Sympathetic sprouting in the central nervous system: a model for studies of axonal growth in the mature mammalian brain

Brain Research Reviews, 12 (1987) 203-233

203

Elsevier BRR 90063

Sympathetic sprouting in the centra1 nervous system: a mode1 for studies of axonal growth in the mature mammalian brain Keith A. Crutcher Department of Anatomy,

University of Utah, School of Medicine, Salt Lake City, UT84132

(U.S.A.)

(Accepted 14 October 1986) Key w0rd.s: Sympathetic; Sprouting; Nerve growth factor (NGF); Hippocampus; Plasticity; Blood vessel; Alzheimer’s disease

CONTENTS 1. Introduction

.............................................................................................................................................

204

2. Sympatheticinnervation of extracerebral targets ................................................................................................ 2.1. Sympathetic targets outside the cranium .................................................................................................... 2.2. Intracranial sympathetic fibers ................................................................................................................

204 204 205

3. Centra1 noradrenergic neurons ......................................................................................................................

206

4. Sympatheticsprouting into the CNS ................................................................................................................ 411. Growth potential of mature sympathetic neurons ......................................................................................... 4.2. Discovery of sympathohippocampal sprouting ............................................................................................ 4.3. Specificity of sympathohippocampal sprouting ............................................................................................ 4.4. Topography of sympathohippocampal sprouting ......................................................................................... 4.5. Sympathetic sprouting in other brain regions .............................................................................................. 4.6. Age and hormone influences on sympathetic sprouting ................................................................................. 4.7. Uhrastructural relationships of septal and sympathetic fibers to hippocampal elements .........................................

208 208 208 209 211 212 213 214

5. Sprouting mechanisms ................................................................................................................................ 5.1. Role of afferent input ........................................................................................................................... 5.2. Target regulation ................................................................................................................................. 5.3. The NGF hypothesis ............................................................................................................................ 5.4. Evidente that NGF is involved in sympathetic sprouting ................................................................................ 5.5. Non-NGF-like growth factors and sympathetic sprouting ...............................................................................

216 216 217 218 220 223

6. Possible functions of sympathohippocampal fibers ............................................................................................. 6.1. Electrophysiological and pharmacological studies ........................................................................................ 6.2. Metabolic studies ................................................................................................................................ 6.3. Behavioral studies ...............................................................................................................................

223 223 224 224

7. Conclusions and future directions ................................................................................................................... 7.1. Are sympathetic axons replacing septal terminals? ....................................................................................... 7.2. 1sNGFinvoIved in sympathetic sprouting? ................................................................................................ 7.3. 1ssympathetic sprouting functional? ......................................................................................................... 7.4. Possible clinica] significante ...................................................................................................................

225 225 226 226 226

Summary .....................................................................................................................................................

227

Acknowledgements References

........................................................................................................................................

...................................................................................................................................................

Correspondence:

K.A. Crutcher,

Department

of Anatomy,

227 227

University of Utah, School of Medicine, Salt Lake City, UT 84132,

U.S.A. 0165-0173/87/$03.50 @ 1987 Elsevier Science Pubhshers B.V. (Biomedica] Division)

204 1. INTRODUCTION

Less than 25 years ago, the concept of neuronal growth in the mature mammalian centra1 nervous system (CNS) was accepted

only by a few individu-

als. Today, the idea that such growth is possible has become widely accepted and is supported by the repeated demonstration that various forms of neuronal growth can and do occur, particularly in response to discrete injury of the brain or spina1 cord. The seminal studies in this field have been reviewed on numer-

most examples of neuronal plasticity have little, if any, functional significante and, in addition, are unlikely to be the result of natura1 selection. Thus, injury-induced

changes in the nervous

system presum-

ably reflect mechanisms which operate during development andlor the norma1 function of mature nervous tissue. In the following review a particular example of neuronal plasticity occurring in the adult mammalian CNS will be examined, namely, the growth of sympathetic axons into the brain and spina1 cord. There are

ous occasions’3,28Z5,53,190 and will not be repeated here. However, it seems safe to conclude that the

severa1 features of this growth response that provide unique opportunities to address some of the ques-

question of whether axonal and terminal growth occurs in the adult mammdian CNS has been answered affirmatively. Yet in spite of the documented exampies of sprouting and regeneration occurring in the CNS, the underlying mechanisms and functional significante of such growth remain, for the most part, unknown. For example, the poor regenerative potential of neurons intrinsic to the CNS could theoretically be due to interna1 limitations of the neurons’ growth potential or, alternatively, to external factors which prevent or fai1 to support such growth. These theoretical possibilities were apparent over a century ago, but it was only in the last 5 years that the appropriate experiments to distinguish between them could be undertaken. Aguayo’ showed quite clearly that neurons residing within the CNS could effectively regenerate for long distances if their axons were exposed to the appropriate environment, in this case a segment of periphera1 nerve. Thus, attention is now focused on factors in the environment of growing axons which permit or interfere with elongation or synaptogenesis. In spite of this theoretical advance, there are still many unanswered questions relating to the regulation and function of neuronal plasticity in the adult mammalian CNS. For example, what are the specific factors which permit, or hinder, axonal growth? Do separate factors account for specific examples of sprouting and regeneration? How does growtb in the mature CNS differ from that which occurs during development? TO what extent is neuronal plasticity functional? In a similar vein, to what extent does such plasticity reflect adaptive mechanisms selected through evolution? This latter question was recently addressed in these pages73 with the conclusion that

tions raised above. Furthermore, the combined use of severa1 different techniques has provided considerable information about the various factors that are involved in the elongation of sympathetic axons within the mature CNS. In order to draw meaningful conclusions regarding the theoretical significante of sympathetic invasion of the CNS, however, it is usefu1 to begin by briefly reviewing what is known about the norma1 function and relationship of sympathetic axons to their extracerebral targets. 2. SYMPATHETIC

INNERVATION

OF EXTRACERE-

BRAL TARGETS

2.1. ~~rnpat~eti~ targets out~i~e the ~ra~ium The autonomie nervous system is classically considered to be divided into sympathetic and parasympathetic components. The sympathetic division consists of a series of paravertebral and prevertebral chain ganglia which house the postganglionic neurons and which receive their preganglionic innervation from the thoracic and upper lumbar spina1 cord segments. In general, preganglionic sympathetic neurons use acetylcholine as their neurotransmitter and postganglionic neurons use norepinephrine (NE) allowing them to be visualized with histofluorescent techniques6’.62,71,‘3”. H owever, there are exceptions to this generalization. For instante, the postganglionic sympathetic innervation of sweat glands is mostly cholinergic 12’. A recent study provides evidente for heterogeneous populations of postganglionic neurons with respect to their neuropeptide content as we11139. Noradrenergic sympathetic fibers are found in virtually every peripheral organ, usually in association

205 with blood vessels, but also with other tissue e.g. salivary glands and cardiac muscle. Ultrastructural studies of sympathetic innervation of various peripheral targets have shown that sympathetic axons rarely form discrete synaptic contacts, i.e. regions of close membrane apposition or pre- and postsynaptic membrane thickening @. The reason for the absence of morphological s~cialization is unclear but is thought to reflect the diffuse modular nature of sympathetic innervation. Thus, the concept has emerged of an autonomic ‘ground plexus’ in which the sympathetic fibers form a more or less dense network of axons with swellings or varicosities representing the sites of vesicle accumulation and transmitter release and re-uptake”‘. As a result, ultrast~ctural features may not be sufficient in determining what cellular elements are actually ‘innervated’ by sympathetic fibers, a point that will be returned to below. The sympathetic innervation to most tissues involves some influente on smooth muscle, usually contraction, but may also influente other cells, including secretory ones. Sympathetic fibers may also have other roles such as regulating vascular permeability, influencing immune function and exerting long-term trophic effects on their targets7. This rather diverse range of functions underscores the difficulty in establishing a role for sympathetic axons when they appear in a new target, as in the sprouting response described below. 2.2. Intracranial sympathetic fibers Within the cranial cavity, sympathetic fibers are associated with structures outside the blood-brain barrier, e.g. pia1 blood vessels, choroid plexus and pineal gland”‘. Occasiona1 fibers may be found along penetrating arterioles, particularly in the primate CNS, but these are most probably limited to the Virchow-Robin space, i.e. the perivascular space formed between the basa1 lamina surrounding the penetrating blood vessel and the external limiting membrane of the brain and spina1 cord. Sympathetic fibers within the pineal gland and choroid plexus affect melatonin synthesis” and cerebrospinal fluid production’“, respectively. The extent to which intracranial sympathetic axons associate with intraparenchymal blood vessels appears to vary considerably across species2’~72*‘6’. Edvinsson and MacKenzie@ stated that ‘adrenergic

