Neuroendocrinology of Mood

334
Current Topics in Neuroendocrinology Volume 8 Editors Detlev Ganten, Heidelberg· Donald Pfaff, New York Editorial Board Yasumasa Arai" Tokyo . Kjell Fuxe, Stockholm Hiroo Imura, Kyoto . Brian Pickering, Bristol Gunter Stock, Berlin

Transcript of Neuroendocrinology of Mood

Editors
Editorial Board
Yasumasa Arai" Tokyo . Kjell Fuxe, Stockholm Hiroo Imura, Kyoto . Brian Pickering, Bristol Gunter Stock, Berlin
Current Topics in Neuroendocrinology
Vol. 1: Sleep. Clinical and Experimental Aspects 1982.47 figures. VII, 129 pages. DM 95,-. ISBN 3-540-11125-5 Neurobiology of REM Sleep. A Possible Role for Dopamine . Endocrine and Peptide Functions in the Sleep-Waking Cycle· Sleep Regulation: Circadian Rhythm and Homeo­ stasis· Haemodynamic Changes During Sleep in Man
Vol. 2: Adrenal Actions on Brain 1982.25 figures. V, 153 pages. DM 95,-. ISBN 3-540-11126-3 Glucocorticoids and Hippocampus: Receptors in Search of a Function· Glucocorticoids and the Developing Nervous System· Feedback Actions of Adrenal Steroid Hormones· Electro­ physiologic and Clinical Aspects of Glucocorticoids on Certain Neural Systems . Adrenal Steroids and Behavioral Adaptation: Relationship to Brain Corticoid Receptors
Vol. 3: Central Cardiovascular Control. Basic and Clinical Aspects 1983.71 figures. V, 192 pages. DM 106,-. ISBN 3-540-11350-9 Functional and Anatomic Aspects of Central Nervous Cardiovascular Regulation· Autonomic Nervous System and Blood Pressure Control in Normotensive and Hypertensive Conditions· Reflex Control of Circulation in Normotensive and Hypertensive Humans· Corticotropin­ Releasing Factor: Central Nervous System Effects on the Sympathetic Nervous System and Cardiovascular Regulation· Neuropeptides and Central Blood Pressure Regulation· Centrally Acting Drugs as a Tool to Study Central Mechanisms of Blood Pressure Control· The Blood­ Brain Barrier and its Role in the Control of Circulating Hormone Effects on the Brain
Vol. 4: Neurobiology of Vasopressin 1985.53 figures. V, 203 pages. DM 98,-. ISBN 3-540-11351-7 Biosynthesis of Vasopressin· Electrophysiological Studies of the Magnocellular Neurons· Volume Regulation of Antidiuretic Hormone Secretion . Vasopressin, Cardiovascular Regulation and Hypertension· Neuroanatomical Pathways Related to Vasopressin
Vol. 5: Actions of Progesterone on the Brain 1985. 61 figures. V, 216 pages. DM 98,-. ISBN 3-540-13433-6 Antiprogestins . Progesterone Receptors in Brain and Hypophysis· Effects of Estrogen and Progesterone as Revealed by Neurophysiological Methods· Progesterone Receptors in CNS Correlated with Reproductive Behavior· Estradiol-Progesterone Interactions in the Repro­ ductive Behavior of Female Rats· Behavioral Effects on Humans of Progesterone-Related Compounds During Development and in the Adult
Vol. 6: Neurobiology of Oxytocin 1986.38 figures. V, 175 pages. DM 98,-. ISBN 3-540-15341-1 Biosynthesis of Oxytocin in the Brain and Peripheral Organs· Regulation of Oxytocin Release· Proteolytic Conversion of Oxytocin, Vasopressin, and Related Peptides in the Brain· Oxytocin and Behavior· Oxytocin as an Ovarian Hormone· Oxytocin and the Milk-Ejection Reflex
Vol. 7: Morpbology of Hypothalamus and Its Connections 1986. 94 figures. VI, 314 pages. DM 148,-. ISBN 3-540-16919-9 The Pitnitary Portal System . Functional Ultrastructure of Gonadotropes: a Review . Ultrastructure of Anterior Pituitary Cells . Catecholamine-Peptide Interactions in the Hypothalamus . Neuroendocrine Projections to the Median Eminence . Afferents onto Neuroendocrine Cells· Interconnectedness of Steroid Hormone-Binding Neurons: Existence and Implications· Ultrastructure of Regulatory Neuroendocrine Neurons and Functionally Related Structures· Synaptogenesis and Neuronal Plasticity to Gonadal Steroids: Implicatious for the Development of Sexual Dimorphism in the Neuroendocrine Brain
D. Ganten and D. Pfaff (Eds.)
Neuroendocrinology o/Mood Coeditor
Contributors
L. F. Agnati, H. Agren, M. Aronsson, M. S. Bauer, G. P. Chrousos A. Cintra, T. 1. Crow, M. A. Demitrack, M.1. Devlin, 1. N. Ferrier K. Fuxe, P. W. Gold, R. N. Golden, 1.-A. Gustafsson A. Harfstrand, D. S. Janowsky, K. Kalogeras, M. A. Kling B. Levant, P. Linkowski, D. L. Loriaux, N. Matussek H.Y. Meltzer, 1. Mendlewicz, 1. F. Nash Jr., C. B. Nemeroff R. M. Post, S. C. Risch, D. R. Rubinow, L. Terenius L. Traskman-Bendz, B. T. Walsh, S. R. B. Weiss, H. Whitfield P. C. Whybrow, F. A. Wiesel, M. Zoli
With 80 Figures
Editors
Dr. DETLEV OANTEN, M.D., Ph.D. Pharmakologisches Institut Universitiit Heidelberg 1m Neuenheimer Feld 366 6900 Heidelberg/FRO
Dr. DONALD PFAFF, Ph.D. Rockefeller University York Avenue, and 66th Street New York, NY 10021jUSA
Coeditor
Karolinska Institute P.O. Box 60400 10401 Stockholm, Sweden
The picture on the cover has been taken from Nieuwenhuys R., Voogd J., van Huijzen Chr.: The Human Central Nervous System. 2nd Edition. Springer·Veriag Berlin Heidelberg New York 1981
ISBN-13: 978-3-642-72740-5
DOl: 10.1007/978-3-642-72738-2
e-ISBN-13: 978-3-642-72738-2
Library of Congress Cataloging in Publication Data. Neuroendocrinology of mood. (Current topics in neuroendocrinology; v. 8) Includes bibliographies and index. 1. Mood (psychology) - Physiological aspects. 2. Neuroendocrinology. 3. Neurotransmitters. 4. Affective disorders - Physiological aspects. I. Ganten, D. (Detlev), 1941-. II. PfatT, Donald W., 1939-. III. Fuxe, Kjell. IV. Agnati,Luigi Francesco. V. Series. QP401.N37 1988 616.89'071 88-4897
This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in other ways, and storage in data banks. Duplication of this publication or parts thereofis only permitted under the provisions ofthe German Copyright Law of September 9, 1965, in its version of June 24, 1985, and a copyright fee must always be paid. Violations fall under the prosecution act of the German Copyright Law.
© Springer-Verlag Berlin Heidelberg 1988
Softcover reprint of the hardcover I st edition 1988
The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use.
Product liability: The publisher can give no gnarantee for information about drug dosage and application thereof contained in this book. In every individual case the respective user must check its accuracy by consulting other pharmaceutical literature.
Typesetting, printing and binding: Briihlsche Universitiitsdruckerei, Giessen 2121/3145-543210
Table of Contents
Principles for the Hormone Regulation of Wiring Transmission and Volume Transmission in the Central Nervous System By K. Fuxe, L. F. Agnati, A. Harfstrand, A. Cintra, M. Aronsson, M. Zoli, and J.-A. Gustafsson With 35 Figures . . . . . . . . . . . . . . . . 1
Clinical Studies with Corticotropin Releasing Hormone: Implications for Hypothalamic-Pituitary-Adrenal Dysfunction in Depression and Related Disorders By P. W. Gold, M.A. Kling, M.A. Demitrack, H. Whitfield, K. Kalogeras, D. L. Loriaux, and G. P. Chrousos With 11 Figures . . . . . . . . . . . . . . . . . . . . 55
Biological Rhythms and Mood Disorders By J. Mendlewicz and P. Linkowski With 3 Figures. . . . . . . . . . . . . . . . . . . . . 79
Recurrent Affective Disorders: Lessons from Limbic Kindling By R. M. Post, S. R. B. Weiss, and D. R. Rubinow With 14 Figures . . . . . . . . . . . . . . . . . . . . 91
The Mechanisms of Action of Antipsychotics and Antidepressant Drugs By F.A. Wiesel and L. Traskman-Bendz. . . .. . ... 117
Catechohimines and Mood: Neuroendocrine Aspects By N. Matussek With 6 Figures. . . . . . . . . . . . . . . . . . 141
Serotonin and Mood: Neuroendocrine Aspects By H. Y. Meltzer and J. F. Nash Jr.. . . . .
