Functions of the Septo-Hippocampal System

Ciba Foundation Symposium 58 (new series)

Elsevier . Excerpta Medica . North-Holland Amsterdam . Oxford . New York

Functions of the Septo-Hippocampal System

The Ciba Foundation for the promotion of international cooperation in medical and chemical research is a scientific and educational charity established by CIBA Limitednow CIBA-GEIG Y Limited-of Basle. The Foundation operates independently in London under English trust law.

Ciba FounLtion Symposia are published in collaboration with Elsevier Scientific Publishing Company / Excerpta Medico / North-Holland Publishing Company in Amsterdam.

Elsevier / Excerpta Media / North-Holland, P.O.Box 21 1, Amsterdam

Functions of the Septo-Hippocampal System

Ciba Foundation Symposium 58 (new series)

Elsevier . Excerpta Medica . North-Holland Amsterdam . Oxford . New York

Cop-vright 1978 Ciba Foundation

All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying and recording, or by any information storage and retrieval system, without permission in writing from the publishers.

ISBN 0-444-90032-2

Published in July 1978 by Elsevier/Excerpta Medica/North-Holland. P.O. 21 I. Amsterdam and ElsevieriNorth-Holland, Inc.. 52 Vanderbilt Avenue, New York, N.Y. 10017.

Suggested series entry for library catalogues: Ciba Foundation Symposia. Suggested publisher’s entry for library catalogues: Elsevier/Excerpta Medica/North-Holland.

Ciba Foundation Symposium 58 (new series) 446 pages. 95 figures. 8 tables

Library of Congress Cataloging in Publication Data

Symposium on Functions of the Septo-Hippocampal System. London, 1977. Functions of the septo-hippocampal system.

(Ciba Foundation symposium. new ser.: 5 8 ) Bibliography: p. Includes index. I . Septum (Brain)-Congresses. 2. Hippocampus (Brain)-Congresses. I . Title. 11. Series:

Ciba Foundation. Symposium. new wr.. 58. [DNLM : I . Hippocampus-Physiology-Congresses. 2. Septum pellucidum-Physiology- Congresses. W3 C161F v. 58 1977 I WL314 S989f 19771

ISBN 0-444-90032-2 QP383.2.S95 1977 599’.01’88 78-18687

Printed in The Netherlands by Casparie. Alkmaar

1v

Contents

L. WEISKRANTZ Chairman’s introduction 1

J . A . GRAY Objectives of the meeting 3

G. s. LYNCH, G. ROSE and c . M. GALL Anatomical and functional aspects of the septo-hippocampal projections 5 Discussion 20

L. w. SWANSON The anatomical organization of septo-hippocampal projections 25 Discussion 44

J . STORM-MATHISEN Localization of putative transmitters in the hippocampal formation (with a note on theconnections to septum and hypothalamus) 49 Discussion 80

P. ANDERSEN Long-lasting facilitation of synaptic transmission 87 Discussion 102

J . F. DEFRANCE. J . c. STANLEY, J . E. MARCHAND and R. B. CHRONISTER Cholinergic mechanisms and short-term potentiation 109 Discussion I22

General Discussion I: Monoaminergic inputs to the hippocampus 127 130 Nicotinic transmission in the hippocampus

Specific pathways between septum and hippo- campus 138

0. s. VINOGRADOVA and E. s. BRAZHNIK Neuronal aspects of septo- hippocampal relations 145 Discussion 1 7 1

J . O’KEEFE and. A. H. BLACK Single unit and lesion experiments on the sensory inputs to the hippocampal cognitive map 179 Discussion 192

V

VI CONTENTS

C. H. VANDERWOLF. R. KRAMIS and T. E. ROBINSON Hippocampal electrical activity during waking behaviour and sleep : analyses using centrally acting drugs 199 Discussion 22 1

s. P. GROSSMAN An experimental dissection of the septal syndrome 227 Discussion 260

J . A. GRAY, J . FELDON. J . N. P. RAWLINS, s. OWEN and N. MCNAUGHTON The role of the septo-hippocampal system and its noradrenergic afferents in behavioural responses to non-reward 275 Discussion 300

General Discussion IZ: Hippocampal units and theta activity 309 Mechanism of theta activity 319 Functional significance of theta activity 321

D. s. OLTON The function of septo-hippocampal connections in spatially organized behaviour 327 Discussion 343

H. URSIN, T. DALLAND, B. ELLERTSEN, T. HERRMANN, T. B. JOHNSEN, P. J. LIVESEY,

z. ZAIDI and H. WAHL Multivariate analysis of the septal syndrome Discussion 369

351

L. WEISKRANTZ A comparison of hippocampal pathology in man and other animals 373 Discussion 388

Final General Discussion: Frequency potentiation and memory 407 Further comments on the hippocampal cognitive

Functions of the septo-hippocampal system in map theory 412

man and other animals 417

Index of contributors 429

Subject index 431

Participants Symposium on Functions of the Septo-Hippocampal System. held at the Ciba Foundation, London, 18th-20th October 1977

Chairman: L. WEISKRANTZ Department of Experimental Psychology,

P. ANDERSEN The Institute of Neurophysiology, University of Oslo,

University of Oxford, South Parks Road, Oxford OX1 3UD

Karl Johansgt. 47, Oslo I , Norway

E. AZMITIA Department of Anatomy, University of Cambridge, Downing

A. BJORKLUND Department of Histology, University of Lund. Biskopsgatan 5,

Street, Cambridge CB2 3DY

S-223 62 Lund, Sweden

A . H. BLACK Department of Psychology, McMaster University, 1280 Main St

J . F . DE FRANCE Department of Neurobiology and Anatomy, Health Sciences Center, University of Texas Medical School, P.O. Box 20708, Houston. Texas 77030, USA