nerve plexuses that enclose the pia1 arteries accompany severa1 of the arteria1 branches which subsequently enter the brain parenchyma’. Lower mammals, however, seem to exhibit fewer intracerebral sympathetic fibers and in the rat, at least, their incidente is extremely low. Such fibers have been reported in the brainstem ‘lo but it is likely that most examples represent the association of the sympathetic fibers with large-diameter arterioles which are surrounded for some distante by a Virchow-Robin space. In the region of the hippocampal formation, which will be addressed in some detail below, such intraparenchymal fibers have never been reported and, in this laboratory at least, are never observed in norma1 rats. The conclusion that sympathetic fibers are normally present ‘within’ the brain parenchyma of some species68reflects certain assumptions about what constitutes’ the boundary between the peripheral (PNS) and centra1 nervous systems. This boundary has rarely received forma1 definition but most investigators accept some criterion, such as the absence or presente of the blood-brain barrier or of the pial-glia1 membrane to determine whether a particular location is, in fact, ‘within’ the PNS or CNS. This boundary is of more than just semantic importance when trying to determine whether axons are within the CNS, as we will see when we examine the sympathetic sprouting response. In the rat, the blood supply to the hippocampal formation36.37as well as the ultrastructural features of the sympathetic innervation of extracerebral blood vessels” have been examined in some detail. Sympathetic axons occur as fascicles of unmyelinated fibers surrounded by collagen and Schwann ce11processes. Individua1 fibers break away from such fascicles and are closely apposed to the basement membrane surrounding the smooth muscle layer. The closest apposition between nerve fiber and blood vessel will usually involve vesicle-laden profiles but rarely any evidente of membrane specialization. Since the vascular fibers are situated in the perivascular region (the subarachnoid space) they are separated from both the brain and the blood vessel by the basa1 lamina. Presumably, as is the case in the neuromuscular junction formed between spina1 motor neurons and striated muscle, sympathetic axons are able to influente vascular smooth muscle by releasing a transmit-

206 ter which acts across

the basement

membranelm.

Whether additional effects might be exerted across the external limiting membrane of the brain and spinal cord (the pial-glia1

boundary)

is not known.

The association of sympathetic fibers with parahippocampal blood vessels is shown schematically in Fig. 1. The longitudinal hippocampal arteries are innervated more densely at ventral hippocampal levels than at dorsal levels. In fact, the most rostral portions of the longitudinal

hippocampal

arteries do not nor-

mally have sympathetic fibers associated with them. The corresponding veins, not shown in Fig. 1, do not have sympathetic

fibers

normally

associated

traparenchymal vasculature. This accumulation appears to be due to a perivascular circulation of extracellular fluid. Sympathectomy prevents this perivascular accumulation. This finding may be relevant to of sympathetic

3. CENTRAL NORADRENERGIC

sprouting

\

with

them. The entire sympathetic innervation of parahippocampal blood vessels arises within, or passes through, the superior cervical ganglion. The function of the sympathetic innervation of extracerebral blood vessels has been surprisingly difficult to establish. Clear effects of sympathetic stimulation on cerebral blood flow are not observed except at extreme pressure@. Some effects on vascular permeability have been documented and it is also possible that sympathetic fibers exert a long-term trophic effect on vascular targets’. Of some interest is the fact that cerebral veins are normally sparsely innervated, if at all. Another feature of sympathetic innervation of extracerebral blood vessels is of some interest, namely, its influente on extracellular fluid circulation within the CNS. Rennels et al.“i have shown in recent elegant experiments that a protein tracer injected into the extracellular compartment (via the ventricular or subarachnoid spaces) will accumulate around the in-

the mechanism below.

septum

discussed

NEURONS

In addition to the peripheral noradrenergic neurons of the sympathetic nervous system, there are centra1 noradrenergic neurons which provide a widespread innervation to most regions of the CNS15* including the hippocampal formation (Fig. 2B). This represents a truly intrinsic innervation since the neu-

Fig. 1. Illustration of the major systems involved in the sprouting phenomenon described in this review. The hippocampal formation (hippo), seen here from a media1 perspective, is innervated by cholinergic and non-cholinergic neurons whose cell bodies reside in the septum and whose axons pass through the fimbria to enter the hippocampal formation. Sympathetic fibers, arising in the superior cervical ganglion (KG), accompany the longitudinal blood vessels which supply the hippocampal formation. The density of the sympathetic innervation to the vasculature decreases from tempora1 (ventral) to septal (rostral) levels. Also illustrated is the potential pathway by which septal neurons could influente preganglionic sympathetic neurons in the lateral horn of the spina1 cord which project to the SCG. The afferent input to the SCG can be removed by transecting the preganglionic nerve (arrow). The arrowheads indicate the approximate leve1 of the slices illustrated in Figs. 3 and 4.

rons reside wholly within the CNS. In spite of the few studies suggesting a role for centra1 noradrenergic neurons in the regulation of intracerebral blood vessels, however, there is no convincing direct evidente that such neurons innervate intraparenchymal vasculature. Of course the same limitation applies here as in the case of the sympathetics, namely, the absence of direct morphological contacts does not negate the possible influente of centra1 noradrenergic axons upon specific targets, including blood vessels. In fact, Descarries et a1.59@rhave argued that centra1 noradrenergic neurons only rarely form classica1 morphologica1 ‘synapses’ with neuronal targets although close membrane apposition is usually observed.

207

Fig. 2. A: Timm-st~ned section through the dorsal hippocampat formation of a rat illustrating the major subfields. The dentate granule cells form a distinct wedge-shaped layet within the dentate gyrus (dg) and give rise to axons (mossy fibers) which terminate in the hilus and CA, region of the hippocampus. The mossy fibers appear black due to their high zinc content which has been precipitated with sodium sulfide. The mossy fibers do not extend into the CA, region which begins at the point approximated by the arrow. B: fluorescence micrograph of the dentate gyrus illustrating the dense noradrenergic innervation of the hilus (H) which arises from the locus coeruleus. A few fibers are also present around the granule celi (GC) tayer. C: darkfield autoradiograph of the dentate gyrus following a [3H]leucine injection of the ipsiiateral superior cervical ganglion in an anima1 receiving a septal lesion. The labeled fibers in the hilus (H) and adjacent to the granute ce11(GC) fayer represent sympathetic fibers. D: fluorescente micrograph of the dentate gyrus from an anima1 that received a media1 septal lesion and was examined one year later.

Since centra1 and peripheral noradrenergic neurons both utilize NE they are both visualized with fluorescence histochemical techniques. Fortunately, it is usually possible to distinguish between both types of fibers due to their different fluorescente morphologies (Fig. 2). In addition, the developmental origin of the two types of noradrenergic neurons, ultimately resulting in their wide physical separation, allows for selettive removal or labeiing of one population or the other. This anatomica1 separation permitted the discovery that sympathetic noradrenergic fibers will invade the CNS foliowing specific lesions. The centra1 noradrenergic innervation to the rat hippocampal formation arises in the locus coeruleus and travels via 3 pathways, one of which passes

through the septum and fimbriaifornix system (Fig. 1). As a result, this inne~ation must be considered both in terms of its potential involvement in ehciting sympathetic sprouting after fimbrial lesions as well as its possible contribution to the recovery of noradrenergic innervation of the hippocampal formation. Each of these issues will be deah with later in this review. Another reason for considering centra1 noradrenergic neurons is their putative responsivity to nerve growth factor (NGF). This point will be returned to later in discussing the possible role of NGF in sympathetic sprouting but for now it is worth noting that sympathetic neurons, but not centra1 noradrenergic neurons, have been consistently shown to exhibit re-