Cholinergic Mechanisms in Mood: Neuroendocrine Aspects
. 183
By D. S. Janowsky, R. N. Golden, and S. C. Risch . . . . . 211
The Psychobiology of Neurotensin By B. Levant and C. B. Nemeroff With 6 Figures. . . . . . . . . . . . . . . . . . . . . 231
VI Table of Contents
Cholecystokinin and Mood By 1. N. Ferrier and T. J. Crow. . ... 263
Opioid Peptides and Mood: Neuroendocrine Aspects By H. Agren and L. Terenius . . . . . . . . . . . 273
The Neuroendocrinology of Anorexia Nervosa By M.J. Devlin and B. T. Walsh With 5 Figures. . . . . . . . . . . . . . . . . . . . . 291
Effects of Peripheral Thyroid Hormones on the Central Nervous System: Relevance to Disorders of Mood By P. C. Whybrow and M. S. Bauer. 309
Subject Index . . . . . . . . . . 329
Principles for the Hormone Regulation of Wiring Transmission and Volume Transmission in the Central Nervous System * K. Fuxe 1, L. F. Agnati 2, A. Harfstrand 1, A. Cintra 1, M. Aronsson 1
M. Zoli 2 and J .-A. Gustafsson 3
Contents
1 Introduction . 2 Humoral Modulation of Wiring Transmission . . . 7 3 Actions of Gonadal Steroids on Wiring Transmission 7
3.1 General Aspects . . . . . . . . . . . . . . 7 3.2 Studies on Presynaptic Features of Monoamine Neurons. 10 3.3 Studies on Monoamine Receptor Mechanisms . 12
4 Actions of Glucocorticoids on Wiring Transmission. . . . . 15 4.1 General Aspects . . . . . . . . . . . . . . . . . . 15 4.2 Morphometric and Microdensitometric Analysis of GR Immunoreactivity
in the Central Nervous System . . . . . . . . . . . . 26 4.3 Studies on Presynaptic Features of Monoamine Neurons. 30 4.4 Studies on Monoamine Receptor Mechanisms . . 32
5 Actions of Thyroid Hormones on Wiring transmission. . . . 34 5.1 General Aspects . . . . . . . . . . . . . . . . . . 34 5.2 Studies on Presynaptic Features of Monoamine Neurons . 35 5.3 Studies on Monoamine Receptor Mechanisms 36
6 The Humoral Modulation of Volume Transmission 38 7 Aspects on the Organization Principles of the CNS 38
7.1 Modules of Wiring Transmission . . . . . . 40 7.2 Modules of Volume Transmission. . . . . . 41 7.3 Functional Aspects on the Modular Organization 41
8 Summary 43 References. . . . . . . . . . . . . . . . . . . . 47
1 Introduction
We have recently suggested the existence in the central nervous system of two types of electrochemical transmission, namely wiring transmission (WT) and volume transmission (VT) (see Agnati et al. 1986a, b). The concepts are summarized in Tables 1 and 2. VT is a humoral type of chemical transmission. However, it
* This work has been supported by a grant (04X-715) from the Swedish Medical Research Council, a grant (MH25504) from the NIH, a grant from the Wallenberg Foundation, and CNR-I, MDI grants
1 Department of Histology and Neurobiology, Karolinska Institute, P.O. Box 60400, S-10401 Stockholm, Sweden
2 Department of Human Physiology, University of Modena, Modena, Italy 3 Department of Medical Nutrition, Huddinge Hospital, Huddinge, Sweden
Current Topics in Neuroendocrinology, Vol. 8 © Springer-Verlag Berlin Heidelberg 1988
T ab
le 1
• TltAN.MI~ltCTl
DltECE~(lIl
/ / '//' / .L/"'/
-1 ~ ~ / ,~
_U.w;tIO ...
Fig. I. Integrative features of the chemical synapse. The multiple transmission lines of the intramembrane receptor-receptor interactions are illustrated. The local circuit is shown as an electrometabolic unit
consists not only of humoral and paracrine signals, diffusing in the extracellular fluid to reach the appropriate receptors, but also of electrotonic signals, which also operate in the extracellular fluid. In fact, the extracellular space of the brain consti­ tutes a restricted microenvironment. Thus, ion fluxes across cellular membranes can induce substantial changes in the ion composition. These ionic fluctuations in the extracellular fluid and the ionic fluxes from sources to sinks may represent
Hormones and Synaptic Transmission 5
7777G~==·=----![9777ZZ I REC£PTOA- RECEPTOR INTERACTION I
Fig. 2. The n-dimensional representation of synaptic transmission
signals for communication between neural groups (Nicholson 1980). In Table 1 the possible role of glia and neurons in WT and VT is summarized. In VT the glial cells control the extracellular fluid ion composition and the shaping of the extra­ cellular fluid pathways (i.e., the communication channels between neural groups) for signal diffusion as well as the release, uptake and metabolism of humoral and paracrine signals. With regard to the function of neurons in VT they represent the location of sources and sinks for electrotonic signals and the sites of release and recognition of humoral and paracrine signals. From a biochemical standpoint the neurons control the sources and sinks of electrotonic signals and are involved in the uptake, release and metabolism of humoral and paracrine signals. When we focus our attention on chemical signals in WT and VT it is possible to recognize some main differential features. Thus, as seen in Table 2, the VT is characterized by low speed and long-term action, a high degree of divergence and plasticity and low safety of the transmission process. WT is the classical type of transmission which is neuron-linked and operates with high speed and safety and short-term actions, the divergency and plasticity being low. It seems clear that the integrative capability of the central and peripheral nervous system is increased by the pres­ ence of VT, which is not submitted to neuroanatomical constraints and may af­ fect the computing charateristics of the neuronal networks.
In order to understand the actions of hormonal and paracrine signals on WT it is important to emphasize that the synapse is now regarded as a highly complex
6 K. Fuxe et al.
- - - - - - - - - 1- - - - - - - - - - .. - - -- --.
, '~ " e
PRESYNAPTIC I SIGNAL
.. z
i~ rJ 2!:! ;:: ...... 2" - rn .. if ffi ~~ ~ 02 a: a: .. " >-% % Z IJ IJ o IJ
INTRACELLULAR
~ MACHINERY
~ FILTERING AND INTEGRATION OF SIGNALS
RESET OFTHE SENSITIVITY AND OF THE INTEGRATIVE FEATURES OF RECOGNITION AND DECODING SYSTE MS FOR EXTRACELLULAR SIGNALS
Fig. 3. Possible interaction between volume transmission and wiring transmission at membrane as well as at intracellular levels. At the intracellular level different targets for the modulatory interactions can be considered, such as rn phosphorylation/dephosphorylation of the receptor complex; [2] cyclic nucleotide cascades; [3] lipid composition and polarization of the membrane
electrometabolic integrative unit (see Fig. 1). It consists of multiple transmission lines (Agnati et al. 1983, 1984a; Fuxe et al. 1984a), which interact with one an­ other at the pre- and postsynaptic membrane via intramembrane receptor-recep­ tor interactions (Fuxe and Agnati 1985) and via postreceptor intracellular signals. Furthermore, the ionic and metabolic responses participate in the integrative ac­ tivity of the synaptic membranes. In Fig. 2 the complexity of the integration is il­ lustrated at the membrane level, where filtration and integration signals take place. The intracytoplasmatic mechanisms control the recognition sites and the decoding mechanisms in the membrane for extracellular fluid signals, in this way, for example, resetting sensitivity in the integrative capacity of the receptor mech­ anisms. Hormonal and paracrine signals can directly modulate the receptor charac­ teristics or the receptor-receptor interactions in the membrane (Fig. 3). The intra­ cellular machinery and its short- and long-term regulation of the receptor mech­ anisms is also influenced, probably mainly via nuclear actions, at least as far as steroid and thyroid hormones are concerned (see Fuxe et al. 1981 b). The electro­ tonic signals in the VT control the membrane polarization, in this way influencing the opening and closure of ionic channels. Changes in membrane polarization
Hormones and Synaptic Transmission 7
probably lead to allosteric changes in intramembrane proteins of the receptor complex and ion pumps. In this way electrical information can be transformed into chemical information. Thus, the two languages of the brain, i.e. the electrical and chemical signals, can be interconverted and the information coded into these two languages effectively integrated.
The above concept on WT and VT makes it easier to understand how mood can be affected by hormones such as steroid and thyroid hormones. We have, for example, observed that glucocorticoid receptor (Fuxe et al. 1985d) immunoreac­ tivity exist in very large numbers of nerve cells all over the cortical hemispheres (archi-, paleo- and neocortex), with predominant nuclear location. Thus, WT in the cortical areas of the brain controlling mood can be massively influenced by these hormones, which represent important signals in VT. Furthermore, these hormones also influence VT on the cortical networks subserving mood, since they can, by direct nuclear actions, regulate the synthesis and release of paracrine sig­ nals such as peptides from the cortical nerve cells to reach distant receptor pop­ ulations in the cerebral cortex (see Fuxe etal. 1985 a; Agnati et al. 1986b). Finally, in the frame of VT we can surmise that psychoactive drugs, even if they work on the wiring transmission (e.g. at synaptic level) reach their targets according to a VT mode. This gives further evidence that that endogenous signals may also effect WT by diffusing in the extracellular fluid of the brain. Mood control can be con­ sidered as the concerted result of a large number of endogenous and exogenous signals affecting the networks subserving mood via actions on WT and VT.
2 Humoral Modulation of Wiring Transmission
Firstly, it must be considered that there exists a blood-brain barrier, so that the central nervous system will not receive a number of peripheral signals. However, there also exist chemical and physical "windows" through which the brain re­ ceives and delivers lipophobic messages. Chemical windows consist of fa­ cilitated transport, active transports etc. The physical windows are represented by brain areas devoid of the blood-brain barrier, such as the area postrema, the me­ dian eminence and the subfomical organ. Also the neuronal inputs represent a part of the physical window (see Fig. 4).
3 Actions of Gonadal Steroids on Wiring Transmission
3.1 General Aspects
By means of auto radiographic and steroid receptor binding techniques nerve cells concentrating steroid sex hormone have been demonstrated in the central nervous system and been found to be concentrated in the limbic forebrain, the medial pre­ optic area and the hypothalamus, especially the medial part (see Cottingham and
8 K. Fuxe et al.
CSF
CNS (VT and WT)
PNS (VT and WT)
Fig. 4. Schematic illustration of interac­ tions between central nervous system (CNS), peripheral nervous system (PNS) and endocrine organs via volume trans­ mission (VI) and wiring transmission (WI). The role of circumventricular or­ gans (CVOs) as "physical" windows for blood signals not passing the blood-brain barrier is indicated (e.g., peptide hor­ mones). CSF, cerebrospinal fluid
Pfaff 1986; Stumpf 1968; Stumpf and Sar 1975b,c, 1978, 1981; Sar and Stumpf 1973, 1977). Recently, Cintra et al. (1986) have demonstrated, by means of a rat monoclonal antibody directed against the human estrogen receptor, purified from MCF-7 human breast cancer cells, that estrogen receptor immunoreactive (IR) nerve cells exist in the limbic forebrain areas, in the hypothalamus and the preoptic area with the same distribution as the estrogen-accumulating nerve cells (Figs. 5 and 6). Of substantial interest was the observation that the estrogen re­ ceptor immunoreactivity was exclusively present within the nuclei and that no translocation of the IR material took place in the cytoplasm following castration. These results strongly support the importance of the genomic actions of es­ trogens, inducing changes in the decoding of various types of proteins.
