismen onderzocht die ten grondslag liggen aan de vroeg optredende, verminderde melatonineproductie tijdens het Alzheimer proces. Melatonine wordt in de epifyse voornamelijk gesynthetiseerd uit

215 serotonine door twee belangrijke enzymen, t.w. serotonine N-acetyltransferase (AA- NAT), het snelheids-limiterende enzym voor melatonineproductie, en hydroxyindole O-methyltransferase (HIOMT). Melatoninesynthese in de epifyse wordt gereguleerd door de SCN via een multi-synaptische neuronale verbinding die ’s nachts de afgifte van noradrenaline (NA) stimuleert, waardoor de β1-adrenerge receptor (β1-ADR) van de pinealocyte het enzym AA-NAT activeert en daarmee ook de melatoninesynthese. We hebben de hoeveelheid monoamines en mRNA niveaus van betrokken enzymen en van β1-ADR in de epifyse van controles (Braak 0), preklinische Alzhei- mer-patiënten (Braak I-II), en klinische eindstadium Alzheimer-patiënten (Braak VI) onderzocht. Hieruit bleek dat het diurnale ritme van β1-ADR mRNA was verdwenen in preklinische en in klinische Alzheimer-patiënten, hetgeen duidt op een verstoorde noradrenerge regulatie van de melatoninesynthese door de SCN. Bovendien werd een opregulatie van expressie en activiteit van monoamine oxidase A (MAOA) in de epifyse van AD-patiënten gevonden, die een vermindering kan veroorzaken van 5-HT, de voorloper in de synthese van melatonine. Deze veranderingen lijken de oorzaak te zijn van de verminderde melatonineproductie en het verstoorde melatonineritme tijdens het AD proces. Verhoogd MAOA werd niet alleen aangetroffen in de epifyse van AD-patiënten maar ook in diverse andere hersenengebieden en kan, als zodanig, worden beschouwd als een algemeen verschijnsel tijdens AD. Tevens werd onlangs aangetoond dat een veel- vuldig voorkomend polymorfisme van een ‘variable number tandem repeats’ (VNTR) in het promoter gebied van het MAOA gen de MAOA transcriptie beïnvloedt. In Hoofdstuk 3 werden niet alleen de effecten bestudeerd van het MAOA-VNTR polymorfisme op de MAOA transcriptie-niveaus en de MAOA activiteit in de epifyse als endofenotype, maar ook de betrokkenheid hiervan bij de verminderde melatonineproductie in AD. Alzheimer-patiënten met een lang MAOA-VNTR genotype vertoonden een hogere MAOA genexpressie en activiteit dan AD patiënten met een kort MAOA- VNTR genotype. Dit zou er op kunnen wijzen dat het MAOA-VNTR polymorfisme geassocieerd is met de opregulatie van MAOA bij AD patiënten. Het was frappant dat Alzheimer-patiënten met een lang MAOA-VNTR genotype en met een hoog MAOA mRNA niveau en activiteit geen verminderde melatonineproductie lieten zien, vergeleken met AD patiënten met een kort MAOA-VNTR genotype. Hieruit bleek dat MAOA opregulatie geen beslissende factor is voor de reductie van melatonineproductie bij AD. Op basis van deze bevindingen kan worden aangenomen dat een dysfunctie van de SCN van grotere invloed is op het optreden van melatonineveranderingen in AD dan MAOA opregulatie. Recent onderzoek heeft aangetoond dat de moleculaire componenten van de biologische klok bestaan uit een serie klok-genen die tezamen de transcriptionele-translati- onele terugkoppelings circuits vormen. Eerdere studies bij knaagdieren hebben laten