West, Hamilton, Ontario, Canada L8S 4K

D. GAFFAN Department of Experimental Psychology, University of Oxford,

J . A . GRAY Department of Experimental’Psychology, University of Oxford,

South Parks Road, Oxford OX1 3UD

South Parks Road, Oxford OX1 3UD

s. P. GROSSMAN Department of Behavioral Sciences (Biopsychology), University of Chicago, Green Hall, 5848 South University Avenue. Chicago, Illinois 60637, USA

Section 4 (Neuro- and Psychopharmacology), Pharmacology Research Laboratories, Pharmaceuticals Division, CIBA-GEIGY Limited, K 125.1 109, CH-4002 Basle, Switzerland

w. P. KOELLA

P. J . LIVESEY Department of Psychology, The University of Western Australia.

G. s. LYNCH Department of Psychobiology, University of California, Irvine,

Nedlands, Western Australia 6009

California 927 17, USA

P. MOLNAR Department of Physiology, University Medical School, H-7643 Pecs, Hungary

VII

VIlI PARTICIPANTS

J . O'KEEFE Department of Anatomy and Embryology. University College (London), Gower Street, London WC 1 E 6BT

D. s. OLTON Department of Psychology, The Johns Hopkins University. Baltimore. Maryland 21218, USA

J . B. KANC'K, Jr. Department of Physiology, Downstate Medical Center, State University of New York, Box 131, 450 Clarkson Avenue, Brooklyn, New York 11203, USA

J . N . P. RAWLINS Department of Experimental Psychology, University of Oxford, South Parks Road, Oxford OX1 3UD

T. E. ROBINSON Department of Psychology, University of California, Irvine,

M. SEGAL Isotope Department. The Wekmann Institute of Science, P.O. Box

Caliibrnia 9271 7. USA

26, Rehovot, Israel

B. SREBRO Institute of Physiology, University of Bergen, Arstadveien 19,

Anatomical Institute, University of Oslo, Karl Johansgt.

N-500 Bergen, Norway

47, Oslo 1, Norway J. STORM-MATHISEN

L. w. SWANSON Department of Anatomy and Neurobiology, Washington University School of Medicine, 660 South Euclid Avenue. St Louis, Missouri 631 10, USA

H. URSIN Department of Physiological Psychology, Institute of Psychology, University of Bergen, Arstadveien 21, 5000 Bergen, Norway

c. H. VANDERWOLF Department of Psychology, Faculty of Social Science, University of Western Ontario, London 72, Ontario, Canada

OLGA s. VINOGRAWVA Department of Memory Problems, Institute of Biophysics, Academy Biological Centre, Puschino-on-Oka, Moscow Distr. 142292. USSR

G. WINOCUR Department of Psychology, Trent University, Peterborough,

J. ZIMMER Institute of Anatomy, University of Aarhus, DK 8000 Aarhus C,

Ontario. K91 7B8, Canada

Denmark

Editors: KATHERINE ELLIOTT (Organizer) and JULIE WHELAN

Chairman’s introduction

L. WEISKRANTZ

Department of Experimental Psychology, University of OxJord

We have a very exciting topic for this symposium: we also have a great number of different disciplines and approaches. We hope that out of the entire discussion some kind of synthesis will emerge, or at least some agreement about where and why we disagree. I personally hope that we can keep certain broad issues in mind during the symposium. The striking aspect of the hippocampus is the anatomical elegance of its structure, revealed in detail in the past few years. In contrast there is really appalling ignorance about what this elegance means. I am not sure that the septa1 area is quite so elegant in appearance but the hippocampus certainly is. Anatomical information will be presented to us. In recent years certain favourite anatomical connections have been stressed in discussions of the septo-hippocampal system, but there are older connections too: anatomists keep on adding connections; they very rarely subtract any ! These ‘older’ connections, between for example the hippocampus and thalamus, are rarely stressed now and it may be worth reminding ourselves that we are dealing with a rather more complex anatomical system than just the pathways that we happen to be interested in at the moment.

We shall be given information on neurotransmitters, and it is worth asking ourselves why God made so many different transmitters when presumably it would be possible for all synapses to work on one transmitter, if the only function is to get signals from one neuron to another. The questicn here is what are the functional implications of a dependence on a particular trans- mitter, if there is one, in this particular anatomical system.

We shall hear electrophysiological information also, and I hope in discussion we can ask ourselves.how we go from the single cell, which is the favourite unit of analysis these days, to populations of cells and their organizational structure. That is a standard problem but by no means a simple one to overcome.

1

2 L. WEISKRANTZ

We shall also be discussing function. There are a great many different approaches to the question of function. Historically, the suggested role of the hippocampus in memory is one with which we are all familiar. We are also familiar with its suggested role in emotional behaviour-in frustration and punishment-and relevant to that will be information on the endo- crinological aspects of the hippocampus. We know now about its possible role in spatial functions. We know that it seems to have some connection with voluntary movement. We know about its suggested role in various kinds of disinhibitory and perseverative behaviour, and in habituation.

I hope we can direct our attention to three questions: first, is it possible to subsume all these different kinds of functions under a single heading? Is there some fundamental scheme in the Platonic sense which we see the shadows of, from time to time, in the various hypotheses that have been put forward? It would be interesting and important to see whether we can get agreement on that. The second question is: what are the ways in which the different suggested functions can be studied independently? Is an animal, for example, who is impaired on spatial function going to be impaired on various other aspects of'behaviour because he has a spatial deficit? We must know the answer to this before we can decide whether we are dealing with several independent functions. Finally, the third question is related to the second but is not quite the same: the title of the symposium contains the words 'septo-hippocampal system'. By the end, we may want to know whether we are in fact dealing with systems in the plural, rather than one single system.