By means of combined autoradiography and immunohistochemistry it has been possible to demonstrate the accumulation of steroid hormones in transmit­ ter-identified neurons such as dopamine (DA), noradrenaline (NA), vasopressin, p-endorphin, y-aminobutyric acid (GABA) and somatotastin nerve cells (see Heritage et al. 1977; Stumpf and Sar 1981; Wuttke et al. 1981). As pointed out
Fig.6. Camera lucida drawing of distribution of estrogen receptor immunoreactivity pre­ sent in nuclei of nerve cells of the preoptic area and adjacent regions in a coronal section of the normal male rat brain. BSTL, bed nucleus striae terminalis, lateral part; BSTM, bed nucleus striae terminalis, medial part;/, fornix; GP, globus pallidus; HDB, nucleus of the horizontal limb of the diagonal band; ICj, island of Calleja; LPO, lateral preoptic area; MPN, median preoptic nucleus; MPO, medial preoptic area; ox, optic chiasm; PVHap, paraventricular nucleus, anterior parvocellular part; PvPO, periventricular preoptic nu­ cleus; sm, striae medullaris of the thalamus; SO, supraoptic hypothalamic nucleus; Tu, ol­ factory tubercle; VP, ventral pallidum; III V, third ventricle
Hormones and Synaptic Transmission 9
• ' . . '
'. "
" : .. ~.
.' , ."
.1 '
I.
·LJt '-, " I I .. I \ ' -ny, I \ 1\ " • , • I \ ~ I :.. \ I \ I ;' I \ \ eSTM /"" \ GP \ I \ " • \
" ., "It ," :~. \
. I' •• ,....." ,; ~·,·H·..:P I. .. ..... "", .:.: ",'" ;r't/,. ,,01- ~ " • .;1
I·'·.···· 1"" .', .. ~., .. ~ ;"" .rr' :. I":: • :.~,:. . :,..""."" . ,' VP . ~ .Ut. .. (sSTI~· ... ,'1.-.: ', \ f'J. :'1'" 'c' ...... ~. ...... r . . . .. \
l~ •• ':S.4.., .'. .•. . \ 1'!~~;\" " " .; '-____ _ i'~' .:::'iL';f'~ .:. : ./. '. '
t ,,4: l,: •• '.1" ,,: ., 0; • .... 1·.. "'MPH ht . { .... 'i !:".,' " ' ~~.t-:. t~ It, " l u-~,I:.. .,/ ""ff- :. .... t.i'~· I .' ,,' "'\ ... V ;. •.•. : .. I .
-l.~ . .cp'dj,hpO " ot
.I'!'---- .............. f Hoe ..
10 K. Fuxe et al.
by Cottingham and Pfaff (1986) sex steroid hormone-binding neurons exhibit a high degree of interconnectedness. A network of steroid hormone-binding neurons is created which allows the amplification of steroid hormone actions and stability in the performance of the hormone-dependent network, as well as appro­ priate channeling of the inputs to this network (see Cottingham and Pfaff 1986). Thus, these hormone-dependent networks can be influenced by sex steroid hor­ mones in the medial hypothalamus, in the preoptic area and in the limbic system, additionally enabling coordination of activity in the networks controlling repro­ ductive behaviours and the secretion of luteinizing hormone-releasing hormone (LHRH) and of prolactin.
It should also be mentioned that regional levels of p-endorphin immunoreac­ tivity and enkephalin immunoreactivity are altered in the neuroendocrine areas of the hypothalamus and of the preoptic area by estrogen treatment (see Dupont et al. 1981). It will be of substantial interest to evaluate the possible presence of estrogen immunoreactivity within these neurons of the mediobasal hypothalamus and of the preoptic area.
In agreement with the importance of genomic actions of estrogen it has been found that estrogens can regulate tyrosine hydroxylase gene transcription in the arcuate nucleus of the rat hypothalamus (Blum et al. 1985). Estrogens can also regulate proopiomelanocortin (POMC) gene expression in the rat hypothalamus. A decrease has been seen in POMC mRNA levels after estrogen treatment which is due at least in part to a decrease in the synthesis ofPOMC mRNA (see Roberts et al. 1985). As a matter of fact it seems likely that the majority of sex steroid ac­ tions on the presynaptic properties of transmitter-identified neurons, such as ef­ fects on synthesis mechanisms and release mechanisms for neurotransmitters, e.g. monoamines, and GABA are secondary to primary actions on the decoding of the genome of the estrogen IR neurons of the brain (for review, see McEwen et al. 1981).
3.2 Studies on Presynaptic Features of Monoamine Neurons
Estradiol-17 p, progesterone and androgens have all been found to induce discrete changes in DA, NA and adrenaline (A) levels and utilization in the hypothalamus and in the preoptic area in male and female rats in various types of endocrine states (see Fuxe et al. 1981 a; Andersson et al. 1981; Wuttke et al. 1981; Lofstrom and Beckstrom 1981; Fuxe et al. 1977a). Estradiol-17P appears to produce its central inhibitory feedback action on LHRH secretion at least in part by a direct action on the hypothalamus, leading to an activation of the lateral tuberoinfun­ dibular DA pathway, which in tum, by axoaxonic influence and/or by effects on tanycytes, may inhibit the secretion of LHRH from the median eminence (see Quimet et al. 1984). This action ofestradiol-17P may be mediated partly by an increase in the secretion of prolactin, which has the ability of increasing DA turn­ over in the lateral palisade -zone of the median eminence, where the LHRH and DA terminals interact (see Fuxe et al. 1984 b). However, it must also be consid­ ered that estrogen receptors are probably located in many of the DA nerve cell bodies of the mediobasal hypothalamus (see Heritage et al. 1977) and that es-
Hormones and Synaptic Transmission 11
trogen treatment of hypophysectomized and castrated animals leads to a marked increase ofDA utilization in both the lateral and the medial palisade zones of the median eminence (Fuxe et al. 1981 a).
The central facilitatory feedback action of estradiol-17 p on LHRH secretion instead appears to involve an increase of NA utilization in the medial preoptic area (see Fuxe et al. 1977a; Lofstrom and Beckstrom 1981; Wuttke et al. 1981). It has been postulated that this action of estrogen involves an inhibition of GABA interneurons in the medial preoptic area (see Wuttke et al. 1981). Thus, muscimol, when given intraventricularly, can reduce NA turnover in the anterior hypothal­ amus (see Fuxe et al. 1979a). Intraventricular injection of GABA also reduces NA turnover in the medial preoptic area (see Wuttke et al. 1981). Thus, the cen­ tral facilitatory feedback action of estrogen appears to involve increased NA re­ lease in the medial preoptic area mediated at least partly via an action on es­ trogen-sensitive GABAergic neurons in this region, leading to a loss of presyn­ aptic inhibition ofNA release (see Wuttke et al. 1981).
Evidence has also been obtained that androgenic steroids can produce discrete change of DA and NA utilization in the hypothalamus and preoptic area by ac­ tivation of androgenic steroid receptors (see Andersson et al. 1981). Thus, evi­ dence was obtained that the activity in noradrenergic mechanisms of the preoptic area can be turned off by the androgenic steroid R1881, while the inhibitory do­ paminergic mechanism in the median eminence is turned on by this agent. As re­ ported above, similar results have been obtained following treatment of castrated female rats with estrogens such that estrogen produces its central inhibitory feed­ back action on LHRH secretion.
The above results taken together indicate that sex steroids, by means of changes in genetic transcription via their nuclear actions, alter the formation of regulatory proteins controlling the chemical transmission in the steroid target cells. By axoaxonic contacts the steroid target cells in the local circuits of various regions of hypothalamus and preoptic area will influence the various DA and NA nerve terminal systems in a discrete way, as observed in the above-mentioned ex­ periments (see Fuxe et al. 1979b).
However, estrogens not only influence catecholamine turnover in regions where estrogen IR nerve cells exist but also influence DA utilization in parts of the forebrain where few estrogen IR nerve cells are found. In the hypophysecto­ mized and castrated female rat it has, for example, been found that estradiol-17 p can markedly reduce DA utilization in various parts of the striatum and of the nucleus accumbens. These results may be induced by activation of estrogen recep­ tors of neural groups within the preoptic area and the hypothalamus which pro­ ject to the ascending DA neurons and thus indirectly regulate DA utilization in the meso striatal and meso limbic systems. These results clearly indicate that es­ trogens can also modulate motor functions and mental activities such as mood (see Fuxe et al. 1981 b). When discussing the actions of estrogens on the striatal mechanisms, behavioural and neurochemical studies have also indicated antido­ paminergic actions, which may be mediated at least in part via the pituitary gland through increases in the secretion of prolactin (see Euvrard and Boissier 1981). Estrogens are known to alleviate extrapyramidal symptoms such as hyperkinesias in patients receiving neuroleptic drugs (see Bedard et al. 1981).
12 K. Fuxe et al.
Of special interest are our recent observations that steroid hormones can in­ fluence the coexistence of peptides and monoamines (Hokfelt et al. 1980) in the monoamine neurons (Fuxe et al. 1985 a), probably mainly via an influence on the synthesis of the peptide comodulator (see Sect. 4.2. paragraph on glucocorticoid receptors).
3.3 Studies on Monoamine Receptor Mechanisms
In a number of papers, estrogen, progesterone and other sex steroids have been found to modulate the binding characteristics of central oc- and fJ-adrenergic re­ ceptors (Fuxe et al. 1979b,c; Wilkinson 1978; Wilkinson et al. 1979a,b, 1981). Our results have demonstrated that the oc- and fJ-adrenergic receptors are sensitive to combined treatment with estrogen and progesterone resulting in the induction of sexual behaviour. The changes induced in the binding characteristics are ligand and region specifIc. In contrast, the gonadal steroids have been found to exert little effect on opiate receptor binding parameters (see Wilkinson et al. 1981; for review see McEwen et aI1970). Estrogen treatment has also been found to influ­ ence the 5-HTt receptors and D2 receptors in the striatum. The effects of estrogen on striatal D2 receptors are complex. It was found at an early stage that estrogen can produce an increase in the number ofD2 receptors in striatal membranes (see Bedard et al. 1981; Fuxe et al. 1979c). These biochemical signs of DA receptor hypersensitivity may be related to the conversion of estrogens to catecholes­ trogens and/or to reduced DA release (Fuxe et al. 1981 a; Gordon 1985). How­ ever, the acute actions of estradiol on DA receptors produce a direct desensitiza­ tion or an uncoupling of the receptor-effector mechanisms, characterized by a de­ crease in the proportion ofD2 receptors in the high-affinity agonist state (Gordon 1985).