208 sponsivity

to NGF. Locus coeruleus

ample, do not demonstrate port of NGF following

neurons,

for ex-

specific retrograde intrahippocampal

transinjec-

tionslsO.

assess its ability to innervate

Collateral neurons

4. SYMPATHETIC SPROUTING INTO THE CNS

Although

there

are many

examples

of neuronal

the iris. These pioneering

studies showed quite clearly the regenerative potential of both damaged in situ sympathetic axons as well as of grafted sympathetic ganglia. sprouting

of postganglionic

has been less commonly

sympathetic

studied. There have

been severa1 studies of such sprouting on the part of preganglionic sympathetic axonsl*’ but these do not

sprouting within both the PNS and CNS3s,53, the rear-

shed light on the growth potential

rangement presented here represents a rare case of intact peripheral neurons growing into the CNS fol-

adrenergic neurons. There is some evidente such collateral sprouting will occur in peripheral

lowing distant injury.

gets, particularly

This remarkable

response

pro-

vides a unique opportunity to identify growing axons within the adult CNS and to study the mechanisms and function of such sprouting. In addition, the specificity of invading sympathetic fibers for CNS regions that have been denervated of putative cholinergic fibers presents an intriguing example of possible transmitter substitution as well as indicating a role for a brain-derived NGF-like growth factor in neuronal sprouting. Each of these issues is dealt with in the following pages. 4.1. Growth potential of mature sympathetic neurons There are two forms of axonal growth that are commonly observed in the mature nervous system. Regeneration is the regrowth of damaged axons and is quite common in the PNS but occurs only rarely in the CNS. Collateral sprouting is the growth of uninjured axons, usually in response to loca1 (denervating) changes in their target. A third form of axonal growth may occur but is difficult to document . This is referred to as the ‘pruning’ effect and involves growth of collateral axons from one region of a neuron’s axonal arbor in response to injury of other axonal branches of the same neuron. This presumably represents a growth response similar to that underlying regeneration, i.e. the signal initiating growth may arise as a result of direct injury to the neuron. The regenerative potential of sympathetic neurons was firmly established by Langley”’ who demonstrated that postganglionic sympathetic axons would regrow following nerve crush and re-establish the degenerated connections. The regenerated connections were functional. Much later, Olson and Malmforslti utilized the regenerative growth potential of sympathetic neurons to successfully transplant superior cervical ganglia into the anterior chamber of the eye and

of peripheral

in response to denervation

ry or parasympthetic

fibers. For example,

northat tar-

of sensoKessler”’

has recently shown that removal of any one of the 3 innervations to the iris, i.e. sensory, sympathetic or parasympathetic, leads to sprouting of the remaining two. Bjorklund et al.” reported that sympathectomy leads to the appearance of neuropeptide Y and tyrosine hydroxylase activity in parasympathetic fibers innervating the iris. Another example of collateral sprouting of postganglionic sympathetic axons apparently occurs following unilateral denervation of the pineal gland. Dornay et al. 63 found that the remaining contralateral sympathetic fibers exhibited biochemical changes which they interpreted as representing compensatory collateral sprouting. Interestingly, they found that the biochemical changes were prevented by decentralization of the superior cervical ganglion. These examples document the ability of mature postganglionic sympathetic neurons to expand or alter their axonal arbor and may provide information relevant to the sympathetic sprouting response occurring in the hippocampal formation. 4.2. Discovery of sympathohippocampal sprouting Bjorklund and Stenevi performed a series of experiments to test the ability of various neuronal implants to innervate the hippocampal formation (HF)“. In addition to embryonic neural tissue from the CNS, they implanted adult superior cervical ganglia (SCG) into a fimbrial cavity and followed the course of regeneration with histofluorescence techniques. They discovered that the implanted sympathetic ganglia were able to establish a dense innervation of the HF. Of particular interest was the observation that the distribution of sympathetic fibers was similar to the distribution of norma1 centra1 norad-

209 renergic

(10~s

(cholinergic) through

coeruleus)

projections;

the fimbria

and

septohippocampal

projections

which

and which were interrupted

the implantation procedure. The relationship septal region and fimbria to the hippocampal tion is shown schematically in Fig. 1. At about the same time that Bjorklund

growth of sympathetic

pass by

of the forma-

and Stenevi

were using the transplantation procedure to assess growth potential, Loy and Moore’33 were studying the centra1 noradrenergic innervation of the rat HF and in the process interrupted the fimbria with a suction lesion. They reported that sympathetic fibers were found within the HF following damage to its anterior pole. They confirmed that the bright fluorescent axons in the HF originated in the periphery since the fibers were eliminated by bilatera1 superior cervital ganglionectomy. They also ruled out any contribution of centra1 noradrenergic neurons to the anomalous innervation of the HF by destroying the locus coeruleus, the brainstem nucleus which gives rise to the centra1 noradrenergic input. Since their hippocampa1 lesion involved damage to extracerebral blood vessels, and associated axons, they suggested that the sympathetic fibers were regenerating in response to direct damage of the vasculature at the anterior pole of the HF. Stenevi and Bjorklundlg9 subsequently reported seeing sympathetic fibers in the HF following fimbrial lesions, or, more surprisingly, destruction of the media1 septum. They hypothesized that the invasion of the HF by sympathetic fibers might be related to the loss of septohippocampal fibers, although they did not test this possibility directly. They also suggested that sympathetic sprouting might interfere with the interpretation of brain lesion studies not only as a source of confusion with centra1 regenerative events but also by providing possible functional influences. The histofluorescent appearance of sympathetic fibers in the hippocampal formation after a media1 septal lesion is shown in Fig. 2D. Fig. 2C shows the results of injecting [3H]leucine into the SCG after a media1 septal lesion and processing the HF for autoradiography. Such results independently confirm the origin of the anomalous sympathetic axons. 4.3. Specificity of sympathohippocampal sprouting The fact that media1 septal lesions elicit the in-

sents a collateral

axons indicates

sprouting

response

that this repreand not regener-

ation of damaged axons. The importance of this distinction lies in the fact that the initiation of axonal growth must come from a stimulus external to the sympathetic neurons. That sympathetic axons could be induced to grow into the hippocampal formation by a distant lesion led Crutcher et al,38,41 to ask what aspect of the media1 septal lesion elicited sympathetic sprouting. They found that sprouting of sympathetic fibers into the HF occurred specifically in response to septohippocampal (presumably cholinergic) denervation. Their conclusion was based on 3 lines of evidente. First, only interruption of septohippocampal axons would elicit the response; destruction of other hippocampal afferent fibers (including locus coeruleus lesions) did not elicit or alter sympathetic sprouting. Second, the distribution of sympathohippocampal fibers was similar to, although not exactly the same as, the topography of the septohippocampal projection as visualized with anterograde transport of horseradish peroxidase from the septum. Third, partial septal lesions, which resulted in partial septohippocampal denervation of the HF, resulted in topographically matched and regionally restricted sympathetic sprouting. By using adjacent sections stained for acetylcholinesterase (AChE) activity and NE histofluorescence, it was possible to show that hippocampal regions that had been denervated of AChE staining were the same regions occupied by sympathetic fibers41. Loy et al. 135undertook a detailed study of the sympathetic sprouting response which also provided support for its specificity. They showed that lateral fimbrial lesions resulted in sprouting mainly within the ventral HF and that media1 fimbrial lesions resulted in sprouting mainly within the dorsal HF. Since this corresponds to the general topography of the septohippocampal projection, their results also supported the conclusion that septal denervation was involved. They also demonstrated that lesions of other afferent pathways, e.g. commissural or entorhinal denervation, did not elicit or enhance sympathetic sprouting. Bjorklund and Stenevit4, again using the transplant paradigm, also studied the specificity of this ingrowth. Sympathetic ganglia transplanted to the region of the HF, without damaging the septohippocampa1 projection, resulted in restricted and sparse