216 zien dat er in de epifyse sprake is van een oscillatie van deze klok-genen. Bovendien worden in de epifyse de klok-genen Per1 en Cry2 direct gecontroleerd door de SCN via een β-adrenerge cAMP signaalroute, hetzelfde pad dat ook de melatoninesynthese reguleert. In Hoofdstuk 4 werd beschreven of en op welke wijze de moleculaire klok in de epifyse wordt aangetast tijdens het Alzheimer-ziekteproces. Zo werd een ritmische expressie van een aantal klok-genen (hPer1, hCry1 en hPer1) gevonden in de epifyse van controles (Braak 0). Tevens bleken hPer1 en hβ1-ADR mRNA niet alleen hetzelfde diurnale expressiepatroon te vertonen maar was er ook sprake van een positieve correlatie tussen deze twee genexpressie-niveaus die zeer waarschijnlijk duidt op een SCN-controle van hPer1 expressie. Opmerkelijk was bovendien dat er zowel in preklinische (Braak I-II) als klinische (Braak V-VI) eindstadium-AD-patiënten geen ritmische expressie werd gevonden van hPer1, hCry1, en hBmal, wat suggereert dat de klok-gen oscillatie van de epifyse al in een zeer vroeg stadium van AD wordt verstoord. Hiernaast was in beide groepen patiënten de positieve correlatie tussen hPer1 en hβ1-ADR-mRNA eveneens verstoord. Onze hypothese dat een verstoorde SCN-controle verantwoordelijk is voor veranderingen van de klok-genen in de epifyse bij AD werd bevestigd door onze resultaten uit dierexperimenteel onderzoek. Zo bleken in de epifyse van de rat zonder SCN-regulatie bijna soortgelijke veranderingen in klok-gen-expressie plaats te vinden als in de epifyse van AD-patiënten: het verlies van het dag/nachtritme in klok-gen-expressie, het verdwijnen van de correlatie tussen Per1 en β1-ADR mRNA en in het bijzonder de verhoogde hoeveelheid Cry1 mRNA. Vervolgonderzoek richtte zich op de vraag of de SCN activiteit afneemt vanaf de vroegste stadia van AD. Het neuropeptide arginine-vasopressine (AVP) is een belangrijke ritmische output van de SCN-klok en reguleert het ritme van de activiteit in de SCN en andere hersen- gebieden. De AVP-expressie wordt nauwkeurig gecontroleerd door de moleculaire klok in de SCN, zoals bleek uit de afwezigheid van het AVP-ritme en de sterk ver- laagde AVP-mRNA-niveaus in de SCN van klok-gemuteerde muizen. In Hoofdstuk 4 werd onderzocht of de klokfunctie van de SCN al in een vroeg stadium van AD is verstoord, waarbij de veranderingen die optreden in de AVP-mRNA-niveaus in de SCN in kaart werden gebracht. Opvallend was dat de AVP-mRNA-niveaus in de SCN reeds laag waren in Braak-stadia I-II en laag bleven in Braak-stadia V-VI vergeleken met Braak-stadium 0. Hieruit bleek dat de SCN reeds vanaf de vroegste stadia van AD zowel een verminderde output als een verstoorde klokfunctie heeft. Bovendien ondersteunen deze bevindingen onze hypothese dat de SCN-regulatie van de epifyse al in een vroeg stadium van AD is verstoord. Een functionele disconnectie tussen de SCN en de epifyse in het vroegste AD-stadium zou niet alleen de oorzaak kunnen zijn van klok-genen melatonineveranderingen in de epifyse, maar ook ten grondslag kunnen liggen aan de circadiane ritmestoornissen bij AD.

217 Door de signaalfunctie van melatoninereceptoren speelt melatonine tevens een rol bij talloze fysiologische processen, waaronder ritmes in hormonen, slaap, stress en voortplanting. De meeste van deze processen worden aangestuurd door de hypothalamus en hypofyse. In Hoofdstuk 5 wordt een beschrijving gegeven van de distributie van de melatoninereceptor MT1 in de hypothalamus en de hypofyse van de mens, waarbij gebruik gemaakt werd van immuunhistochemische technieken. MT1 was gelocaliseerd in de hypothalame kerngebieden: de SCN, nucleus paraventricularis (PVN), nucleus periventricularis, nucleus supraopticus (SON), de seksueel dimorfe kern, de diagonale band van Broca, de nucleus basalis van Meynert, de nucleus infun- dibularis, ventromediale en dorsomediale nucleus, tuberomamillary nucleus, nucleus corporis mamillaris en de nucleus paraventricularis van de thalamus. Er werd geen kleuring gevonden in de nucleus tuberalis lateralis en de bed-nucleus van de stria terminalis. De MT1 receptor vertoonde colocalisatie met vasopressine in SCN-neuronen, in de parvocellulaire en magnocellulaire neuronen van de PVN en SON, en met de corticotropine-releasing hormoon (CRH) bevattende parvocellulaire neuronen van de PVN. In de pars tuberalis van de hypofyse werd een sterke MT1-expressie waarge- nomen, in tegenstelling tot de voor- en achterkwab van de hypofyse. Deze resultaten suggereren dat melatonine een rol speelt bij de regulatie van verschillende hypothalame processen en hypofysefuncties. Colocalisatie van MT1 en CRH geeft eveneens aan dat melatonine een modulerende functie heeft in de regulatie van de HPA-as in de PVN, wat implicaties kan hebben voor stresscondities zoals depressie. Melatonine is betrokken bij de regulatie van circadiane ritmes zowel op het perifere als het centrale niveau. Eﬀecten op centraal niveau worden veroorzaakt door de feedback op de SCN via de melatoninereceptoren. Gebleken is dat de melatoninespiegels verlaagd zijn bij oudere personen en depressieve patiënten, terwijl bij AD-patiënten de melatoninespiegels zelfs sterk verlaagd zijn. Daarom wordt de toediening van melatonine gezien als een potentiële eﬀectieve behandeling van verouderings-, depressie- en Alzheimer-gerelateerde circadiane stoornissen. Wij hebben met immuunhistochemische technieken de veranderingen onderzocht in de eiwit-expressie-niveaus van de MT1-melatonine-receptor in de SCN en vergeleken met veranderingen in de neuropeptides vasopressine (AVP) en vasoactief intestinaal peptide (VIP) gedurende veroudering en AD (Hoofdstuk 6), en bij depressie (Hoofdstuk 7). In Hoofdstuk 6 tonen wij aan dat het aantal en de dichtheid van SCN-neuronen die de neuropeptides vasopressine (AVP) en/of vasoactief intestinaal peptide (VIP) bevatten, niet verandert. Echter, het aantal en de dichtheid van SCN-neuronen die MT1 bevatten in de groep oude controles was verlaagd vergeleken bij de groep jonge controles. Dit zou er op kunnen wijzen dat tijdens veroudering de expressie van de MT1-receptor in de SCN in een eerder stadium verminderd is dan de expressie van de neuropeptides AVP en VIP. Hiernaast werd aangetoond dat het aantal neuronen dat zowel MT1 als AVP/VIP