Objectives of the meeting

J. A. GRAY

Department of E.rperimental Psychology. University of O.uford

The sciences of brain and behaviour are beginning to knit together. So far. they have produced little more than a fringe, outlining the way in to the brain (perception) and the way out (motor processes). But it is now becoming possible to weave at least a coarse fabric with which to clothe the middle; and the septo-hippocampal system looks to be a favourable point at which to get on with the weaving.

The anatomy of the septo-hippocampal system is relatively well worked out. by comparison at least with other parts of the limbic system. It is, furthermore. an anatomy which is above all orderlv: so orderly, in fact, that guessing from its structure to its function has become a favourite pastime of arm-chair theorists. All too often, indeed, such theorists have relied solely on anatomy (bolstered perhaps by computer analogies) in their speculations. But this is not because other data are lacking. On the contrary, since the early 1960s there has been a’growing flood of reports of the behavioural effects of various kinds of damage to, or stimulation of, the different parts of the septal area and the hippocampal formation ; and of the behavioural correlates of various kinds of electrical recording from these structures. Thus there is plenty of opportunity for controlled speculation as to the functions of the septo- hippocampal system; and that, I believe, is the chief business of this symposium.

Of course, such speculation has been going on non-stop for a long time. But it is an odd feature of research in this fieid that workers who use a particular technique take account only of the results of other workers using similar methods to their own. This has led to quite divergent views all becoming firmly accepted, each by the adherents of only one technique. For example, students of the behavioural correlates of the hippocampal theta rhythm for the most part believe this to be related to motor behaviour; those who study the effects of lesions in animals have generally been persuaded that the septal

3

4 J. A. GRAY

area and the hippocampus exercise some sort of inhibitory function ; and those who study lesions in man have usually credited the hippocampus with a role in memory. These different groups of workers may refer to each other's results; indeed, they often publish within the covers of the same volumes: but one gets the strong impression that they never really talk to each other. We hope that they will do just that during the next three days.

The situation I have described has been going on for quite some time: why, then, is now a better time to talk than hitherto? There are several reasons. First, there have recently appeared a couple of volumes which have brought together many of the data about the hippocampus (Isaacson & Pribram 1975) and the septal area (DeFrance 1976) taken separately; though, oddly, this is the first symposium to consider them together, in spite of the well-known close connections between the two structures. Thus we can concern ourselves principally with understanding the significance of these data, rather than with recording them. Second, in the past five or six years there have been very important advances-and some dramatic changes-in our knowledge of the basic anatomy and physiology of the septo-hippocampal system; but the implications of these advances have not yet 'been fully worked out. Third, there have developed, in the same period, a number of radically new approaches, often based on new experimental methods, to the behavioural functions of the septo-hippocampal system ; but these new ideas have not yet been fully brought up against each other, or against the data which supported older ideas.

We hope, therefore, that the next three days will force us to look carefully at all of the data, old and new, anatomical, physiological and behavioural, dealing with the septal area and with the hippocampal formation, in an effort to get to grips with the question: what are the functions of the septo- hippocampal system? At each step in our scrutiny, we should ask: What do we know? What ought to be done to resolve those points about which we are ignorant? Which are the issues which really divide us, and which are the ones which we merely describe differently?

References

DEFRANCE. J. (ed.) (1976) 7he Septa1 Nuclei. Plenum Press. New York ISAACSON, R. L. & PRIBRAM, K. H. (eds.) (1975) The Hippocampus, 2 vols.. Plenum Press. New York

Anatomical and functional aspects of the septo-hippocampal projections

GARY LYNCH*. GREG ROSE? and CHRISTINE GALL*

Department of Psychobiology* and School of Social Sciencest, University of California. Irvine. California

Abstract The origins, distribution, and cellular targets of the septo-hippocampal projections are reviewed. It appears that the distribution of acetylcholinesterase- positive neurons in the medial septum and diagonal bands and those cells labelled after injections of horseradish peroxidase into the hippocampus coincide ; however, the possibility of a non-acetylcholinesterase septal projection remains. Good agreement is found between the distribution of hippocampal acetylcholinesterase and the patterning of silver grains after injection of [3H]leucine into the medial septum. A major target of septal efferents to the hippocampus is the interneuron population ; the possibility of septal mediation of intrahippocampal circuitry via this anatomical arrangement is discussed.

The hippocampus of mammals receives two major projections from the telencephalon, and these are in several important respects quite different from one another. By far the dominant afferent of the hippocampus is that from the entorhinal cortex ; this projection generates a tightly packed, extremely well developed terminal field which covers nearly 70% of the dendritic tree of the granule cells of the dentate gyrus as well as a significant portion of the apical dendrites of the pyramidal cells (Blackstad 1958; Raisman et al. 1965 ; Hjorth-Simonsen 1972; Hjorth-Simonsen & Jeune 1972).

Quantitative electron microscopic studies have indicated that in some regions the entorhinal terminals generate as much as 90% of the total synaptic population (Matthews et al. 1976; Lee et al. 1977). In marked contrast to this is the septo-hippocampal afferent system. These fibres are few in number and rather than restricting themselves to specific laminae they are found scattered throughout several sites in hippocampus (Raisman 1966;

Mosko et al. 1973). The contrasting anatomical properties of the two input systems would imply that the septum exerts a quite different type of influence

5

6 G. LYNCH et al.

on the physiology of the hippocampal formation than does the entorhinal cortex. It is this idea which serves as the theme of this review.