It seems likely that most of the effects of estrogen and progesterone treatment on monoamine receptors in steroid receptor-rich areas of the brain are produced via actions on the estrogen and progesterone receptors present in these areas. The activation of the steroid receptors may in turn produce changes in genetic tran­ scription which can affect various aspects of biochemical signals regulating monoamine and other types of receptors (see Agnati et al. 1981; McEwen et al. 1981). This hyothesis is supported by the ability of in vivo estrogen administra­ tion to produce sustained and delayed increases in the receptor density values of transmitter receptors in regions where estrogen receptor immunoreactivity exists. Nongenomic actions of sex steroids are probably also involved, however, since changes in the binding characteristics of monoamine receptors can also be dem­ onstrated in membrane preparations upon in vitro addition of the steroid. This may be the case in the above-mentioned estrogenic modulation of striatal DA re­ ceptor sensitivity. Membrane actions have also been demonstrated in a number of electrophysiological and biochemical experiments (see Baulieu 1981; Moss and Dudley 1985). Thus, within seconds estradiol-17oc-butyric acid inhibits the firing rate of nerve cells applied directly to the membrane of these cells (Carette et al. 1979). These estrogen derivatives cannot cross the cell membrane. Furthermore, progesterone on the surface of maturing Xenopus laevis oocytes promotes oocyte
C H
E M
IC A
N -4
C O
E X
IS T
E N
C E
GENETIC TRANSCRIPTION
GENETIC TRANSCRIPTION
Fig.8. Schematic illustration of the possible interrelationships between catecholamine (Rc.J and estrogen (RE) receptors in target nerve cells. LAC, low-affinity component; HAC, high-affinity component
meiosis without entering the cell (see Baulieu 1981). Also, progesterone has re­ cently been found in vitro to activate LHRH and DA nerve terminals by a non­ genomic mechanism (Ramirez 1985). These effects, however, appear to be medi­ ated via the metabolite pregnenolone, which has shown to exhibit highly potent activation of DA terminals in the hypothalamus and in the striatum. It was sug­ gested that membrane actions of steroids may lead to changes in the transduction mechanisms, resulting in changes in adenylate cyclase and phospholipase C activ­ ity. In this way a chain of events may be started, leading finally to the release of LHRH and DA demonstrated in vitro on incubation with progesterone or preg­ nenolone. Thus, it seems likely that sex steroids can regulate membrane excitabil­ ity, possibly via a special membrane site, since the action of estrogen on the mem­ brane appears to be antagonized by antiestrogens (see Moss and Dudley 1985). An effect of membrane fluidity should, however, also be considered. These mem­ brane changes could also affect the interactions among intramembrane macro­ molecular complexes, such as ion pumps, ion channels, etc. In particular, it seems likely that the membrane actions of gonodal steroids could also modulate the re­ ceptor-receptor interactions (see the paragraph below on glucocorticoid recep­ tors). This steroid action could thus have a role on information handling by neural networks. In fact, as discussed above, intramembrane receptor-receptor interactions represent an important integrative mechanism in synaptic trans­ mission allowing for a divergence and convergence of information flow in the synapses (see Fig. 7). They also increase the number of information signals which can be produced.
It should also be mentioned that gonadal steroids such as estrogen influence the transduction mechanisms; this is illustrated by a reduction ofhistamine-stim­ ulated adenylate cyclase following 7 days of treatment with estrogen and reduc­ tion in isoproterenol-stimulated adenylate cyclase upon chronic estrogen treat­ ment (see McEwen et al. 1981).
It must be emphasized, however, that not only can VT influence WT, but WT can also influence VT, as illustrated in Fig. 8. Thus, indications exist that cate­ cholamine and serotonin (5-HT) receptors also can regulate steroid receptor syn-
Hormones and Synaptic Transmission 15
thesis (see Ginsburg et ai. 1977; Kitayama et aI., unpublished observations). Thus, the networks of the brain regulate their own sensitivity to hormonal signals by controlling the amount of steroid receptors present in the nerve cells. Of special interest in this regard is the fact that unsaturated fatty acids can affect sex steroid hormone receptors in the brain (see Kato 1985). Thus, transmitter receptors such as catecholamine receptors may, via regulation of phospholipase C, control the formation of unsaturated fatty acids, such as arachidonic acid, which when reach­ ing cytoplasmatic steroidal receptors inhibit the binding capability of these recep­ tors (see Kato 1985).
4 Actions of Glucocorticoids on Wiring Transmission
4.1 General Aspects
The existence of specific receptors for glucocorticoid steroids in the central ner­ vous system was first provided in biochemical studies by McEwen et ai. (1969, 1970). The nuclear concentration of 3H-corticosterone, as seen using autoradiog­ raphy , indicated that the receptors for glucocorticoid hormones are mainly nu­ clear in location. They were found principally within hippocampal formation, the septal area and the amygdaloid cortex. Corticosterone target neurons were also observed in the thalamus, but not within the hypothalamus, and in the preoptic area (see Stumpf and Sar, 1975a, 1981; see McEwen 1982; see Ganten and Pfaff 1982). However, by means of monoclonal antibodies against the rat liver glu­ cocorticoid receptor (GR) in combination with the indirect immunoperoxidase technique, we have been able to demonstrate GR IR nerve and glial cells all over the brain and spinal cord of the male rat (Fig. 9) (Fuxe et ai. 1985 b-d, 1987; Ag­ nati et al. 1985). The GR immunoreactivity in the nerve cells was found mainly in the nucleus, but also in the cytoplasm (Figs. 10 and 11). Following 2-4 day ad­ renalectomy the nuclear GR immunoreactivity of the nerve cells was found to dis­ appear but the weak cytoplasmatic GR immunoreactivity remained. Following 4 h of treatment with corticosterone the GR immunoreactivity reappeared in the cell nuclei (Fig. 11). GR IR glial cells were mainly found in the white matter (Fig.9), where they formed bands of cubic-like structures between the fiber bundles, probably mainly representing glial cells of the oligodendroglia type.
"
• • • • • • • •
• •

• • • , • • • • • •
• • .. • Fig. 10. Cellular localization of glucocorticoid receptor (OR) immunoreactivity (OR IR) in nerve cells of the frontoparietal cortex of the male rat. In the pyramidal nerve cells the OR immunoreactivity is seen not only in the nucleus but also in the pericaryon and in the apical dendrite. Bar, 200 Ilm
Fig.n. Localization of glucocorticoid receptor (GR) immunoreactivity in pyramidal nerve cell nuclei and sur­ rounding cytoplasm in the hippocam­ pal subregion CAl and its modulation by adrenalectomy with or without sub­ sequent corticosterone treatment (4 h, 10 mg/kg, i.p.). Transverse sections. After adrenalectomy (ADX), the OR immunoreactivity is observed exclu­ sively in the pericarya of the pyramidal nerve cells. After corticosterone (cort) treatment OR immunoreactivity even stronger than that present in the con­ trol animals is found in the nerve cell nuclei. Konig-Klippel level A4100 Ilm
GR control CAl
GR ADX CAl
GR ADX.cort. CA1
18 K. Fuxe et al.
Fig. 12. Distribution of glucocorticoid receptor (OR) immunoreactivity in the anterior and dorsal periventricular part of the of the hypothalamus of the male rat. Densely packed strongly OR immunoreactive (IR) nerve cell nuclei are found in the anterior parvocellular nucleus of the paraventricular hypothalamic nucleus (ap) and in the periventricular hypo­ thalamic nucleus (pv). Weakly to moderately OR IR nerve cell nuclei are found in the an­ terior hypothalamic nucleus (AHy), the medial preoptic nucleus (MPO) and the bed nu­ cleus of the stria terminalis, preoptic part (BSTPO). 3, third ventricle
In agreement with our observations showing a widespread distribution of OR IR neurons in the central nervous system, an in vitro quantitative auto radio­ graphic analysis of adrenal steroid binding sites showed a widespread distribution of glucocorticoid binding sites in the rat central nervous system. The highest con­ centrations were found in the dentate gyrus, the lateral septum, the nucleus trac­ tus solitarius, the nucleus paraventricularis and the amygdaloid complex (see de Kloet 1985; Rostene et al. 1985).
Fig. 14. Distribution of glucocorticoid receptor (OR) immunoreactivity in the frontoparie­ tal cortex. The highest densities of OR immunoreactive nuclei are observed in layers 2, 3 and 6 of the cortex and in the nucleus caudatus putamen. DAB-nickel combination was used. Level: 1.3 mm behind Bregma. FrPaM, frontoparietal cortex, motor area; FrPaSS, frontoparietal cortex, somatosensory area; CC, corpus callosum; ec, external capsule; cg, cingulum; LV, lateral ventricle; fl, fimbria of hippocampus; CPu, caudate putamen; AD, anterodorsal thalamic nucleus; DG, dentate gyrus; PCg, posterior cingulate cortex; CA3, field CA3 of hippocampus. Bar, 250 ~m
~- . - ~
'- -- "'
-: :-
:- .:;
~~ ~
_ ... - ~ .
, . .
> .... ~
..
.. •
, ....
... .. . ... ;-
. .' . . . Fig. 16. Distribution of glucocorticoid receptor immmunoreactive (OR IR) nerve cell nuclei in the anterior cingulate cortex. High densities of strongly OR IR nerve cells are found especially in layers 2 and 3. Bar, 250 Jlm
ox
Fig. 17. Distribution of glucocorticoid (OR) immunoreactivity in the hypothalamus, dorsal striatum and piriform cortex. Note the low number of weak OR immunoreactive (IR) pro­ files in the globus pallidus, ventral pallidum and substantia innominata. High densities of strongly OR IR profiles are localized in the claustrum, in the endopiriform nucleus and in the olfactory and piriform cortex. Approximate level: 1.3 mm below bregma. DAB-nickel combination was used. ie, internal capsule; ox, optic chiasm; LH, lateral hypothalamic area; SO, supraoptic hypothalamic nucleus; GP, global pallidus; CPu, caudate putamen; FStr, fundus striati; VP, ventral pallidum; Sf, substantia innominata; AA, anterior amygdaloid area; CxA, cortex-amygdala transition zone; ct, claustrum; En, endopiriform nucleus; LOT, nucleus of the lateral olfactory tract; PO, primary olfactory cortex; Ce cen­ tral amygdaloid nucleus; Bar, 250 Jlm