210 innervation

of the HF. Ganglionic

transplantation

transplanted

SCG will grow into the rat HF following

followed by fimbrial transection or media1 septal lesions, however, resulted in massive ingro~h from the transplants. Although the innervation estab-

lesions that interrupt the septohippocampal projection but not follo~ng damage to other hippocamp~ afferents. This evidente has been cited as supporting

lished in the transplant paradigm represented regenerative growth on the part of cells which had been

the conclusion that sympathohippocampal sprouting exhibits lesion or denervation specificity, that is, it only occurs following a specific lesion4’. There have,

axotomized

as a consequence

of ganglion

removal,

the receptivity of the hippocampal formation to regenerating sympathetic fibers was clearly improved by septohippocampal denervation. It should be noted that unti1 recently the septohippocampal pathway was thought to be primarily, if not exclusively, cholinergic in nature. As a result, ali of the studies cited above were interpreted as indi~ating that it was cholinergic denervation per se that elicited sympathetic sprouting4”. Although there is still some reason to accept this hypothesis (see below) it is now clear that a large percentage, if not, in fact, the majority, of septohippo~ampal neurons are non-cholinergic ‘9~. Some may use glutamic acid decarboxylase (GAD) as a neurotransmitter and others may use yet unidentified transmitters. As a result, it is possible that septal and sympathetic fibers could contain a common neurotransmitter or ‘neuromodulator’. For instance, severa1 neuropeptides have been identified . Some of these peptides in sympathetic ganglia 139 might also be present in septal neurons. Alternatively, the ability of sympathetic neurons to express cholinergic activity in vivo’28 and in vitro”j9 suggests the possibility that some s~pathohìppocampal fibers may express cholinergic activity. Such cholinergic expression might be small compared to the norma1 noradrenergic activity but could be functionally relevant. These possibilities emphasize the difficulty in establishing whether sprouted sympathetic fibers are functional in terms of electrophysiological or biochemical transmission. Whether sympathetic sprouting is elicited by ‘cholinergic’ denervation per se, or some other aspects of septohippocampa1 fiber removal. remains an unanswered question. Although much of the evidente is consistent with the hypothesis that chohnergic denervation is involved, for the time being it is more appropriate to identify ‘septohippocampal’ denervation as the necessary condition for eliciting sympathohippocampal sprouting. Collectively, then, these studies provide strong evidente that sympathetic fibers from in situ and

however,

also been sug~estions

that other hippocam-

pal damage, such as occurs following kainic acid injections into the hippocampal formation may elicit sympathetic sprouting93p117. Such results need to be confirmed with sympathectomy experiments and if confirmed

it will be necessary

to determine

whether

this might reflect a response to damage of septal fibers in the region of the injection. As discussed in a recent paper’l, examples of neuronal sprouting involve changes in the tissue which elicit the growth of uninjured axons. If only certain lesions or dene~ations elicit the sprouting then the response can be said to exhibit lesion or denervation specificity. However, the extent to which different populations of afferent fibers respond to a particular lesion may indicate the presente of response specificity as well. In the example of sympathetic sprouting considered here the question of response specificity was addressed by determining whether non-sympathetic perivascular axons sprout in response to septohippocampal denervation”‘. That such fibers might be expected to sprout was suggested by their close association with perivascular sympathetic fibers which were found to increase their perivascular innervation at the same time as they invade the denervated HF. However, sprouting on the part of non-sympathetic perivascular axons was not observed. The fact that the non-sympathetic fibers did not sprout suggests that sympathetic sprouting may also exhibit response specificity, i.e. only certain fibers in the area of denervation respond to the lesion. The suggestion has been made that locus coeruleus fibers regenerate faster in response to neurotoxin injections if simultaneous septohippocampal denervation is present’4. However, this conclusion is tempered by the fact that septal lesions or fimbrial transections interrupt some locus coeruleus fibers and may therefore stimulate growth of remaining fibers through a ‘pruning’ effect. Support for this conclusion comes from the observation that NE levels in the HF which arise in the locus coeruleus never exceed

TABLE 1 Specifcity of sympathohippocampalsprouting Lesion (denervation) specificity (1) Septal denervation is necessary and sufficient (2)

lesions (3) Loss of granule cells, but not CA, pyramidal cells, prevents sprouting

the norma1 levels provided by this centra1 innervation’43. It seems, therefore, that locus coeruleus fibers do not sprout in response to septohippocampal denervation. Of course, many other hippocampal inputs could potentially respond to a media1 septal lesion with a sprouting response that is hard to detect, in which case the sympathetic sprouting response would not be specific. The evidente for specificity in the sympathetic sprouting response is summarized in Table 1. The extent to which sympathohippocampal sprouting exhibits lesion and/or response specificity indicates that a discrete signal is involved. A good candidate for such a signal has been identified and will be discussed in detail below. 4.4. Topography of sympathohippocampal sprouting In addition to the preceding evidente for lesion and response specificity of sympathetic sprouting, the restricted distribution of sympathohippocampal axons also indicates that there are constraints on the regional extent of such growth or, in other words, that there is topographic specificity. The distribution of septohippocampal fibers and terminals in the rat is shown schematically in Fig. 3. The densest septal input is to the dentate gyrus and CA, region of the hippocampus. A sparser innervation is present in the CA, region. Sympathetic fibers are normally confined to extracerebral arteries (curved arrows, Fig. 3). Following fimbrial transection or a media1 septal

Fig. 3. Schematic illustrating the major ce11 populations in a transverse slice of the dorsal hippocampal formation. The stippling indicates the distribution of septal fibers and terminals which cnter the region by way of the fimbria (fim). The densest innervation is to the dentate gyrus and CA, of the hippocampus. The arrow indicates the endpoint of the mossy fiber layer which terminates at the transition from CA, to CA,. Sympathetic fibers are restricted to the parahippocampal blood vessels (curved arrows).

lesion, sympathetic fibers appear only within the dentate gyrus and CA3 region of the hippocampus (Fig. 4). They do not extend beyond the transition zone (CA,) into CA, (arrow, Fig. 4). The majority of

Fig. 4. Illustration of a section comparable to that in Fig. 3 following septohippocampal denervation. Sympathetic fibers invade the dentate gyrus and CA, region of the hippocampus terminating at the CA,-CA, transition zone (arrow). The sympathetic axons enter along penetrating blood vessels but do not ramify within the dentate molecular Iayer, extending, rather into the hilus and around the major ce11layers in the dentateCA, region.

212 are also within the tate gyrus the area and below granule ce11 Sympathetic axons in association bIood vessels layer but this

ra,

the dentate

not ramify the outer even in presente of simultaneous lesion which

denervates

of en-

this

that sympathetic

species as and the rat except sympathetic the

ramify to dentate gyrus

CA, and pyramidal ce11 the stratum and, in

occurs in is similar fibers

These variations response

gion32. Sympat~etic hilus of

CA, of guinea pig, area that have a cholinergic input in the Recent results studies in ferret, a of the Carnivo-

in

the sympathetic species may

extent within course along

ctues concerning

particularly withcases, within

4.5. Sympathetìc sprouting in other brain regions

stratum oriens. there are few fibers in the radiatum or molecuiare of hippocampus. One the most features of fiber topography, is the termination of distribution at transition zone CA,. This a very and dramatic in the (although see below on studies by and Davis the guinea Although this of sympathetic in the is generally Loy and and Milner Loy 148~149 reported some tions between and female In particular, found that rats had less sproutand a restricted distribution the sympafibers than This apparent dimorphism be returned in the of horfactors that affect the response (section Another feature this sprouting is that sympathetic innervation parahippocampal blood increases as This is pronounced the rostral of the where sympafibers are normally present. addition, following denervation vascular fibers in association parahippocampal veins though they not normally such vessels5’. of the described above obtained from of the HF. Sympathetic has also observed in guinea pig” in the ret HF in preparation). results in guinea pig somewhat surprising. topography of in the area and region is similar to observed in rat but, addition, axons were to appear

regulation

that in within

of

ingrowth.