The understanding of the entorhinal-hippocampal connections has benefited greatly from recent neuroanatomical and electrophysiological investigations. Stated simply, entorhinal fibres terminate on the outer portions of the dendrites of the granule and pyramidal cells (Nafstad 1967; Steward 1976) and when activated produce robust extracellular postsynaptic potentials (EPSPs) in those cells (Andersen et al. 1966; Dudek et al. 1976). It seems evident that this projection provides a powerful relay between thalamic and cortical sites (Van Hoesen & Pandya 1975) and the primary cells of the hippocampus (i.e. the granule cells of the dentate gyrus).

Unfortunately the system of septo-hippocampal connections has not proved as amenable a subject for either anatomical or electrophysiological work. The region from which the system is generated is small and is traversed by fibre bundles travelling to and from the hippocampus, confounding analysis of the origin of the projections which actually arise from within the septum. Furthermore, the sparse and scattered nature of the distribution of the septal axons within the hippocampus provides any number of difficulties for tracing experiments (Mosko et al. 1973). However, the application of recently developed neuroanatomical methods has provided valuable new information, with the result that a clearer picture of the organization of the septal afferents of the hippocampus has begun to emerge. In this paper we shall make use of these data to reexamine the origins, distribution. and cellular targets of the septo-hippocampal projections. Following this, brief consideration will be given to the question of what role the septal afferents play in the operation of the hippocampus and an attempt will be made to link the data from the anatomical studies with the results of neurobehavioural investigations.

THE ORIGINS OF THE SEFTAL PROJECTION INTO HIPPOCAMPUS

Until quite recently, information on the location of neurons which generate the septo-hippocampal projection came primarily from studies using the anterograde degeneration method and acetylcholinesterase histochemistry (AChE; EC 3.1.1.7). With regard to the former, Raisman (1966) reported that the medial but not lateral septum projected into the hippocampus and dentate gyrus. These results were somewhat compromised by the fibre-of- passage problem; that is, it was difficult to be certain that the observed degeneration originated from neurons located in the septum rather than from fibres interrupted by the lesion which pass through the septal area. Studies by Lewis & Shute (1967) using the AChE histochemical method demonstrated

SEPTO-HIPPOCAMPAL PROJECTIONS 7

FIG. 1. After injection of horseradish peroxidase (30% w/v) into subfield CAI of the rostra1 hippocampus of the rat (illustrated schematically in A by the stippled area) retrogradely labelled neurons were observed in the nucleus of the diagonal bands both ipsilateral and contralateral to the injection site. B shows a coronal section through the septa1 area which demonstrated such a labelling pattern. The small boxes in the lower portion of the figure identify the location of two neurons shown at higher power in c and D. Calibration: B, bar = 200 pm; C,D, bar ~ 25 pm.

8 G . LYNCH et al.

that the medial septum-diagonal band area contained a collection of densely staining (i.e. AChE-positive) cells and by means of a series of clever lesion experiments the authors were able to conclude that these cells generated a septo-hippocampal projection.

The experiments of Segal & Landis (1974) and, more recently, Meibach & Siege1 (1976) using the method of retrograde transport of horseradish peroxidase (La Vail et al. k973) have shed new light on the question of the origins of the septal afferents to the hippocampus. Their results largely confirm the conclusions of the earlier work but add the additional and important information that the septo-hippocampal projection exhibits a topographical relationship. It appears that the most medial aspects of the septum innervate the rostra1 hippocampus while more lateral portions of the medial nucleus and adjacent area send fibres to progressively more caudal and ventral regions of the hippocampal formation. Whether this organization holds true for the entire length of the medial septal nucleus-diagonal band complex is not clear.

From these studies it appears that hippocampal afferents from the septum are exclusively ipsilateral in origin. However, our own experiments using horseradish peroxidase histochemistry demonstrate that a small but significant number of cells which send their axons to the hippocampus are located in the contralateral medial septum4iagonal band complex of the rat (Fig. 1). This result supports an earlier observation (Mellgren & Srebro 1973) of a bilateral loss of hippocampal acetylcholinesterase after discrete unilateral lesions of the medial septum. That others using horseradish peroxidase were unable to identify this bilateral origin for the septo-hippocampal projection may have been the result of using a less sensitive chromagen in their development procedure (Mesulam 1976).

Taken together the evidence from the horseradish peroxidase and acetyl- cholinesterase studies is quite good that a septo-hippocampal system exists and originates within the medial and ventral aspects of the septal complex; however, it is not clear that the projections described by the two methods are the same. The possibility remains that there exist septal afferents of the hippocampus which do not contain AChE. A necessary first step in investigating this question is a comparison of the distribution and appearance of the AChE-positive cells with those cells which are labelled after large injections of horseradish peroxidase into hippocampus.

Fig. 2 summarizes the location of the AChE-positive neurons in the septum and nucleus of the diagonal bands. The animal used for this figure was treated with diisopropylfluorophosphate (DFP), an irreversible inhibitor of AChE, and then allowed 24 hours to synthesize new enzyme before sacrifice.

SEPTO-HIPPOCAMPAL PROJECTIONS 9

.... ..’ ::. .... :: .*:,

. . .

1200 p 880 p

FIG. 2. Schematic representation of acetylcholinesterase-positive neurons located in the septal region after pretreatment with diisopropylfluorophosphate. The number of pm below each drawing indicates the position of the section from the anterior end of the nucleus of the diagonal bands.