218 bevat sterk verminderd was in de late neuropathologische Braak-stadia (Braak V-VI), maar niet in de vroege stadia (Braak I-II) vergeleken met de groep oude controles (Braak-stadium 0). Ons onderzoek laat zien dat de melatonine-effecten op de SCN, gemedieerd door MT1, verstoord zijn tijdens veroudering en ook in sterkere mate in de late stadia van AD, waardoor er klinisch relevante circadiane stoornissen ontstaan. Toediening van melatonine zou onder deze omstandigheden een effectieve behandeling kunnen zijn. In Hoofdstuk 7 wordt beschreven dat de dichtheid en het aantal MT1-immuunreactieve neuronen in de centrale SCN van depressieve patiënten meer dan 1,8 maal hoger was dan in controles. Tevens bleek er een negatieve correlatie te zijn met de leeftijd waarop de depressie zich openbaarde, terwijl er juist een positieve correlatie was met de duur van de ziekte. Deze resultaten duiden op een verhoogde beschikbaarheid van MT1-receptoren in de SCN bij depressieve patiënten, die toe- neemt naarmate de ziekte langer duurt. Therapeutische toediening van melatonine aan depressieve patiënten zou wellicht een effectieve behandelingsmethode kunnen zijn. In dit proefschrift wordt aangetoond dat een verminderde activiteit van de SCN reeds aanwezig is vanaf het moment dat de eerste tangles in de entorhinale cortex worden aangetoond (Braak-stadium I). Dit beïnvloedt niet alleen de synchronisatie van de moleculaire klok maar ook het ritme van de melatoninesynthese in de epifyse. Het verlies van functionele connectiviteit tussen de centrale (meester-)klok (de SCN) en de perifere (slaaf-)klokken (bijvoorbeeld de epifyse) lijkt ten grondslag te liggen aan de circadiane stoornissen die optreden tijdens het Alzheimer-proces. Blijkbaar is het circadiane systeem bijzonder kwetsbaar tijdens het ontstaan van de ziekte van Alzheimer. Wij anticiperen dat toekomstig onderzoek naar het beloop van de circulerende melatoninespiegels en hun dagelijkse ritmiek nader inzicht op kan leveren over de vroegste AD-stadia, die tot op heden niet klinisch geïdentificeerd kunnen worden tijdens het leven. Nader onderzoek zal de mechanismen verduidelijken die ten grondslag liggen aan de opmerkelijke afname van SCN-activiteit in het vroege stadium van AD. Het lijkt van essentieel belang zo vroeg mogelijk te beginnen met de toediening van melatonine aan de groep ouderen met klachten over slaap en cognitie en aan depressie-patiënten, gezien de aanwezigheid van melatoninereceptoren in de SCN in deze groep patiënten.