If sprouting of sympathetic fibers is specifically due to loss of cholinergic septohippocampai axons, it is possible that cholinergic denervation of other brain regions will aIso result in such ingrowth. Following septal lesions, Crutcher and Davis4’ observed sympathetic fibers in the media1 habenula of the rat. They suggested that sympathetic sprouting occurred in response to loss of septohabenular fibers, some of which are probably cholinergic”. Gottesfeld9 reported an increase in media1 habenular fhtorescence and NE content following lesions of the stria medullaris that she attributed to sprouting of both centra1 and peripheral noradrenergic neurons. Since some investigators have identified a norma1 ~ontribution of sympathetic fibers to the media1 habenula9~203, it is not clear, however, whether the increase in media1 habenular fluorescente reflects a a de novo invasion of this region by sympathetic axons or an increase in a normally sparse inne~ation. Chafetz and Gagezs found that sympathetic sprouting occurs in the medial habenula and in the pineal gland following medial septal lesions and suggested that such results reflect a common ce11body response on the part of those sympathetic neurons which sprout into the HF. TO test the generality of the hypothesis that cholinergic denervation of a brain region leads to sympathetic sprouting it is necessary to know a major source of cholinergic fibers to the region in question. A major extrinsic cholinergic projection to the neocortex arising in the basa1 forebrain1~,1’4,1~ provides a system in which to test this hypothesis. As in the studies of sprouting in the HF, Crutcher* monitored the loss of cortical cholinergic fibers with AChE histochemistry. Adjacent sections were stained for histofhtorescence to determine whether sympathetic sprouting had occurred. In agreement with the hypothesis, sympathetic fibers were observed in neo-

213 cortical regions denervated

of AChE-positive

fibers.

cidating the sources of cholinergic

fibers. The diffuse

The peripheral origin of the fluorescent fibers was confirmed by their disappearance ipsilateral to a su-

group of large AChE-positive

perior cervical ganglionectomy.

the neocortex, may be the source of cholinergic fibers to other brain regions such as the amygdala. Unlike the situation in the HF, olfactory bulb and neo-

In addition,

destruc-

tion of other neocortical afferent fibers, e.g. centra1 noradrenergic fibers arising in the locus coeruleus, did not lead to sympathetic sprouting. One exception to this ‘cholinergic denervation’

basa1 forebrain,

cortex, hy-

however,

which provides

AChE

perikarya cholinergic

histochemistry

within

the

fibers to

cannot

al-

pothesis was found in the olfactory bulb which receiv-

ways be relied on as an indicator of the integrity of the cholinergic fibers in other brain regions. It may,

es cholinergic innervation from the nucleus of the horizontal limb of the diagonal band. Again, AChE his-

therefore, be difficult to obtain convincing histochemica1 evidente of the specificity of sympathetic

tochemistry

sprouting

was used to monitor

the loss of choliner-

gic fibers in the bulb following electrolytic lesions of the horizontal limb nucleus. Surprisingly, lesions which resulted in extensive AChE depletion did not elicit detectable sympathetic sprouting89. In order to contro1 for the possibility that sympathetic axons did not have access to the denervated tissue (the sympathetic innervation to blood vessels in that region is quite sparse), SCG were transplanted to the dorsal surface of the main olfactory bulb, again without any evidente of sympathetic innervation. This finding provides an interesting exception to the other studies in which cholinergic denervation does lead to sympathetic sprouting. The reasons for this are not clear but may ultimately shed light on the mechanism underilying sympathetic sprouting in other brain regions including the HF. The capacity for sympathetic sprouting in brain regions other than those already noted has not been addressed in light of the cholinergic denervation hypothesis, but there are reports of sympathetic fibers invading the CNS following certain injuries. Among these are the presente of sympathetic fibers in the cingulate cortex following damage to the cingulum’89 and the presente of sympathetic fibers in the spina1 cord caudal to a transection’45. In the latter study there was, unfortunately, no attempt to confirm the peripheral origin of the fibers with sympathectomy. The possibility remains, however, that the spina1 cord is also subject to this autonomie invasion following injury. If sympathetic innervation of the transected spina1 cord can be documented, it would be interesting to determine whether loss of a descending cholinergic pathway accounts for this response. Whether or not the specificity of sympathetic sprouting for cholinergic-denervated cortex can be generalized to other brain regions will depend on elu-

for other cholinergic-denervated

brain re-

gions. If, on the other hand, their cholinergic identity is not an essential attribute of the fibers whose loss elicits sympathetic sprouting then non-cholinergic systems might be examined as possible candidates for eliciting sympathetic sprouting in other brain areas. 4.6.

and hormone

on sympathetic

Some studies addressed possible influences on sprouting. Scheff a1.176 reported sympathetic ingrowth fimbrial was dramatically in aged compared to ones. Since levels increase age, DeKosky aI.” suggested elevated corticosteroid may account examples reduced sprouting in aged Scheff et reported preliminary indicating exogenous corticosterone in fact, sympathetic sprouting. in sympathetic in aged has recently quantified and by Booze al. 18. effect of hormones has suggested the work Loy and and Milnand Loy 148~149 . As mentioned previously, they reported that adult male rats exhibited significantly less sprouting compared to adult female rats and that the pattern of sympathetic sprouting in males was less extensive than that in females. In particular, sympathetic fibers were typically present within the supragranular region of the dentate gyrus and in the straturn oriens of CA3 in female rats but only rarely in males. TO examine possible influences of sex steroids on sympathetic sprouting they studied norma1 male and female rats ranging in age from 3 to 120 days. They also examined the effect of castration and steroid hormonal administration. Both neonata1

214 males and females exhibited the same pattern and amount of sprouting as adult femaies, i.e. there was no obvious sexuai dimorphism. They also found that

4.7. Ultrastructural relationships of septal and sympathehc fibers to hippocampal elements There are two questions relating to the sympathet-

castration

of aduit animals

but that sprouting

castration equivalent

ic sprouting response that can only be answered with ultrastructural anaiysis. The first is the identification

neonata1

testosterone

males or intact

did not affect sprouting

of neonata1 males permitted to that in adult females. Also,

females

administration reduced

to castrated

sprouting

to that

seen in intact males. Thus, it appears that sympathetic sprouting is regulated, in part, by sex steroid hormones but that this influente occurs sometime during development. Loy and Milnert3* have suggested that there is sexual dimorphism in the development of AChE-positive cells in the hiius which might account for this organizational differente in sympathetic sprouting foilowing fimbrial lesions. A study by Davis and Martin5’ss7 suggests that thyroid hormones may also affect sympathetic sprouting, again indirectly. Neonata1 thyroxin treatment appeared to result in more extensive sympathetic ingrowth in the stratum oriens of CA, compared to saline-treated controls. Since thyroxin treatment results in a greater density of granule celi axons (mossy fibers) in this region, the authors suggest that these results support an earlier conclusion that sympathetic sprouting is reiated to the distribution of mossy fibers46. These hormonal studies provide interesting possibilities concerning the regulation of sprouting. However, it should be pointed out that one serious limitation in interpreting the results of each of these studies is the lack of quantification of the amount of sympathetic sprouting. For example, the pattern of sprouting in adult male rats reported by Crutcher et aL4t is almost identica1 to that reported by Loy and Miiner for adult females”36.‘48. It is possible, for example, that there are differences in the rate of sprouting foliowing hormone treatments or that there are strain differences in the amount or topography of sprouting. Without rigorous quantitative methods it is difficult to firmly establish the validity of the conclusions regarding age and sex differences in this sprouting model. The use of quantitative methods such as those described by Chafetz et a1.22.23and by Booze et al.17 may provide the opportunity to confirm the regulatory effects of different hormones on sympathetic sprouting.

of the norma1 targets of those septohippocampal bers whose ioss elicits the sprouting response. second

is whether

sympathetic

fibers

that

fiThe have

sprouted into the CNS form synapses and, if SO, what cellular eiements are contacted by them. Since the sprouting of sympathetic fibers appears to depend on the loss of septohippocampal

neurons,

the norma1 targets of septal fibers may become

tar-

gets of sympatheti~ fibers. The unequivocai identification of neuronal targets is difficult, however, particularly since axons can influente cells in the absente of morphological synapses, as appears to be the case for sympathetic innervation of peripheral tissues (see section 2.1.). Non-synaptic influences can only be assessed with biochemicai or electrophysiologica1 methods. A deeper confusion arises from the use of the term ‘target’. The cells which are influenced by a particular innervation may be considered as the non-synaptic or synaptic physiological targets of that innervation but there also may be cells which play a role in eliciting axonal growth but which do not respond to signals generated by the afferent fibers. Such cells may be considered as tropic or trophic targets of growing axons even though they do not enter into discrete morphological relationships with the innervating axons. The question of ‘synaptic’ targets will be addressed here and the possible existence of trophic targets will be returned to in the section on the mechanism of sympathetic sprouting (section 5.2.). The identification of septohippocampai fibers has been accomplished with severa1 light microscopie techniques (see ref. 42 for a review), but ultrastructu-. ral studies have, unti1 recently, been rare. Since sympathetic fibers normaily innervate extracerebral vascuiature, and may continue to be associated with intrahippocampal blood vessels, septal fibers might also be expected to innervate intraparenchymal hippocampal vasculature. Chandler and Crutcher2h analyzed the septohippocampai projection using anterograde transport of horseradish peroxidase at the electron microscopie level. Previous light microscopie studies demon-