Since this enzyme is produced only in the cell soma, at this short survival interval AChE-containing neurons stain heavily while the dendritic and axonal processes (which have not yet received the enzyme by somatofugal transport) remain unstained. The result is that the location and features of the AChE- positive cells can be seen unobscured by neuropil staining (Lynch et al. 1972). A comparison of material treated in this way with control (i.e. no DFP) rats provides a more complete picture than is found with AChE histochemistry alone.

In general, AChE-positive cells are found throughout the medial septal nucleus and diagonal bands region. These cells do not, however, constitute a homogeneous population. On the basis of differences in morphology and

10 G . LYNCH et al.

FIG. 3. Low power photomicrograph of the septum stained by AChE histochemistry in a rat previously treated with diisopropylfluorophosphate. At this coronal level the various AChE- positive cell types appear in separate fields. The cells of the central portion of the medial septal nucleus (MSN) are noticeably more faintly stained than those of the dorsal MSN ‘cap’. The cells of the diagonal band region, seen in an ovoid configuration near the ventral face of the section, are similarly distinct from central MSN cells on the basis of staining intensity and gross morphology. A small population of intensely stained ‘midline cells’ comprise the fourth distinct AChE-positive subgroup. Bar = 500 pm.

staining intensity several distinct cell groups can be described. This heterogeneity is most apparent at mid-septa1 levels where the various groups occupy distinct regions of the complex (Fig. 3). At this level the majority of the histochemically identified somata in the medial septal nucleus belong to small, lightly staining-bipolar cells. The few visible processes of these neurons are aligned almost exclusively dorsoventrally. In contrast to the medial septal nucleus cells, AChE-positive cells of the diagonal band region are larger and much more intensely stained, with branching dendritic and occasional axonal processes visible. As can be seen in Figs. 3 and 4c, these

SEPTO-HIPPOCAMPAL PROJECTIONS

Fit;. 4. Photomicrographs of AChE-positive cells in the medial septum-diagonal band complex in a rat pre-treated with diisopropylfluorophcsphate. The densely staining cells of micrograph ( c r J

found at the periphery of the diagonal bands ‘ovoid group’ illustrate the excellent delineation of basic morphology possible with this technique. Micrograph ( b ) , taken at the midline of the medial septa1 nucleus, shows the process (arrow) of a cell on the right half of that nucleus cross the midline to bifurcate on the contralateral side. Although not frequent, such ‘crossing’ processes are regularly observed. The ovoid configuration of the diagonal band AChE-positive group is apparent in micrograph (c). The disposition ofcell bodies as well as the alignment of their processes creates an ovoid AChE-positive ring. Calibration: a,b, bar = 2 5 pm; c, bar 100 pm.

12 G. LYNCH et al.

neurons are arranged in an ovoid cluster ventral and medial to the nucleus accumbens septi. The dendritic processes appear to conform to the curve of the ovoid ring with the centre remaining cell- and process-poor. Two smaller but distinct AChE-positive cell populations are found along the midline and at the dorsal tip of the medial septal nucleus. The midline neurons stain intensely and have an oval, elongated soma; occasionally one thick process aligned with the midline is evident. The cells which ‘cap’ the medial septal nucleus seem to represent yet another distinct group in that they are somewhat larger and much more intensely stained than the neurons of the medial septal nucleus and more frequently display dendritic processes out of the dorsoventral alignment.

The distribution of the neurons which were labelled after a large injection of horseradish peroxidase in the rostral hippocampus is shown in Fig. 5 . The labelled cells are found along the entire anterior-posterior extent of the medial septal nucleus and nucleus of the diagonal band and in this sense are co-extensive with the AChE-positive cells. In the case illustrated the neurons are restricted to a medial position but this may be due to the fact that the enzyme injection was limited to the rostral hippocampus. (As previously mentioned, Meibach & Siege1 (1977) found more laterally placed neurons after injections of horseradish peroxidase into the ‘temporal’ aspects of the hippocampus). The retrograde horseradish peroxidase method provides only a minimum of cytological detail but it was evident from these experiments that several classes of cells corresponding approximately to those described using the AChE histochemical method were labelled.

The observation that AChE- and horseradish peroxidase-positive cells are co-extensive in their distribution suggests that the septo-hippocampal system may be composed entirely of cholinesterase-positive elements. As will be seen, this conclusion would also be in agreement with the results of studizs comparing the organization of the septal projections in hippocampus as demonstrated by autoradiography and AChE histochemistry. Despite this, a caveat is in order4xperiments in which AChE and horseradish peroxidase histochemistries have been combinzd have indicated the presence of nume rous retrogradely labelled cells which do not stain for acetylcholinesterase (Rose & Lynch, unpublished data; Mesulam et al. 1977). Whether this is due to inadequate staining with the AChE method or provides evidence of non-AChE septal projections to hippocampus remains a subject for further work. At the present, it can be concluded that: (1) the distribution of acetylcholinesterase- containing neurons corresponds quite well with the origins of the septo- hippocampal system and (2) it is likely that several morphologically distinct types of AChE-positive cells project to the hippocampus.

SEPTO-HIPPOCAMPAL PROJECTIONS 13

240 p

A 560 u

D I. . ,

Q

880 p 1200 u

FIG. 5 . Schematic representation of the distribution of retrogradely labelled neurons after an injection of 30 ”/, horseradish peroxidase into the rostra1 hippocampus. Anterior-posterior levels are indicated as in Fig. 2.