219 List of Publications

Wu, Y. H., Feenstra, M. G., Zhou, J. N., Liu, R. Y., Torano, J. S., Van Kan, H. J., Fischer, D. F., Ravid, R., and Swaab, D. F. (2003) Molecular changes underlying reduced pineal melatonin levels in Alzheimer disease: alterations in preclinical and clinical stages. J Clin Endocrinol Metab 88, 5898-5906 Wu, Y. H., and Swaab, D. F. (2005) The human pineal gland and melatonin in aging and Alzheimer’s disease. J Pineal Res 38, 145-152 Bao, A. M., Fischer, D. F., Wu, Y. H., Hol, E. M., Balesar, R., Unmehopa, U. A., Zhou, J. N., and Swaab, D. F. (2006) A direct androgenic involvement in the expression of human corticotropin-releasing hormone. Molecular Psychiatry 11, 567-576

Wu Y.H., Zhou J.N., Bao A.M., Jockers R., Van Heerikhuize J., Swaab D.F. (2006) Increased number of neurons expressing melatonin receptor MT1 in the suprachiasmatic nucleus in depression, and its relation to age at onset and disease duration. Neurobiology of Aging 2006, Jul 10; [Epub ahead of print] Wu Y.H., Fischer D.F., Kalsbeek A., Garidou-Boof M.L., Van Der Vliet J., Van Heij- ningen C., Liu Y., Zhou J.N., Swaab D.F.(2006). Pineal molecular clock oscillation is disturbed in Alzheimer disease, due to a functional disconnection with the “master clock”. The FASEB Journal 20:1874-1876.

Wu Y.H., Zhou J.N., Balesar R., Bao A.M., Unmehopa U.A., Jockers R., Van Heerikhu- ize J., Swaab D.F. (2006) Distribution of MT1 melatonin receptor immunoreactivity in the human hypothalamus and pituitary gland: colocalization of MT1 with vasopressin, oxytocin, and corticotropin-releasing hormone. J Comparative Neurology 499:897-910. Wu Y.H., Swaab D.F. Disturbance and strategies for reactivation of the circadian rhythm system in aging and Alzheimer’s disease. Sleep Medicine, accepted, 2006. Wu Y.H., Fischer D.F., Zhou J.N., Swaab D.F. A promoter polymorphism in the monoamine oxidase A gene is associated with the increased MAOA activity in Alzhe- imer’s disease. Submitted. Wu Y.H., Zhou J.N., Bao A.M., Jockers R., Van Heerikhuize, J., Swaab D.F. Increased number of neurons expressing melatonin receptor MT1 in the suprachiasmatic nucleus in depression, and its relation to age at onset and disease duration. Submitted.

220 Abbreviations

5-HT serotonin 5-HIAA 5-hydroxyindoleacetic acid ASPS advanced sleep phase syndrome AVP arginine vasopressin AD Alzheimer’s disease ApoE apolipoprotein E beta-ADR beta-adrenergic receptor Bmal brain and muscle aryl hydrocarbon receptor nuclear translocator (ARNT)-like BST bed nucleus of the stria terminalis CCGs clock controlled genes CKI casein kinase I CREB cAMP response element binding protein CREs cAMP-responsive elements CRH corticotropin-releasing hormone Cry Cryptochrome CSF cerebrospinal fluid DA dopamine DBB diagonal band of Broca or vertical limb of the diagonal band of Broca DBP D-element binding protein DMN dorsomedial hypothalamic nucleus DSPS delayed sleep phase syndrome GABA Gamma-amino-butyric acid GFAP glial fibrillary acidic protein GRP gastrin-releasing peptide hDBB horizontal limb of the diagonal band of Broca HIOMT hydroxyindole-O-methyltransferase HPA hypothalamus-pituitary-adrenal HPLC: high performance liquid chromatography HVA homovanilic acid IGL the intergeniculate leaflet IML interomediolateral cell INF infundibular or arcuate nucleus IR immunoreactivity

222 LHA lateral hypothalamic area LH pulsatile luteinizing hormone LS lateral septum MB mamillary body NBM nucleus basalis of Meynert MAOA monoamine oxidase A ME median eminence MHPG 3-methoxy-4-hydroxyphenylglycol MT1 melatonin receptor 1a MT2 melatonin receptor 1b NA noradrenalin NAT N-acetyltransferase NTL nucleus tuberalis lateralis NPY neuropeptide Y OXT oxytocin PACAP pituitary adenylate cyclase-activating polypeptide Per period gene PeVN periventricular nucleus PHI peptide histidine isoleucine PMD: postmortem delay PT pars tuberalis PVT paraventricular nucleus of thalamus PVN paraventricular nucleus RHT retinohypothalamic tract RIA radioimmunoassay SDN-POA sexually dimorphic nucleus of the preoptic area or intermediate nucleus of the anterior hypothalamus-1 SCN suprachiasmatic nucleus SCG superior cervical ganglia SOM somatostatin SON supraoptic nucleus TM tuberomamillary nucleus TPH tryptophan hydroxylase Trp tryptophan VIP vasoactive intestinal peptide VMN ventromedial hypothalamic nucleus. VNTR variable number tandem repeats

223