215 strated the feasibility

of using this technique

alize the septohippocampal

projection4*.

to visu-

Since it is

regions of the axon will generally this approach.

be overlooked

with

not possible to completely rule out a possible contribution of fibers of passage to the labeled profiles in

Ideally, then, it would be useful to have a marker that will unequivocally labe1 sympathetic fibers over

the HF, the precise targets of septal fibers cannot be unequivocally identified. However, they found that al1 of the labeled profiles exhibiting synaptic speciali-

their entire length. Such labeling is feasible with anterograde markers such as horseradish peroxidase (HRP). In a recent study, sympathetic fibers in the

zations targets.

dentate gyrus were identified using anterograde transport of wheat germ agglutinin-conjugated HRP

were exclusively associated with neuronal There was no indication of synaptic associa-

tion with vascular

or glia1 elements.

Recent

studies

from other laboratories have provided similar results27,78,79(although the association of a choline acetyltransferase

(ChAT)-positive

profile with a paren-

chymal vessel in the amygdala has been reported in one study3). As a result, in order for sympathohippocampa1 fibers to synaptically ‘replace’ septal fibers, they presumably must ‘innervate’ neuronal targets. The ultrastructural identification of sympathohippocampal fibers has been surprisingly difficult to accomplish. Loy and Moore’34 referred to preliminary results obtained using 5-hydroxy-dopamine (5OH-DA) as a marker for noradrenergic varicosities. Although micrographs were not presented, the authors reported that labeled profiles were present around vascular elements as well as within the neuropi1 where synaptic contacts with neurons were seen. McGinty et al. 144undertook light microscopie studies in which the blood vessels and sympathetic fibers were visualized in the same tissue section. They also reported that sympathetic fibers left the vasculature and ramified within the hippocampal tissue. Although these studies indicate that sympathetic fibers enter the hippocampal neuropil and form synapses with neuronal targets, there are technical limitations that prevent outright acceptance of these conclusions. The light microscopie data are difficult to evaluate since the resolution is limited and it is not clear if the vascular bed was completely labeled. As far as the ultrastructural data are concerned, one of the limitations of using a false transmitter marker, such as 5-OH-DA, is that al1 noradrenergic, and possibly some non-adrenergic, terminals will be labeled. As a result, the only way to be sure that sympathetic terminals are exclusively labeled is to remove al1 other centra1 noradrenergic terminals. Such manipulations may, in turn, have an effect on sympathetic fibers. Another limitation when using 5-OH-DA is that only the terminals will be labeled. Non-terminal

following injections into the SCGs5. Taken together with earlier studies in which standard electron microscopic procedures were used to study the blood vessels penetrating the dentate molecular layer following sympathetic sprouting5’, some preliminary conclusions can now be drawn. Normally, sympathetic fibers are restricted to perivascular fascicles associated with extracerebral arteries supplying the HF. These arteries give rise to arterioles which penetrate the dentate molecular layer and, when viewed with the electron microscope, are seen to consist of endothelial and smooth muscle cells separated from the adjacent neuropil by a layer of the basa1 lamina. Following media1 septal lesions, and subsequent sympathetic sprouting, some of the penetrating blood vessels are accompanied by perivascular bundles of axons which lie outside the vascular basa1 lamina but are separated by an additional basa1 lamina from the neuropil”. This duplication of the basa1 lamina is also observed around some capillaries in the dentate hilus region. With the use of HRP as an ultrastructural marker it will be possible to extend these observations to areas where no clear morphological marker exists, such as the basa1 lamina duplication, with which to identify the sympathetic fibers. SO far, most of the HRPlabeled profiles that have been found are still closely associated with blood vessels even though no basa1 lamina separates them from the neuropils5. However, some labeled profiles have been found within the neuropil. Clearly further work is needed before definitive conclusions can be drawn about the synaptic relationships, if any, of sympathetic fibers with specific hippocampal targets. The close association of sprouting sympathetic fibers with blood vessels is not surprising in light of their norma1 association with extracerebral vascular targets. Their predominant association with vascular elements, even after entering the CNS, however. is

216 somewhat surprising since the suspected signal giving rise to the sprouting is thought to arise from hippocampa1 cells and not from the blood vessels (see be-

rons, such as a change in afferent arising in the denervated HF.

low). Two possibilities are worth considering. One is that growing sympathetic axons require the kind of

5.1. Role of afferent input

substrate provided by the basa1 lamina. This is consistent with what is known about the role of matrix factors such as laminin in supporting neurite outgrowth52~‘8’. In fact, laminin is only found associated with blood vessels in the mature CNS. A second possibility relates to the studies of Rennels et al. “’ cited above. If a diffusible factor arises from the HF and is released into the extracellular fluid then it might accumulate within the perivascular region, as shown for tracer proteins such as HRP. In either case, the perivascular disposition of sympathetic sprouts may be providing clues concerning the constraints on axonal growth in the mature mammalian CNS. 5. SPROUTING

MECHANISMS

It is likely that no single mechanism can account for such a complex biologica1 phenomenon as neuronal sprouting, but it is possible that certain factors assume more importance than others in the initiation, guidante, and maintenance of such growth. Investigators of other examples of neuronal plasticity, including peripheral nerve regeneration, have offered severa1 hypotheses to account for those phenomena. These include direct effects of injury (including the pruning hypothesis), changes in afferent input, loss of inhibitory factors, production of centra1 or peripheral growth factors, and possible interactions between al1 of these (for a review see ref. 119). In the example of neuronal sprouting considered here, it has been possible to obtain evidente in support of some of these mechanisms and against others. In the origina1 implant studies, Bjorklund and Stenevi suggested that the invasion of the HF by sympathetic fibers was under the contro1 of factors released by denervation ” . Loy and Moore, on the other hand, suggested that sympathetic neurons were aberrantly regencrating into the denervated HF following direct injury to their axonsi34*135. It is now reasonably certain that sympathetic fibers do not need to be damaged in order to be able to invade the CNS SOthat attention can be focused on those mechanisms which involve some change extrinsic to the sprouting neu-

input or a signal

Because the entire afferent neural input to the SCG can be eliminated, Crutcher et al.38 were able to determine whether such input was necessary for sympathetic neurons to sprout in response to a media1 septal lesion.

A possible

pathway

by which septal

neurons could influente sympathetic neurons is shown in Fig. 1. Descending fibers from the septum could alter the activity of preganglionic sympathetic neurons in the lateral horn of the spina1 cord via hypothalamospinal pathways. Transecting the preganglionic trunk to the SCG (arrow, Fig. 1) before placing the septal lesion prevents any preganglionic input from influencing the sprouting response. Crutcher et al.38 observed that sprouting still occurred in the absente of preganglionic input (also known as decentralization) and concluded that the initiation of sprouting was independent of changes in the firing rate of preganglionic fibers. Although afferent input was not necessary for sprouting to occur, the results of the study by Crutcher et al. suggested that sprouting may have been reduced on the decentralized side38. Since quantitative comparisons were not made, however, this conclusion was not definitive. Subsequent studies by Madison and Davis142,‘43 extended these earlier observations. They showed that there is indeed less NE from the SCG in the HF of decentralized animals, but uptake of exogenous NE in these animals is not significantly different from that in which the preganglionic input to the SCG is intact. They also observed a significant increase in NE content of the HF when the SCG was reinnervated by preganglionic fibers. More recently, Dornay et al.63 have shown that decentralization prevents the biochemical changes (homotypic collateral sprouting, according to Dornay et al.) measured in the pineal gland following unilateral sympathectomy. In other words, afferent input appears to be required for the biochemical changes which affect transmitter production but is not required for the initiation and maintenance of axonal growth. The apparent discrepancy between the effect of decentralization on homotypic collateral sprouting in the pineal gland and the ‘heterotypic’ sprouting response in the HF may be more apparent

217 than real. In fact, there is very little evidente nal growth in the study by Dornay

for axo-

et al.63 and the re-

sults could be interpreted as reflecting increased numbers of terminals and increased NE content within the sympathetic

fibers rather than axonal growth.