THE DISTRIBUTION AND TARGETS OF THE SEPTO-HIPPOCAMPAL SYSTEM

Again work using the anterograde degeneration method and AChE histo- chemistry has provided useful descriptions of the distribution of the septal projections in the hippocampus ; furthermore, studiescomparing the projections as shown by the two methods have found good although not complete agreement. Fig. 6 illustrates the distribution of the acetylcholinesterase staining in a cross-section of hippocampus and compares this with the pattern seen after a large injection of concentrated [3H]leucine into the medial septal area. In the dentate gyrus the two patterns are quite similar in that the densest innervation is in the infragranular region while a thin band is detected immediately above the granule cells. AChE histochemistry provides evidence

14 G . LYNCH et al.

FIG. 6. Coronal sections through the rostral hippocampus stained for acetylcholinesterase (top) and after a large injection of [3H]leucine into the medial septum (bottom). In general, there is good correspondence between AChE-positive sites in the rostral hippocampus and those areas demonstrating radioactive labelling after injection of the tritiated amino acid into the septum. For further details see text. Bar = 500 pm.

for a modest projection in the middle molecular layer which has been difficult to establish satisfactorily with the anterograde degeneration method (Mosko et 01. 1973) but is somewhat better illustrated by autoradiography. In the regio inferior pyramidal cell fields, the heaviest innervation as revealed by both methods is in the stratum oriens. The projections of the septum to the apical dendritic fields of the pyramidal cells is sparse and comparisons between methods are accordingly difficult; the stratum moleculare of CAI is notably devoid of input while a very thin zone between this region and the

SEPTO-HIPPOCAMPAL PROJECTIONS 15

stratum radiatum can be seen to receive a projection by both AChE histo- chemistry and autoradiographic tracing. A band of moderate intensity of AChE activity is found along the basal axis of the CAI pyramidal cells but this has no apparent counterpart with either autoradiographic or Fink-Heimer tracing methods.

These results support earlier conclusions (Mosko et al. 1973; Mellgren & Srebro 1973) that the pattern of the acetylcholinesterase staining in hippo- campus corresponds quite closely to the distribution of the septo-hippocampal projection. The AChE pattern is the slightly more extensive of the two and this may indicate that the hippocampus receives or generates AChE-positive afferents other than those from the septum (Storm-Mathisen & Blackstad 1964). Alternatively, it is possible that these discrepancies are due to the sparse nature of the projections in question-as the number of fibres and terminals which contain radiolabel decreases it becomes increasingly more difficult to distinguish them with certainty from background.

In a p9sitive sense, it is evident that the septum does not send any major projections into hippocampus which terminate outside the zones receiving the AChE-positive fibres. In light of the earlier discussion of possible non-AChE components of the septo-hippocampal system it seems likely that if two types of septal inputs exist, they are intertwined and have the same regions for targets.

The distribution of the septo-hippocampal projections as described above is highly suggestive with regard to the cellular targets of these axons. The infra-granular region contains a heterogeneous population of interneurons and is essentially devoid of dendrites from either the granule or pyramidal neurons (Ramon y Cajal 1911). Therefore, the observation that this area contains a dense concentration of septal axons and terminals strongly suggests that these fibres are targeted for one or more types of interneurons (Mosko et al. 1973). The distribution of septal projections in the regio inferior also suggests this conclusion, in that the location of the terminals and a previously described group of interneurons coincides.

That this interneuron population is the target of septal afferents has been partially confirmed by autoradiographic studies using [3H]adenosine. This nucleoside readily crosses synaptic junctions, thereby providing a means of identifying the target cells of neurons which have transported it down their axons (Schubert & Kreutzberg 1975). After the injection of [3H]adenosine into the medial septum, the compound (or one of its derivatives) is transported into the hippocampus and with appropriate survival periods is found in a number of postsynaptic targets (Rose & Schubert 1977). Fig. 7 shows an example of a densely labelled cell in the infragranular region of the dentate

16 G. LYNCH el a/.

FIG. 7. Lefr. Camera-lucida drawing of a Golgi-stained cell in the hilus of the dentate fascia generally described as an interneuron. Right: Intrahilar cell labelled two days after the injection of [3H]adenosine into the septum. Radioactive label is seen to extend into the proximal portions of the cell dendrites (arrow). Both the horizontal orientation of its dendritic process and its position suggest that this neuron is of similar type to the cell drawn on the left. Bar = 50 pm.

gyrus which resembles a type of interneuron commonly encountered in Golgi material. Cajal (191 I ) discussed this class of cells and noted that their axons ramified in the dendritic field of the granule cells. Other interneuron 'types' have been labelled in stratum oriens of CA3, but identification of these is less certain because of possible confusion with the displaced pyramidal cells also occasionally found in this layer. The granule cells are also labelled after septal injections of [3H]adenosine, although not as heavily as are the interneurons This may suggest that although septal axons terminate on both these hippocampal cell types, a greater percentage of the innervation is targeted for the interneuron population.

From these observations it is not possible to decide if any selectivity exists in the septal innervation of interneurons; of particular interest is the question of whether those cells whic$ generate the presumably inhibitory basket plexuses on the granule cells are involved. In any event, the relationship between septal fibres and interneurons provides a mechanism whereby the numerically inferior septal projections to the dentate could exert potent physiological effects in the face of the relatively massive perforant path input. Specifically, the interneurons generate extensive axonal arborizations and in

SEPTO-HIPPOCAMPAL PROJECTIONS 17

some cases these make synaptic contacts with hundreds of granule cells (Ramon y Cajal 191 I). Furthermore, the location and inhibitory character of the basket endings is such as to exert a profound control over the physiology of the granule cells (Andersen et al. 1964, 1966). Thus by innervating the interneuron population the numerically inferior septal inputs to the hippo- campus could exert an influence which would be far greater than could be achieved by their direct termination on hippocampal pyramidal or granule cell dendrites.