22 K. Fuxe et al.
BST
ic
Fig. IS. Distribution of glucocorticoid receptor immunoreactivity in nerve cells of various anterior thalamic nuclei. DAB-nickel combination was used. Approximate level: 1.3 mm behind bregma. 3, third ventricle; sm, striae medullaris thalamus; PVA, paraventricular thalamic nucleus, anterior part; PT, paratenial thalamic nucleus; AD, anterodorsal tha­ lamic nucleus; A V, anteroventral thalamic nucleus; BST, bed nucleus striae terminalis; AM, anteromedial thalamic nucleus; Rt, reticular thalamic nucleus; Re, reuniens thalamic nucleus; ie, internal capsule; f, fornix; Pe, periventricular hypothalamic nucleus. Bar, 250 !lm
Hormones and Synaptic Transmission 23
Table 3. Sites of coexistence of GR JR, monoaminergic and peptidergic neurons
Neurons
Noradrenaline A1-A7
Growth hormone - releasing factor IR Arcuate nucleus
Somatostatin IR Periventricular hypothalamic nucleus
Neuropeptide Y IR C1-C3 area A1 area Arcuate nucleus
(I.-Melanocyte - stimulating hormone IR Arcuate nucleus Perifornical nucleus Central amygdaloid nucleus
Cholecystokinin IR Paraventricular hypothalamic nucleus Hippocampus Periventricular hypothalamic nucleus
p-Endorphin IR Arcuate nucleus
50 o
75
100
Nomenclature of monoaminergic neurons according to Dahlstrom and Fuxe (1964) GR, glucocorticoid receptor; JR, immunoreactive
Fig.19. High densities of glucocorticoid receptor immunoreactive nerve cell nuclei are found in the nucleus caudatus putamen DAB-nickel combination was used. Star, fibrae capsulae internae; ec, external capsule. Bar, 200 11m
24 K. Fuxe et al.
OR Immunor .. cllye
igh
(den,'I1 .. )
1mm
Fig. 20. Camera lucida drawing ofthe distribution of glucocorticoid receptor (OR) and cor­ ticotropin-releasing factor (CRF) immunoreactivities in the septal and preoptic area of the male rat, based on two-colour immunocytochemistry (see Harfstrand et al. 1986). Double­ labelled nerve cells are represented by triangles. Stars indicate CRF immunoreactive cells without any demonstratable OR immunoreactivity. ac, anterior commisure; ox, optic chiasm; LV, lateral ventricle; LSD, laterodorsal part of the septal area; SFi, nucleus fim­ brialis septalis; SHy, nucleus septohypothalamicus; MnPo, nucleus preopticus medianus; BSTM, BSTL, medial and lateral components of the bed nucleus of striae terminalis; LPO, lateral preoptic nucleus; MPO, medial preoptic nucleus
By means of two-color immunocytochemistry we have demonstrated that the vast majority of the NA, A and 5-HT neurons of the lower brain stem contain strong OR immunoreactivity (see Fuxe et al. 1985 b; Harfstrand et al. 1986). These results indicate that glucocorticoids can in part control brain function via modulation of the synthesis and release ofNA, A and 5-HT and their respective peptide comodulators. In view of the known function of NA, A and 5-HT path­ ways it is now possible to better understand how glucocorticoids can regulate cardiovascular and neuroendocrine function, the sleep-wakefulness cycle, attention and mood in man.
The various DA cell groups of the brain showed variability with regard to the number of DA nerve cell bodies positive for glucocorticoid receptors. Approxi­ mately 50% of the DA cells of groups A8, A9 and AlOin the ventral midbrain contain a weak to moderate degree of OR immunoreactivity. As seen in Table 3, the arcuate DA cell bodies all contain OR immunoreactivity, while the All DA
Hormones and Synaptic Transmission 25
Fig. 21. Demonstration of nuclear glucocorticoid receptor immunoreactivity by means of FITC fluorophor in all neuropeptide Y (NPY) immunoreactive nerve cells of the medial parvocellular part ofthe arcuate nucleus (mARC) . NPY immunoreactivity is located in the cytoplasm and demonstrated by rhodamine fluorophor. Two-colour immunofluorescence methodology. Arrows point to some of the double-labelled cells. Arrows with bars indicate glucocorticoid receptor immunoreactive cells lacking NPY immunoreactivity. ME, median eminence; 3, third ventricle. Bar, 100 !lm
cell group of the posterior hypothalamus showed no GR immunoreactivity in any of its cell bodies (see Hiirfstrand et al. 1986).
As seen in Table 3, all the CRF IR cells of the paraventricular hypothalamic nucleus projecting into the median eminence, as well as all the arcuate growth hormone-releasing factor (GRF) IR neurons projecting into the median emi­ nence, contain GR immunoreactivity (Cintra et al. 1987). Many of the CRF IR cells of the preoptic nuclei and of the bed nucleus of the nucleus striae terminalis are also GR immunoreactive (Fig. 20). Thus, glucocorticoids can directly control the CRF and GRF neurons by an action at the nuclear level, probably represent­ ing the mechanisms underlying the central feedback action of glucocorticoids on GRF and CRF synthesis leading to an inhibition ofGRF and CRF secretion (the delayed feedback). Also the neuropeptide Y (NPY) neurons of the parvocellular part of the arcuate nucleus all contain GR immunoreactivity (Fig. 21). This GR
26 K. Fuxe et al.
IR is as strong as the one demonstrated in the CRF IR neurons. About 50% of the a-melanocyte-stimulating hormone (a-MSH) IR neurons of the arcuate nu­ cleus and of the perifornical area exhibited GR immunoreactivity. Furthermore, all the p-endorphin IR neurons of the arcuate nucleus were GR immunoreactive (Table 3). Moreover it was discovered that a-MSH immunoreactivity exists in cortical pyramidal cells, which are also GR immunoreactive.
4.2 Morphometric and Microdensitometric Analysis of GR Immunoreactivity in the Central Nervous System
These analyses have been carried out using a computerized system for image anal­ ysis (IBAS II, Zeiss Kontron Munich, FRG). For details on the methodology, see Agnati et al. (1984 b, 1985). A semiquantitative method to assess the relative amount of antigen has been developed: discrimination is performed at a level ca­ pable of excluding the background and the field area, (F AC)o is measured. The same procedure is repeated at different levels of discrimination allowing the selec­ tion of higher and higher grey tone values. After each discrimination procedure the corresponding field area, (F AC)i, is measured. Thus, the percent rations Yi = (F AC)if(F AC)o can be calculated. By considering the plot of the Yi values as a function of the respective levels of discrimination, a curve can be obtained (Fig. 22) which expresses the relative content of antigen per area in the sampled field. The ED25, ED50 and ED75 of the curve represent indices oflow, medium and high content of antigen, respectively (see Zoli et al. 1986).
It is also possible to convert the exponential decay of the curve into a straight line. This linear transformation is useful, since the slope of the straight line can be used as a relative quantitative index of antigen content, which is more sensitive than the ED values (see below).
One result from the morphometric analysis is demonstrated in Fig. 23, which shows a density map of the GR IR neurons within the area of the paraventricular hypothalamic nucleus. The number of GR IR nerve cells is given per unit of square, in relation to the various subnuclei of this region. It is seen that the largest amounts of GR IR neurons are observed within the parvocellular part (FP). High densities are also found within the periventricular hypothalamic nucleus. Scat­ tered GR IR cells are found within the magnocellular part, and none are found at this rostrocaudallevel within the dorsal parvocellular subnuclei. A density map illustrating the coexistence and distribution of the tyrosine hydroxylase (TH) and GR immunoreactivity in the arcuate nucleus is shown in Fig. 24.
A microdensitometric analysis has been performed in a number of regions of the telencephalon and diencephalon, namely the caudate putamen, nucleus amygdaloideus medialis, somatosensory frontoparietal cortex, periventricular hypothalamic nucleus and the parvocellular part of the paraventricular hypotha­ lamic nucleus. As seen in Fig. 22, the ED50 index shows similar amounts of GR immunoreactivity in the striatum, the amygdaloid cortex and the neocortex. All the indices are higher for the parvocellular part of the paraventricular hypotha­ lamic nucleus, giving evidence for higher amounts of immunoreactivity in this re­ gion than in the other regions analysed. It should be noticed, however, that the
Hormones and Synaptic Transmission 27
Striatum (CPu) Amygdala (Me)
ED25 = 32.5 ED25 = 32.5
\ ED75 = 26.5
\ ED75 = 26.8
75
50 . . <:> \ \ :;:! 25 . x ., '. ' . "0 .......... ::::::. '-. u 0 I ,r I ,. -< 35 30 25 20 15 35 30 25 20 15 ~ , Fronto-parietal cortex (FtPa) N. paraventricularis (Pa)
~ 100 N. periventricularis (Pe)
• ED25 = 32.0 • Pa Pe
75 ED50 = 27.1 32.0 EED75 = 21.8 30.5
50 \ 25 \
35 30 25 20 15 35 30 25 20 15
Striatum (CPu) Amygdal<l (Me)
2
.~ <:> . :;:!
~ . .
'" "0 0 I I I I· I I I j.
.( 35 30 25 20 15 35 30 25 20 15 ~ Fronto-parietal cortex (FtPa) N. paraventricularis (Pa) ,
N. periventricularis (Pe) u -<
.", .. ~ . ~ ' . "'-
0 I I i I I i i I II
35 30 25 20 15 35 30 25 20 15
levels of discrimination
Fig. 22. Schematic representation of the procedure introduced in immunocytological stud­ ies to obtain a relative quantitative estimation of the antigen content. In the figure· the ef­ fects of subsequent discriminations are illustrated. Mter each discrimination the field area of positive immunoreactivity, (FAC)i, is measured. The percent ratios Yi=«FAC)i/ (FAC)o) x 100 are calculated and by means of these values the estimation of the antigen content is performed. The Yi values are plotted as a function of the different levels of dis­ crimination. A characterization of the curve is obtained by means of the ED values (upper panels). A more precise characterization of the curve is obtained after linear transformation by considering the slope values (lower panels), which are given with the respective 95% con­ fidence interval
Fi g.
2 3.