5.2. Target regulation The results of the decentralization

because the septal innervation

led

to the suggestion that a specific mechanism, involving the presente of an NGF-like growth factor, accounted for sympathetic sprouting3s (see section 5.3.), yet other possibilities should be considered. For example, sympathetic fibers might normally be actively inhibited from sprouting by the release of a growth-inhibiting substance from septal fibers. Altematively, a degeneration product released by damaged septal axons or the increased availability of synaptic or transmitter receptor sites could permit growth. Finally, a loca1 breakdown in the bloodbrain barrier could release growth-promoting substances into the HF. Each of these possibilities has been addressed experimentally to some degree. Although it is difficult to completely eliminate the various mechanisms listed above, Crutcher and Davis43 addressed the possibility that loss of septal fibers per se elicited sympathetic ingrowth. Normally the media1 septum provides a sparse innervation to the dentate molecular layer (Fig. 3) and the entorhinal cortex provides a massive innervation to this layer. Following lesions of the entorhinal afferent fibers, a dense band of septal fibers appears in the denervated portion of the molecular layer34~‘40~‘56~158~188~1~. By performing sequential entorhinal and septal lesions it was possible to test whether sympathetic fibers would proliferate within the zone previously occupied by sprouted septal fibers. The fact that they did not led to the conclusion that septal fiber degeneration per se was not responsible for the initiation or topography of sympathetic sprouting. The results of the sequential lesion experiment also suggested that the sympathetic fibers were not normally simply inhibited from invading the HF by a factor released by septal fibers since the loss of inhibition should permit sympathetic sprouting in the dentate molecular layer as well. Further experimental evidente indicating that loss of septal fibers alone does not account for sympathetic sprouting is found in the observation that neonata1 septal lesions, which produce very litile denervation

still

tration of a factor, or factors, that would induce the sympathetic axons to invade the CNS. In fact, Kiernan has proposed

experiment

is just developing,

elicits tremendous sympathetic ingrowth47. Another possibility is that a regional change in the blood-brain barrier (BBB) could allow for the infil-

that axonal elongation

can only oc-

cur in the presente of serum proteins that are normally excluded by the BBB ng . In order to test for possible changes in the BBB following septal lesions, Madison et al.141 studied the extravasation of systemically injected HRP or Evans blue at various times following a septal lesion. At no time was there evidente of extravasation of the systemic markers beyond what was observed in norma1 animals. Although it is still possible that there is a selettive increase in vascular permeability for a factor that induces sprouting, there is no general breakdown in the BBB. Changes in circulating factors, such as hormones, could affect the perikarya in the SCG and thereby influente sprouting. However, such factors cannot be solely responsible for the directed growth response, since the sympathetic fibers only sprout ipsilateral to the denervation and only into regions denervated of septal fibers 41. As a result, norma1 circulating systemic factors that are required for axonal growth, such as oxygen, glucose, amino acids, hormones, etc., cannot account for the topographical specificity of this sprouting response. It would seem then, that the initiation of sympathetic sprouting must stem from a change in the denervated HF which is transmitted to extracerebral perivascular sympathetic axons. TO test the possibility that specific hippocampal cells are sources of the sprouting signal arising after septal denervation, i.e. trophic targets, Crutcher and Davis eliminated entire ce11 populations in the HF before making a media1 septal lesion. Granule cells, the principal ce11 type in the dentate gyrus, were destroyed with colchicine in one group of animals and the pyramidal cells in CA3 were eliminated by injecting kainic acid in another group. They found that loss of granule cells, but not pyramidal cells, prevented sympathetic sprouting in response to media1 septal lesions. The close association of sympathetic fibers with the granule ce11axons (mossy fibers) led them to conclude that the dentate granule cells were a likely

218 source of the sprouting signal. This hypothesis was strengthened the thyroxin

cells. by the results of

study of Davis and Martin56,57 cited ear-

lier in which a greater density of mossy fibers in the stratum oriens of CA, correlated with more extensive sympathetic

sprouting

in this same region com-

pared to untreated animals. Interpretation results is of course limited by the ability quantitative

comparisons

between

of these to make

animals.

Another experimental approach to this question would be to induce the growth of mossy fibers into a region where they are not normally found and then determine

whether septal lesions will result in sympa-

thetic sprouting into the same region. Such an experiment is feasible since mossy fibers can be induced to grow into the CA1 region of the hippocampus by destroying CA3 pyramidal cells during developmentX3. Unfortunately, such an experiment would be of limited value since mossy fibers may induce changes in other ce11types in CA, such as glia1 cells, which could then influente sprouting or some other ce11 type in the dentate-CA, regions may be necessary in addition to the presente of granule cells. A more direct approach would be to study the topography of sympathetic sprouting in a species such as the European hedgehog which normally has mossy fibers in the CA1 region”. The suggestion that granule cells are required for sympathetic sprouting to occur has been called into question by Peterson and Loy”’ who repeated the neurotoxin experiments of Crutcher and Davis with slightly different conditions. They confirmed that the elimination of pyramidal cells with kainic acid did not prevent sympathetic sprouting but they also reported that destruction of granule cells with colchicine did not prevent sprouting in the ventral HF*. They concluded that neither granule cells nor pyramidal cells were obligatory for sympathetic sprouting and that some other ce11 type, such as interneurons, represents the target of sympathetic fibers. One possible explanation for the discrepancy between the two studies is that Crutcher and Davis injected the colchicine directly into the HF, possibly resulting in damage to ce11types other than the granule

Peterson

and Loy’s more

distant

injections

could have resulted in more specific lesions. Another possibility

is that the different

schedules of injections

used in the two studies accounted

for the contradicto-

ry results. The loss of granule cells may have to be complete for some time before the sprouting Signa1 is completely removed (see below). In support of this interpretation is the recent study by Kesslak et a1.74,‘17 confirming Crutcher and Davis.

the

origina1

observation

of

A very recent study, however, suggests that the mossy fiber hypothesis, in its simplest form, cannot explain the topography of sympathetic sprouting in another species. Booze et al.” undertook a study of sympathetic sprouting in the guinea pig HF. They found that such sprouting does occur but that, unlike the situation in the rat, sympathetic fibers extend into the CA, region of the guinea pig hippocampus. This occurs in spite of the fact that mossy fibers are restricted to the CA, region. The dissociation between the topography of granule ce11axons and the distribution of sprouted sympathetic fibers indicate that in the guinea pig, mossy fibers alone do not account for the topography of sympathetic sprouting. It would be interesting to determine whether removal of granule cells in the guinea pig HF would prevent or alter sympathetic sprouting as occurs in the rat, 5.3. The NGFhypothesis In the origina1 study of sympathetic sprouting regulation, Crutcher et al. proposed a mechanism to account for the specificity of this particular neuronal rearrangement3*. They suggested that NGF might normally be produced in the rat HF but transported away by septal fibers. Loss of the septal innervation was postulated to result in an accumulation of the factor which would then attract sympathetic fibers. In support of their hypothesis, Crutcher et al. cited two studies. The first was a study by Menesini-Chen et a1.146who showed that NGF injected intracerebrally into neonata1 rodents results in sympathetic ingrowth into the brainstem and spina1 cord. The second was the

unexpected

labeled

NGF

demonstration injected

that

into the adult

radioactively rat HF is selec-

* Interestingly, Peterson and Loy did find that sympathetic sprouting was prevented in the dorsal HF following colchicine injections which is the area where Crutcher and Davis made most of their observations. However, no comment was made on why sprouting would be inhibited at dorsal but not at ventral hippocampal levels.

219 tively transported locus coeruleus

by septal perikarya’$’ neurons.

A similar

but not by

hypothesis

was

tained

in earlier

studies.