These anatomical observations provide a suggestion about the type of role that the septum might perform in the operation of the hippocampus. If, as argued above, these projections are primarily targeted for interneurons, then activation of the septum could be expected to alter the inhibitory control that these local cells exert over the granule and pyramidal cells. In this context, a recent electrophysiological study (Alvarez-Leefmans & Gardner- Medwin 1975) demonstrated that septal stimulation could regulate to a remarkable degree the efficacy of perforant path volleys in eliciting granule cell discharge. An explanation for this result could well be supplied by the hypothesis that septal contacts on interneurons modulate their activity and thus allow the 'biasing', in effect, of the response of hippocampal neurons to their excitatory inputs.

SOME FUNCTIONAL IMPLICATIONS OF THE ANATOMICAL DATA

The anatomical data discussed above suggest that the septal projections may serve to modulate the response of the hippocampus to the activation of its massive afferent input from the entorhinal cortex. Ultimately, evaluation of this and similar hypotheses must come from experiments in which the activity of hippocampal neurons after the activation of entorhinal projections is studied over time and the contributions of the septal inputs are carefully analysed. Neurobehavioural studies done in collaboration with Dr S. A. Deadwyler have been conducted which, we hope, will be appropriate for these purposes, and the initial results have been encouraging. Obviously, the first condition that had to be met was the development of a paradigm which led to the reliable discharge of the entorhinal projections. After considerable experimentation, it was found that a tone cue would generate a robust evoked response in the outer molecular layer of the dentate gyrus as a rat learned that operant responses performed within seconds of the signal would be reinforced.

Since the laminar distribution of the tone-evoked response (as recorded with a movable microelechode) was nearly identical to that produced by electrical stimulation of the entorhinal cortex it is very likely that the recorded

18 G . LYNCH, et a/.

potentials are due to a synchronous discharge of the perforant path. Sur- prisingly, the evoked response produced under these circumstances did not elicit reliable discharge of the granule cells. However, when a second, non- rewarded tone was added to the situation and the rat was required to dis- criminate between tones, the evoked responses in the outer molecular layer were followed by prolonged high frequency discharges of the granule cells, although addition of the second tone caused no discernible increase in the amplitude of the evoked potential to either tone. Eventually very different patterns of granule cell discharge were established to two tones, the prolonged burst of cell tiring being seen in association with the positive (i.e. rewarded) tone, and an initial burst followed by a rapid return to background levels after the negative (non-rewarded) tone.

From these results it appears that activation of the entorhinal projections, even in a synchronous fashion, is not by itself an adequate condition to discharge the granule cells and hence initiate activity in the first segment of the intrahippocampal circuit. This strongly implies that the ‘responsivity’ (threshold) of the granule cells is under the control of an agency other than the entorhinal cortex and it is likely that this control is exerted at the level of the cell body. A regulatory knechanism of this type resembles the function ascribed to the septo-hippocampal system in the hypothesis discussed above. Studies in which the firing behaviour of the neurons in the medial septa1 nucleus and diagonal bands complex are followed in the one- versus two-tone behavioural situation (i.e. in two circumstances in which the response of the granule cells to an entorhinal input is very different) could provide a partial test of these ideas.

That the primary hippocampal circuit is not triggered even in a behavioural situation which results in the synchronous discharge of the entorhinal afferents to the dentate gyrus serves as a reminder that even the suggestively ‘simple’ organization of the hippocampus does not make this structure an easy target for functional analysis. Afferent strength is normally predicted by terminal number; however, from these experiments the conclusion must be drawn that in a functional sense this may not always hold true. In the dentate gyrus afferent projection fibres from the perforant path constitute by far the dominant input to this structure. It could be, however, that in normal (i.e. behavioural) circumstances even this massive innervation is unable to trigger discharge of the postsynaptic granule cells without some kind of ’cooperation’ from other inputs. Appreciation of the circumstances which lead to the appropriate interactions of these very different afferent systems may provide some of the clues needed to understand the physiological and behavioural processes in which the hippocampus participates.

SEPTO-HIPPOCAMPAL PROJECTIONS 19

ACKNOWLEDGEMENTS

We wish to acknowledge the skilful technical assistance of Ms Brigitta Flick and Ms Jo Ann Hendrickson and the perseverance of Ms Darlene Thompson in typing the manuscript. This work was supported in part by grant BNS 76-17370 from the National Science Foundation as well as a Career Development Award to G.L. C.M.G. was supported by a predoctcral fellowship from the National Institutes of Health.

References

ALVAREZ-LEEFMANS, F. J. & GARDNER-MEDWIN, A. R. (1975) Influences of the septum on the hippocampal dentate area which are unaccompanied by field potentials. J. Physiol. (Lond.) 249. 14-16

ANDERSEN, P.. ECCLES, J. C. & LBYNING, Y. (1964) Pathway of postsynaptic inhibition in the hippocampus. J . NeurophjJsiol. 27, 608-6 19

ANDERSEN, P.. HOLMQVIST, B. & VOORHOEVE. P. E. (1966) Entorhinal activation of dentate granule cells. Acta Physiol. Scand. 66, 448-460

BLACKSTAD, T. W. (1958) On the termination of some afferents to the hippocampus and the fascia dentata. Actu Anur. 3.7, 202-214

DUDEK. F. E.. DEADWYLER, S. A,, COTMAN, C. W. & LYNCH, G. (1976) Intracellular responses from granule cell layer in slices of rat hippocampus: perforant path synapse. J . Neurophysiol. 39. 384-393.