D en
si ty
d is
tr ib
ut io
n pl
ot o
Hormones and Synaptic Transmission 29
Coexistence of TH and GR immunoreactivity In nerve cells in medlobasal hypothalamus
microns 1000~--r---'---~---r--~--~r---r---~--'---~
v+
Fig.24. Density map illustrating the coexistence of tyrosine hydroxylase (TH) and glu­ cocorticoid receptor (GR) immunoreactive (IR) nerve cell bodies of the arcuate nucleus. The number of positive cells is shown in each square. In the black triangle is indicated the number of OR IR nerve cell bodies and in the white triangle the number of TH IR nerve cell bodies. The circles around the numbers in the white triangles indicate that the TH IR nerve cells have a nuclear OR immunoreactivity (coexistence). The orientation is indicated in the upper right part of the picture. The outline of the ventral surface and of the third ventricle is also given in the density map. ME, median eminence; M, medial; V, ventral; D, dorsal; L, lateral
curve within the periventricular hypothalamic nucleus is highly skeweq, and therefore a certain amount of high grey tones may have escaped in this type of evaluation. By considering the slope values of the respective straight lines (see Fig.22) an improved estimation of GR immunoreactivity may be obtained. Again, the parvocellular part of the paraventricular hypothalamic nucleus has the highest index as inferred from the low slope value. Using this type of semiquan­ titative method it is possible to study changes in the GR immunoreactivity in highly discrete neuronal populations of the central nervous system.
30 K. Fuxe et al.
4.3 Studies on Presynaptic Features of Monoamine Neurons
The effects of corticosterone on regional DA and NA levels and utilization in the hypothalamus depend upon the endocrine state of the animal. However, a consis­ tent finding has been the ability of corticosterone to increase NA utilization in the median eminence, even in hypophysectomized and adrenalectomized rats (see Andersson et al. 1981). In addition, corticosterone was able to increase NA levels selectively in the subependymallayer of the median eminence, indicating an ac­ tivation also of NA synthesis. These results may reflect the existence of an inhibi­ tory noradrenergic mechanism in the median eminence, controlling CRF release via an axoaxonic modulation. It is suggested that the primary action is on a GR IR local circuit neuron in the hypothalamus which can interact via an axoaxonic influence with the noradrenergic network in the median eminence (Andersson et al. 1981). Indications were also obtained that glucocorticoids can reduce NA uti­ lization in the magnocellular part of the paraventricular hypothalamic nucleus. NA utilization in this region was also turned off by sustained hypersecretion of adrenocorticotropic hormone (ACTH). These results may indicate the existence of a facilitatory noradrenergic mechanism involved in vasopressin regulation. It is again suggested that the GR IR local circuit neurons are involved in producing this highly selective influence on NA turnover in this part of the paraventricular hypothalamic nucleus involving axoaxonic interactions in that nucleus. We have
e " .. >
• ~ .. e t ~ '0 ~
'" i Adll • • SoIY.nl
Adll . ·Cortlcoelerone 10mg/ kg e.C 4h :: Adll. -RU 28g88 10mg/ kg LC 4h
Adll.·Aldoeterone 100ug/ kg e.c 4h
Figs.25-27. Effects of corticosterone, RU 26988 and aldosterone on regional noradrena­ line (NA) levels in discrete nuclei of the preoptic area (Fig.25) and of the hypothalamus (Fig. 26) and on cathecolamine (CA) levels of the median eminence (Fig. 27) of adrenalec­ tomized male rats (4 days after adrenalectomy). Means+SEM (n=6) . Significances refer to rats treated with solvent alone: *p<0.05; **p<O.Ol (Dunn test). PVO, area anterior to the anterior periventricular hypothalamic nucleus; PV II, posterior periventricular hypo­ thalamic region; POP, pose, POM, periventricular, suprachiasmatic and medial parts of the preoptic area. NIST, ventral part of the bed nucleus of striae terminalis. P AFM, P AFP, magno- and parvocellular parts of the paraventricular hypothalamic nucleus; DM, dor­ somedial hypothalamic nucleus; SEL, subependymallayer of the median eminence; MPZ, medial palisade zone of the median eminence; LPZ, lateral palisade zone of the median eminence; BZ, border zone, an area located at the border of the medial and lateral hypo­ thalamus between the fornix and the ventral surface of the brain; PV I, anterior periven­ tricular hypothalamic region
Fig. 26
• :J iii ~
~ 1 U • a. • ~ "0 ~
MPZ
.~ Ad" •• RU 28988 10mg/ kg e.c 4h
; Ad".· Aldoeterone 100ug/ kg e.C 4h
LPZ
administered the selective GR agonist RU26988 (Moguilevski and Raynaud 1980; Veldhuis et al. 1982) in a dose of 10 mg/kg s. c. to adrenalectomized male rats. Marked depletion ofNA stores is observed after 4 h in large numbers of preoptic and hypothalamic NA nerve terminal systems but not in the median eminence CA nerve terminals (Figs. 25-27). These changes are associated with a lowering ofNA utilization in several regions outside the median eminence using the TH inhibition method. It seems possible that these marked and widespread actions of the selec­ tive GR agonist on NA neurons are mediated via activation ofGR in the NA cell bodies in the medulla oblongata, giving rise to the hypothalamic and preoptic NA innervation. Thus, via genomic actions GR may control the activity of NA-syn­ thesizing enzymes and the production of receptor proteins controlling the excit­ ability and the firing rate of the NA neurons. Recent findings also indicate that glucocorticoids control coexistence in both monoamine and peptide neurons.
32 K. Fuxe et al.
Thus, upon adrenalectomy there is an increase in the number of NPY IR nerve terminals in the hypothalamus, partly caused by an increased amount ofNPY IR in the phenylethanolamine-N-methyltransferase (PNMT) IR terminals (Fuxe et al. 1985a). Also, a number of papers (see Swanson et al. 1986) have shown that upon adrenalectomy CRF IR neurons of the paraventricular hypothalamic nu­ cleus begin to exhibit vasopressin and angiotensin immunoreactivity. These ef­ fects are also reversed by treatment with glucocorticoids. These results indicate that glucocorticoids normally act to suppress the gene expression of peptide co­ modulators, at least in some monoamine and peptide neurons. When discussing the effects of glucocorticoids on the CRF neurons it is of special interest to note that the acetylcholine-induced secretion of CRF from CRF IR neurons in vitro is reduced upon in vitro treatment with glucocorticoids. These results again underline that glucocorticoids, by regulating the gene expression of various types of receptor-linked proteins belonging to the cholinergic receptor mechanism, may appropriately regulate the sensitivity of the CRF IR neurons (see Vale et al. 1985). In line with this interpretation it has also been noted by Schonbrunn (1985) that glucocorticoids decrease the number of somatostatin receptors in GH4 C1 cells.
Other presynaptic transmitter mechanisms which are influenced by glucocor­ ticoid receptors are 5-HT utilization in the hippocampal formation (de Kloet 1985). Furthermore, vasoactive intestinal peptide (VIP) immunoreactivity is increased in the hippocampus of adrenalectomized animals (see McEwen 1982). Adrenalectomy also increases GABA uptake into synaptosomes of the hippo­ campal formation. All these effects are probably mediated via actions by glucocorticoids on hippocampal and/or raphe glucocorticoid receptors.
4.4 Studies on Monoamine Receptor Mechanisms
Evidence has been obtained for a glucocorticoid receptor involvement in the reg­ ulation of fJ-adrenergic receptors in the hippocampal formation and in the neo­ cortex (Roberts et al. 1981; Ogren et al. 1981). It was found that the 6-hydroxy­ dopamine-induced increase in the density values of 3H -dihydroalprenolol binding sites in the hippocampal formation was significantly greater upon adrenalectomy (Roberts et el. 1981). Adrenalectomy alone did not influence the binding charac­ teristics of the fJ-adrenergic receptors. In view of the learning impairments noted upon lesion of the ascending NA bundle to the cerebral cortex in combination with adrenalectomy (see Ogren et al. 1981), the further increase of the biochemical supersensitivity in the fJ-adrenergic receptors observed may reflect a potentiation of a compensatory biochemical phenomenon in the hippocampal formation to re­ store NA receptor activity and hippocampal function. The results amplify the ex­ istence of interactions between central noradrenergic mechanisms and central glucocorticoid receptors in avoidance learning, an interaction which may take place predominantly within the hippocampal formation and the cerebral cortex. It must also be considered that the NA cells in the locus coeruleus all contain a strong GR immunoreactivity predominantly located in their nuclei (Harfstrand et al. 1986), which must contribute to the interactions between NA and glucocor­ ticoids in the intact rat.
Hormones and Synaptic Transmission 33
Glucocorticoid receptors have also been shown to affect transduction mech­ anisms within the hippocampal formation. Thus, glucocorticoids facilitated his­ tamine-stimulated cyclic AMP formation but suppressed NA-stimulated cyclic AMP formation as well as VIP-dependent cyclic AMP stimulation (see McEwen 1982; McEwen et al. 1985a).
It has also been shown that the receptor-receptor interactions (see Fuxe and Agnati 1985) are modulated by glucocorticoids. Thus, glucocorticoids increase the ability of VIP to increase the density values of 5-HT 1 receptors within the dor­ sal subiculum (see McEwen et al. 1985a). Not only the transduction mechanisms but also the substrates for the protein kinases are affected by the glucocorticoid receptors within the hippocampal formation. Thus, the phosphoprotein synap­ sin I is found in increased concentrations upon treatment with glucocorticoid hormones (see Nestler et al. 1981). Synapsin I is regulated via activation of cyclic AMP-dependent and calcium-dependent protein kinases.
In Fig. 28 it is illustrated that steroid hormones via nuclear actions may con­ trol the synthesis of receptors for peptides and catecholamines. As examples, the
Principle for steroidal regulation of CA transmission
GR
Homoregulation via GR on / CA receptor sensitivity
'" Heteroregulatlon by GR via NPY receptors on CA synthesis and CA receptor sensitivity
GR - Effects on the gain of the receptor amplifier mechanism
Fig. 28. Illustration of some principles for the steroid regulation of the pre- and postsyn­ aptic features of catecholamine (CA) transmission and of corticotropin-releasing factor (CRE) neurons. Noradrenaline and adrenaline neurons both innervate the CRF immu­ noreactive neurons in the paraventricular hypothalamic neurons, and glucocorticoid recep­ tor (G R) immunoreactivity has been demonstrated in both the nerve cell nuclei and the peri­ carya of the catecholamine as well as of the CRF neurons. Thus, the glucocorticoid recep­ tors may regulate both the pre- and postsynaptic features of the classical transmission and the cotransmission lines in the catecholamine synapses. The direct regulation of the catecholamine transmission is termed "homoregulation" and the indirect regulation of the catecholamine synthesis and catecholamine receptor sensitivity, for example via modula­ tion of neuropeptide Y (NPy) receptor mechanisms and NPY stores is called "heteroregu­ lation". The available evidence indicates a profound effect of glucocorticoids on the regu­ lation also of the synthesis of the cotransmitter, in this case NPY (see Fuxe et al. 1985 a). It is also indicated that the glucocorticoid receptor, via the nuclear actions, can also reg­ ulate the receptor transducer mechanism (amplifier mechanism)
34 K. Fuxe et al.
NPY receptors and the Q(radrenergic receptors are shown. Thus, glucocorticoids may regulate the synthesis of the proteins containing the transmitter recognition sites leading to changes in the receptor density values and lor may regulate the coupling proteins such as the GTP-binding proteins which control the gain of the receptor-amplifying mechanism and/or may regulate the protein kinases involved in the phosphorylation of proteins, e.g. those in the membranes controlling trans­ duction of signals. Another important aspect of Fig. 28 is the fact that glucocor­ ticoids can regulate CRF synthesis and release not only via a direct action on the CRF IR neurons, but also via an influence on the afferent input. In this case the effects on the adrenergic input is illustrated. Thus, the glucocorticoids can influ­ ence the CRF neurons via both direct and indirect actions. The final outcome is the result of the integration of all these signals influencing the CRF neurons. Again, it should be emphasized that the glucocorticoids can modulate coexistence in neurons. In this case the effect on coexistence of NPY and A is exemplified.