HIORTH-SIMONSEN. A. (1972) Projection of the lateral part of the entorhinal area to the hippo- campus and fascia dentata. J . Comp. Neurol. 146, 219-232

HJORTH-SIMONSEN, A. & JEUNE, B. (1972) Origin and termination of the hippocampal perforant path in the rat studied by silver impregnation. J . Comp. Neurol. 144. 215-232

LA VAIL, J. H.. WINSTON, K. R. & TISH, A. (1973) A method based on retrograde intra-axonal transport of protein for identification of cell bodies of origin of axons terminating within the CNS. Bruin Rrs. 58, 470-477

LEE, K. S., STANFORD. E. J . . COTMAN, C. & LYNCH, G. (1977) Ultrastructural evidence for bouton proliferation in the partially deafferented dentate gyrus of the adult rat. E.rp. Bruin Res. 29. 475-485

LEWIS, P. R. & SHUTE. C. C. D. (1967) The cholinergic limbic system: projections to hippocampal formation, medial cortex, nuclei of the ascending cholinergic reticular system and the sub- fornical organ and supra-optic crest. Bruin 90. 521 -537

LYNCH, G. S., LUCAS, P. A. & DEADWYLER, S . A. (1972) The demonstration of AChE containing neurons within the caudate nucleus of the rat. Bruin Res. 45. 617-621

MATTHEWS, D. A., COTMAN. C. & LYNCH, G. (1976) An electron microscopic study of lesion- induced synaptogenesis in the dentate gyrus of the adult rat. I . Magnitude and time course of degeneration. Bruin Res. 115, 1-21

MEIBACH. R . C. & SIEGEL. A. (1977) Efferent connections of the septal area in the rat: an analysis utilizing retrograde and anterograde transport methods. Brain Res. 119, 1-20

MELLGKEN, S. 1. & SREBRO, B. ( 1973) Changes in acetylcholinesterase and distribution of degener- ating fibres in the hippocampal region after septal !esion in the rat. Bruin Rex 52, 19--35

MESULAM, M. ( 1976) The blue reaction product in horseradish peroxidase neurohistochemistry : incubation parameters and visibility. J . Hisrochem. Cytochem. 24, 1273-1280

MESULAM, M., VAN HOESEN, G. & ROSENE, D. G. (1977) Substantia innominata, septal area and nuclei of the diagonal band in rhesus monkey: organization of efferents and their acetyl- cholinesterase histochemistry. Neurosci. Abstr. 3, 202

MOSKO, S., LYNCH, G. & COTMAN, C. W. (1973) The distribution of septal projections to the hippocampus of the rat. J . Comp. Neurol. 152, 163-174

20 DISCUSSION

NAFSTAD, P. H. J. (1967) An electron microscopic study of the termination of the perforant path fibers in the hippocampus and the fascia dentata. Z . ZeNforsch. Mikrosk. Anat. 76, 532-542

RAISMAN, G. (1966) The connexions of the septum. Brain 89, 317-348 RAISMAN, G., COWAN, W. M. & POWELL, T. P. S. (1965) The extrinsic afferent, commissural and

association fibres of the hippocampus. Brain 88, 963-996 RAMON Y CAJAL, S. (191 I ) Hisrologie du SysfPrne Nerveux de /'Hornme et drs VPrtPbres, vol. 11,

Maloine, Paris ROSE, G. & SCHUBERT, P. (1977) Release and transfer of [3H]adenosine derivatives in the cholinergic

septa1 system. Bruin Res. 121, 353-357 SCHUBERT, P. & KREUTZBERG. G. W. (1975) ['Hladenosine, a tracer for neuronal connectivity.

Bruin Res. 85, 317-319 SEGAL, M. & LANDIS, S. (1974) Afferents to the hippocampus of the rat studied with the method

of retrograde transport of horseradish peroxidase. Brain Res. 78, 1-1 5 STEWARD, 0. (1976) Topographic organization of the projections from the entorhinal area to the

hippocampal formation of the rat. J. Comp. Neurol. 167, 285-314 STORM-MATHISEN, J. & BLACKSTAD, T. W. (1964) Cholinesterase in the hippocampal region.

Distribution and relation of architectonics and afferent systems. Acra Anar. 56, 216-253 VAN HOFSEN, G. W. & PANDYA, D. N. (1975) Some connections of the entorhinal (area 28) and

perirhinal (area 35) cortices of the rhesus monkey. I. Temporal lobe afferents. Brain Res. 95,l-24

Discussion

Gray: In your neurobehavioural experiment I didn't understand whether the changed unit response of the granule cells occurred first to the unreinforced second tone and only then started occurring to the reinforced tone, or whether (after you introduced the unreinforced tone) it first occurred to the reinforced tone. That is critical for the interpretation of the experiment.

Lynch: Because of the random pattern in which the tones are presented and the averaging procedures we used, it is not possible to give a satisfactory answer to your question.

Gray: If you keep going until the discrimination is learnt perfectly, does the unit response to the correct tone start dying out in the granule cells or does it stay permanently?

'

Lynch: The unit response appears to be permanent. Winocur: Does that pattern entirely reverse when the reward value of the

Lynch: When the value of the tone is reversed, yes. the pattern of unit

Winocur: Have you done this over a series of reversals? Lynch: We did two reversals-we can't go much further than that because

of the difficulty of maintaining electrodes over several days. Weiskrantz: To what extent is this pattern dependent on the actual emission

of a response?

stimuli reverses?

responses also reverses.