It must be considered that glucocorticoids not only influence information handling, but also exert effects on the metabolic state of the neurons. These ac­ tions are probably of importance for the survival of the neurons. It has thus been seen that glucocorticoids can reduce the lesion-induced axon sprouting in the gyrus dentatus (see Scheff et al. 1980). Furthermore, McEwen and colleagues (1985a) have demonstrated that treatment with glucocorticoids produces not only a downregulation of the glucocorticoid receptors in the hippocampal forma­ tion, but also a nerve cell loss. Thus, glucocorticoids may, for example, inhibit the synthesis of a trophic factor in the pyramidal nerve cells. In this way the inhibitory effect of the hippocampal formation on the CRF neurons is lost (see Angelucci 1985), and the corticosterone serum levels become markedly increased, especially after stress.
Our observations that there are glucocorticoid receptors in glial cells, espe­ cially in the oligodendroglia cells, explain the ability of glucocorticoids to induce glycerol phosphate dehydrogenase in oligodendroglia cells (see McEwen 1982). WT in brain may be influenced by these effects of glucocorticoids on oligodendro­ glia cell, since these effects may lead to alteration in the myelination and nerve conduction velocities (Scheff et al. 1980; Friedrich and Bohn 1980; Henkin 1970). In addition, morphological changes have been noticed, such as hypertrophy of as­ troglia under the influence of glucocorticoids (see Scheff et al. 1980).
5 Actions of Thyroid Hormones on Wiring Transmission
5.1 General Aspects
In 1958 Ford and Gross provided evidence for the existence of a hypothalamic site of action of thyroid hormones by demonstrating the accumulation of radio­ active T 4 and T 3 in the paraventricular region of the hypothalamus. Thyroid-con­ centrating neurons have also been demonstrated within other areas of the hypo­ thalamus (Stumpf and Sar 1978). The thyroid hormone receptors have been char­ acterized as nuclear receptors (Oppenheimer et al. 1974; Eberhart et al. 1976).
Hormones and Synaptic Transmission 35
Light-microscopic autoradiograms of brain after intravenous administration of 125I_T 3 gives evidence for the localization of 125I_T 3 in discrete neuronal systems (Dratman et al. 1982). Furthermore, the circulating form of the hormone T4 has been shown to be differentially taken up in various brain areas by a high-affinity transport mechanism. These results indicate that the central nervous system must be an important site of action for thyroid hormones. It has also been speculated that the amino acid hormones T 4 and T 3 may be substrates in metabolic pathways in the brain leading to the formation of adrenergically active neurotransmitters (see Dratman et al. 1984). It should be noticed, however, that T 4 does not produce any change in energy metabolism in the adult brain.
5.2 Studies on Presynaptic Features of Monoamine Neurons
Andersson and Eneroth have demonstrated that the long-term feedback action of thyroid hormones involves an action on hypothalamic catecholamine nerve terminal networks involved in the regulation of thyroid-stimulating hormone (TSH) secretion (Andersson et al. 1985; Andersson and Eneroth 1987). It was found that thyroidectomy leads to an activation of NA nerve terminal systems in the paraventricular hypothalamic nucleus and to an inactivation of the DA nerve terminal systems of the external layer of the median eminence. These effects were reversed by restitution therapy with T 3 or T 4' Chronic but not acute admin­ istration of T 3 and T 4 to the hypophysectomized rat was capable of increasing DA utilization within the external layer of the median eminence, as well as pro­ ducing a reduction ofNA utilization within the paraventricular hypothalamic nu­ cleus. Thus, a long-term feedback action of thyroid hormones on TSH secretion involves inter alia the activation of DA nerve terminals in the median eminence, inhibiting the release of TSH-releasing hormone (TRH). It was suggested that these actions of thyroid hormones involve an activation of nuclear receptors in the medial basal hypothalamus, possibly located in some of the tuberoinfundibu­ lar DA nerve cell bodies. Furthermore, the feedback action of thyroid hormones on TSH secretion also involves the inactivation of a facilitory noradrenergic mechanism in the paraventricular hypothalamic nucleus. It seems possible that the thyroid hormone receptors in this case are located in the cell bodies of the TRH IR neurons, which via recurrent collaterals interact with the paraventricular NA nerve terminal systems, explaining the highly localized change in the paraven­ tricular hypothalamic nucleus.
It is of substantial interest that the modulation of the pituitary thyroid activity could also produce highly discrete changes within the DA nerve terminal net­ works of the forebrain. Thus, thyroidectomy reduces DA utilization in the ante­ rior part of the nucleus accumbens, while increasing DA utilization in the poste­ rior part of the nucleus accumbens, where the costoring cholecystokinin (CCK)/ DA nerve terminals exist. In agreement, chronic but not acute treatment with T 3
or T 4 of the hypophysectomized rat produces an increase in DA utilization within the anterior part of the nucleus accumbens. These results are of substantial inter­ est, since DA nerve terminal systems within the nucleus accumbens participate in the control of locomotion and in the reward mechanisms (Fuxe et al. 1977 b,
36 K. Fuxe et al.
1986a; Ljungberg and Ungerstedt 1978). It may therefore be speculated that the hyperactivity found in patients with hyperthyroidism may be related at least partly to an activation of the dopaminergic mechanism within the anterior part of the nucleus accumbens. The results also imply the existence of nuclear recep­ tors for thyroid hormones within the nucleus accumbens itself.
5.3 Studies on Monoamine Receptor Mechanisms
Many investigations have demonstrated that the number and functional activity of p-adrenergic receptors in peripheral tissues increase within increasing concen­ trations of thyroid hormones in the circulation (Williams and Lefk:ovits 1977).
3H_ Spiperone (striatum)
Sham operation 4 weeks
o 01 0.2 03 0.4 0.5 0.6 B pmol/mg prot.
o 4~~------------------_, ~ ~. Bmi• = 0.433 pmollmg prot.
~ 3 .~KO=0.104nM
o
o
Thyroidectomy 4 weeks
Fig. 29. Effects of thyroidectomy and restitution therapy with T 3 on the binding character­ istics of 3H-spiperone-Iabelled D2 receptors in striatal membranes. The Scatchard plots are shown. A reduction of the Bmax levels is shown following thyroidectomy, as well as preven­ tion of such a fall by replacement therapy with T 3 (10 /lg/kg, twice daily) (see Fuxe et al. 1984a)
Hormones and Synaptic Transmission 37
Within the striatum of the rat brain we have made the observation that changes in the pituitary thyroid activity can produce substantial changes in the density of D2 receptors without changing DA utilization in various parts of the striatum (see Fuxe et al. 1984 a). These results open up the possibility that thyroid hormones may accomplish a heterostatic regulation of the striatal DA synapses, i.e. they may change the sensitivity of the postsynaptic DA receptor mechanism by in­ creasing the number of D2 receptors without any associated changes in the pre­ synaptic features of the DA transmission line (Fig. 29). This regulation has been defined as a heterostatic regulation of the synaptic transmission. As seen in Fig. 30, it allows a change of the set point of the synapse. Thus, hormones such as thyroid hormones may control synaptic heterostasis, which probably repre­ sents an important part of functional synaptic plasticity (see also Fuxe et al. 1984a, 1986a, b; Agnati et aI1986a). Again it is visualized that the thyroid hor­ mones control the DA receptor mechanisms via actions on nuclear thyroid hor­ mone receptors present in the striatal neurons, leading to changes in the ex­ pression of the gene for the proteins carrying the dopamine recognition sites.
HOMEOSTATIC CONTROL V K CO.I High "delil, 10, lhe .ig.... Ironoml .. ion (low o.c illollo ns
o,ound lhe •• 1 poinl)
. .. ; PAc:m!N PHO~VLAnON
HETEROSTATIC CONTROL V (I( , 61() CO.l Funclional o1""Plic pl"licil, (di.placem.nl of Iho HI poonl)
~. R-;!f_v t , I (S.R.CII/OO» 1( , 61( S R-I O
.,~y& FROM THE LOCAL ENDOCRINE
CIRCUIT ~ACRINE SIGNALS
TRANSMISSION LINE
FB LOOP
S S,nlhHio R R ...... I • T,onomiU., binding o T,.noduc:t ion V Inpul 10 lhe irrt, .... lu .. ,
olloClo, X Reo -Reo interaction
Fig. 30. Schematic representation of the mechanisms which may contribute to two types of basic behaviours of synapses: the constancy of the efficacy of the transmission line (synaptic homeostasis) and the change of the level at which this constancy is maintained (synaptic heterostasis)
38 K. Fuxe et al.
6 The Humoral Modulation of Volume Transmission
It should be considered that steroid hormones, thyroid hormones and also pep­ tide hormones can affect some of the circumventricular organs of the central ner­ vous system and in this way influence the release of signals and the neuronal path­ ways from these organs. Some chemical signals may affect the neuronal networks via the ventricular system and the extracellular fluid. GR immunoreactivity exists within the nerve cells of the subfomical organ. The peptide hormones, such as angiotensin and atrial natriuretic peptides, may be active here, since there is no blood-brain barrier in the circumventricular organs and a high density of angio­ tensin II (Saavedra et al. 1986; Healey et al. 1986) and atrial natriuretic factor re­ ceptors (Saavedra et al. 1986; Bianchi et al. 1986) exists in this region.
As stated above, it seems likely that the steroid and thyroid hormones may influence the uptake of messengers and trophic factors in glial cells and neurons, as well as their release from glial cells and neurons. Thus, the release and the up­ take of paracrine signals from neurons reaching a distant receptor population may be highly influenced by steroid and thyroid hormones, which thus can pro­ foundly influence VT in the nervous system.
It is also conceivable that the extracellular fluid pathways which are of sub­ stantial importance in VT may be influenced by hormones in view of the ability of glucocorticoids to control astroglia functions.
7 Aspects on the Organization Principles of the eNS
It seems that the basic texture of some eNS structures is made up of elementary units. Thus, it can be surmised that in the eNS there are at least three types of basic organization: (a) Nuclei (which may be represented by subnuclei of classical anatomical nuclei, e.g. nucleus tractus solitarius), i.e. groups of neurons which operate to perform in an integrative fashion a certain elaboration of information, or to exert a certain trophic action. In particular, it is possible to distinguish the diffuse type of nuclear organization, such as the one of reticular formation, from the compact nuclear organization of the thalamus and the hypothalamus. (b) Elementary circuits, which show a highly repetitive geometry, without clear­ cut morphological boundaries. This organization is represented by the cerebellum and hippocampus. (c) Modules, i.e. repetitive units made up by organized ele­ ments of neuronal structures. We can recognize different types of modules in the eNS. 1. "Structural" modules, i.e. repetitive clusters of cell bodies and/or terminals,
characterized on the basis of transmitters and/or recognition sites for trans­ ducer mechanisms and/or metabolic and trophic features, as found by methods of chemical neuroanatomy. Modules may be heterogenous regarding size and shape. This organization is present in the striatum ("striosomes") (see GraybieI1986).
2. Local circuit modules, i.e. repetitive aggregates of nerve terminals making synaptic contacts with one another controlling a neuronal input and/or a
Hormones and Synaptic Transmission 39
MORPHOfUNCTIONAlORGA ZAllO Of THE DIAN EM! NCE
THE MEDIANOSOME CONCEPT
'ENTITY OF COEIlISTENCE· ... ATTERN OF PEPTIDE FRAG"'ENT'
• NUCLEAR STEROID RECEPTORS
SECRETION Of RElEASl'fG ANO _TORY FACTORS
TRANSPORT OF ~STANCES
Fig. 31. Schematic illustration of the medianosome concept and of the sites of action of pep­ tide, steroid and thyroid hormones
neuronal output. They can be recognized in the median eminence ("mediano­ somes") (Figs. 31 and 32) and in the olfactory bulb.
3. Columns, i.e. repetitive sets of cells which form a vertical structure throughout a brain region and are involved in the integration of specific inputs. They are homogenous regarding size and shape. This organization is present in the somatosensory, visual, motor and frontal association cortex.
Other principles of organization such as somatotopy (i. e. the segregation of a set of neurons or axons in the nuclei or pathways, respectively, according to the body surface area, to which they are connected) can be superimposed on the basic buildup of neuronal circuits. Genetic and epigenetic influences can exert their ac­ tions in the frame of these organizational principles. In the following paragraphs we will develop the concept of the modular organization of the CNS in the frame of the WT and VT.
40 K. Fuxe et al.
MODULAR ORGANISATION OF THE MEDIAN EMINENCE
(THE LHRH INNERVATION AND ITS FUNCTIONAL MODULES)
MEDIAN EMINENCE LEVEL
7.1 Modules of Wiring Transmission
Fig. 32. Illustration of the structural luteinizing hormone­ releasing hormone (LHRH) medianosome and its subdivi­ sion into functional or integra­ tive medianosomes, repre­ sented by the various interac­ tion zones (LHRH/TH, LHRH/ENK etc.). The modu­ lation caused by LH and sex hormones is indicated (for fur­ ther aspects see Fig. 34)
The cortical columns were identifIed on the basis of functional criteria, e.g. the neurons of a column respond to the same sensory stimulus applied to a specific area of the skin. The columns have been shown to be modality specifIC and site specific and were subsequently demonstrated to have an anatomical correlate. Recently, studies in chemical neuroanatomy have supplied indications that the modular organization is not unique to the cerebral cortex but also exists in other brain areas, such as the striatum and the median eminence (Olson et al. 1972; Ten­ nyson et al. 1972; Graybiel 1984; 1986; Fuxe et al. 1971, 1986c; Andersson et al. 1984). Structural modules, defined on the basis of markers of chemical trans­ mission, have been recognized in the striatum. We have recently introduced the
Hormones and Synaptic Transmission 41
idea that there exist "integrative" modules in the brain, formed in the interaction zones between various structural modules (Agnati et al. 1986c; Fuxe et al. 1986 c). The overlap zones between two, three or more structural modules corre­ spond to integrative modules of higher and higher level (Fig. 32). Each integrative module of any level may have different spatial and temporal patterns of activity, resulting in a very ample spectrum of functional states. Such an integrative mod­ ule may be considered as the functional counterpart of the module, which is basi­ cally defined by means of criteria borrowed from chemical neuroanatomy.
7.2 Modules of Volume Transmission
Recently, by means of automatic image analysis, we have also been able to estab­ lish the existence of islands of neurons which contain glucocorticoid receptor im­ munoreactivity within the striatum and the nucleus accumbens (Fig. 33). These results provide the first evidence for the existence, at least in parts of the eNS, of structural modules characterized on the basis of humoral inputs. GR IR is­ lands have a rather uniform distribution and are different from the striatal mod­ ules of the WT mentioned above (see Zoli et al. 1988).
7.3 Functional Aspects on the Modular Organization
The best example of functional interactions between WT and VT modules is pre­ sented by the local circuits formed by various types of transmitter-identified nerve terminals at the median eminence level (Fuxe et al. 1986c). In fact, at median emi­ nence level (an interface area between brain and the endocrine system), the vari­ ous types of transmitter-identified nerve terminal networks form aggregates of nerve terminals, the medianosomes (see Fig. 31), located in distinct parts of the median eminence forming rostrocaudal strips. Such aggregates of nerve terminals regulate the secretion of one hypothalamic hormone (e.g. a releasing hormone) (see Fig. 34). The releasing factor secreted by the median eminence controls the release of one adenohypophyseal hormone. The differential activation of the vari­ ous elements forming a medianosome results in a functional state which will pro­ vide the appropriate output to control anterior pituitary hormone secretion. Re­ ceptors for various hypothalamic and hypophyseal hormones, such as thyreo­ tropic releasing hormone and corticotropin releasing hormone, probably exist in various portions of the cell membranes of the nerve terminals constituting the lo­ cal circuits of the medianosomes (Taylor and Burt 1982; De Souza et al. 1985). Thus, one important role of the hypothalamic and of the hypophyseal hormones is to modulate the activity of the medianosomes, an action which probably under­ lies the ultrashort and short-loop feedback action of hypothalamic and of the hy­ pophyseal hormones, respectively. From the above it becomes clear that one im­ portant site of action of hormones in the brain is a local circuit module, where the activated hormonal receptors of humoral modules interact with activated transmitter receptors of wiring modules to adjust the functional output.
42 K. Fuxe et al.
Fig. 33 A-D. Example of the elaboration of the sampled fields. A Original image as it ap­ pears on the IBAS screen. B The discrimination function has been performed on the orig­ inal image to differentiate grey matter (black in the image) from white matter. This pro­ cedure allows the determination of the field area of the grey matter present in the sampled field. C The discrimination function has been performed on the original image to separate the specific staining (glucocorticoid receptor immunoreactive profiles: white dots in the discriminated image) from the background. This procedure allows the evaluation of the number of the discriminated profiles. D The close function has been performed on the discriminated image C. This procedure allows determination of the field area covered by the islands and the number and the size of the islands, defined as aggregates of three or more original profiles
LHRH MEDIANOSOME
OF PROLACTIN BINDING SITES
PIF MEDIANOSOME
OF NERVE TERMINALS REGULATING LHRH
+ PlF SECRETION RESPECTIVELY (WIRING MODULES)
Fig. 34. Example of a possible interaction between volume transmission modules (humoral modules of prolactin binding sites) and of wiring transmission modules [wiring modules; structural medianosomes for luteinizing hormone-releasing hormone (LHRH) and prolac­ tin release-inhibiting factor (PIP) secretion]
8 Summary
There are two types of chemical transmission in brain, namely wiring trans­ mission (WT) and volume transmission (VT). WT is the classical type of trans­ mission which is neuronally linked and operates with high speed, high safety and short term actions, their divergency and plasticity being low. VT is a type of transmission mainly operating via electronic and paracrine signals, diffusing in the extracellular fluid to reach the appropriate targets. To understand the actions of hormones and paracrine signals on WT, the complexity of the individual synapse must be understood. It consists of multiple transmission lines, which interact with one another at the pre- and postsynaptic membrane via intra­ membrane receptor-receptor interactions and via intracellular postreceptor sig­ nals.
There exists a blood-brain barrier, leading to the exclusion of a number ofpe­ ripheral hormonal signals. However, there also exist chemical and physical "win­ dows" through which the brain receives and delivers messages. Physical windows are represented by brain areas devoid of a blood-brain barrier, such as the area
44 K. Fuxe et al.
postrema, the median eminence and the subfornical organ. The chemical win­ dows are represented by facilitated transport, active transports etc.
Steroid sex hormone-accumulating and estrogen IR nerve cells have been demonstrated in the central nervous system and are concentrated in the limbic forebrain, the medial preoptic area and the hypothalamus. The nuclear location of the estrogen immunoreactivity strongly supports the importance of the genomic actions of estrogens. Estradiol-17fJ, progesterone and androgens pro­ duce discrete changes in dopamine (DA), noradrenaline (NA) and adrenaline (A) levels and utilization within the hypothalamus and the preoptic area in male and female rats in various endocrine states. The monoamines participate in both the central inhibitory and facilitory feedback actions of estradiol-17 fJ on LHRH se­ cretion. The various catecholamine (CA) nerve terminal networks are probably influenced via local circuit interactions, in which the steroid target cells partici­ pate. However, an accumulation of steroid hormones has also been observed in DA and NA nerve cell bodies. Estrogens also influence striatal DA mechanisms, indicating that estrogens can also modulate motor functions and mental activi­ ties, such as mood. These effects may be produced via indirect actions involving the hypothalamus and/or the pituitary gland. However, the vast majority of the effects of estrogens and other gonadal steroids on the presynaptic pro