The high concentration

NGF in the HF would be diluted

of

by other brain re-

subsequently made by Bjorklund and Stenevi based on the results of their transplant studies14. The growth of axons from implanted sympathetic ganglia

whole brain are used. It is also possible that different NGF precursor molecules are present in the brain69.

into the HF was much greater in the presente

The presente

tohippocampal denervation companied by hypertrophy

of sep-

and the ingrowth was acof the ganglionic peri-

gions containing

low NGF

levels when extracts

of NGF in the rat HF is certainly

sary, but not sufficient,

to implicate

of

neces-

this growth fac-

karya. The authors suggested that a neuronotrophic factor of glia1 origin might be involved in sympathetic

tor in the sympathetic sprouting response. Additional experiments are required to verify a role for NGF in such sprouting and these will be reviewed in the

sprouting and in the acceleration regeneration following injections

following section. Before examining

tryptamine.

In a subsequent

of locus coeruleus of 5,7-dihydroxy-

study, Gage et al. found

that denervation increased survival of transplanted embryonic sympathetic gangliag3. Unfortunately, at the time the NGF hypothesis was proposed there was considerable controversy as to whether or not NGF was present in the brain75,‘05* 108,182r86,198.In fact, most investigators concluded that NGF was not present in significant amounts in the CNS80,173,193and most studies of the effects of NGF on centra1 neurons focused on centra1 noradrenergic fibers 8*10,113.Yet, attempts to demonstrate effects of NGF on locus coeruleus neurons in culture had consistently failed64y’67. Given the remarkable specificity of the sympathetic sprouting response, however, the question of the presente of NGF in the CNS seemed to warrant further examination. Using a sensitive bioassay developed by Collins29*30, Crutcher and Collins49 found that extracts of norma1 rat HF accelerated the rate of neurite elongation from both sympathetic and parasympathetic embryonic chicken neurons in culture. The effect on sympathetic neurons was antagonized by antiserum to NGF, whereas the effect on parasympathetic neurons was not affected by NGF antiserum. Such results supported the conclusion that NGF-like activity is present in the rat HF but the identification of the attive molecule(s) as authentic NGF could not be determined with a bioassay alone. Severa1 recent studies, however, have demonstrated that authentic NGF, as well as the messenger RNA for NGF, are present in the rat HF124~‘84~185~201. In fact, the concentrations of both protein and message are much higher in this brain region than in any other region. This regional concentration probably accounts, in part, for the inconsistent results ob-

the evidente

in support

of the

NGF hypothesis, however, it is worth asking what the norma1 role of hippocampal NGF might be. The origina1 observation of Schwab et al. lgoon the affinity of septohippocampal neurons for exogenous NGF implicated these cells as having NGF receptors and, therefore, as possible targets of endogenous hippocampa1 NGF. Similar results were obtained in the basa1 forebrain following injections of labeled NGF in the cortex’83. The presente of NGF receptors in the brain has been confirmed biochemically’94 and there is increasing evidente that NGF exerts effects on septal cholinergic neurons10031”“*106as well as on other forebrain cholinergic neurons150.151 or transplanted cholinergic cells196. This includes in vivo’O”l 150.151,204 and in vitrolO4,106 studies demonstrating that exogenous NGF will stimulate cholinergic enzyme activity. Exogenous NGF also appears to be capable of reducing the extent of injury-induced cholinergic and non-cholinergic septal ce11loss following fimbrial transection’0’.205. Additional support for a role of endogenous NGF-like activity in septohippocampal function is provided by the recent finding that entorhinal lesions, which result in septohippocampal sprouting in the dentate gyrus34,140+156-158.188.192 and greater innervation by septal septal implants’*, also result in greater hippocampal NGF-like activity54. This, in turn, suggests that entorhinal inputs may regulate NGF levels in the HF. Such regulation is presumably not due to retrograde transport since no evidente for such transport was found in the study by Schwab et al.igO. In summary, increasing evidente suggests that NGF has specific functions in the mammalian CNS and that septohippocampal neurons represent at least one population of neurons responsive to this particular growth factor.

220 5.4. Evidente that NGF is involved in sympathetic sprouting

As mentioned above, the mere presente of a growth factor is necessary, but not sufficient, in of itself, to implicate it in a particular growth response. Recent data, however, do provide evidente for such a role of NGF in the sympathetic sprouting response. Collins and Crutcher3’ measured NGF-like activity in medium that had been conditioned by slices of the rat HF taken from animals that either received a septal ‘esion or a contro1 lesion not interrupting the septohippocampal pathway. They found that one week following a septal lesion, a time point just before sympathetic sprouting is observed in the HF147, there was significantly more NGF-like activity in slice-conditioned medium when compared with contro1 medium (conditioned by non-denervated slices). Of particular interest was the finding that slices of the dentateCA3 region of the HF, where sympathetic sprouting occurs, resulted in significantly greater activity compared with slices from the CAI region of the hippocampus, where sprouting does not normally occur. This was true both in contro1 tissue as well as in tissue from septal-lesioned animals. Korsching et al. 125recently measured hippocampal NGF, as well as mRNA for NGF, following a fimbrial transection and also found that NGF protein levels, but not mRNA for NGF, increased significantly by 2 weeks but returned to contro1 levels by 4 weeks. They concluded that the increased NGF was due to loss of retrograde transport by septal fibers and not due to increased protein synthesis. Interestingly, they also concluded that the increase in hippocampal NGF was unlikely to be sufficient to elicit sympathetic sprouting although the reasons for this conclusion were not given. Whittemore et al.202 recently reported similar results to that of Korsching et al. but added the observation that fimbrial transection in 2week-old rats resulted in increased NGF mRNA levels in the HF. Yoshida et al.*% also found that septal deafferentation results in greater growth-promoting effects of hippocampal extracts on CNS ce11cultures although the effect of anti-NGF antibodies on the activity was not reported. Otten et al.168recently confirmed previous reports of increased NGF after cholinergic denervation but also added the important observation that there is no increase in hippocampal NGF following centra1 noradrenergic denervation.

It is worth noting that the approach used by Collins and Crutcher is quite different from that used by other investigators t255168*202. In the latter studies, extracts of hippocampal tissue have been assayed for

NGF levels with sensitive radioimmunochemical assays where Collins and Crutcher have studied medium conditioned by hippocampal slices using a biological assay. As a result it is difficult to directly compare the results obtained in the studies. Nevertheless, both approaches demonstrate that hippocampal NGF increases in response to septohippocampal lesions. Springer and Loy”’ tested the NGF hypothesis by injecting NGF antiserum into the HF after fimb~al transection. They found that such injections prevented the sympathetic sprouting response. They also reported that such injections resulted in the loss of dentate granule cells. If these effects are due to direct antagonism of endogenous NGF-like activity, then the hypothesis that NGF is involved in sympathetic sprouting would be strengthened considerably. In spite of these promising studies, severa1 unanswered questions remain. For example, what cells give rise to the NGF? The colchicine experiments described previously suggest that the granule cells may contribute to generating the trophic Signa1 but other cells, such as glia or interneurons, have been suggested to be involved as we11’4,170. In fact, one of the early suggestions was that NGF, or a similar factor, is present in the granule celi axons (mossy fibers) and that the high zinc content of the mossy fibers reflects stores of NGF or NGF precursors. This suggestion was based on the fact that the 7s form of NGF is normally complexed with zinc749’91.Another piece of TABLE 11 Evidente that NGFis involved in sympathohippocampal ing

sprout-

References

49,124,201 NGF is present in the hip~campal formation NGF mRNA is present in the hippocampal 124,172,185,201 formation Hippocampal NGF increases following 31,125,202 septal denervation Greatest NGF levels and increases occur in 31,124 areas of sprouting 187 Anti-NGF a~tibodjes block sympathetie sprouting Granule cells, whose absence blocks sprouting, 46,74,117,172 contain NGF mRNA

221 evidente

that might support granule

cells as a source

of NGF-like activity is that new granule cells are added throughout adult life6 and NGF production is commonly associated with mitotic ce11populations. Alternatively, NGF-like activity has been suggested to arise from glia1 cells14 or intemeurons17’. Lindsay 13’ showed some years ago that astrocytes in culture can support NGF-dependent neurons. There is evidente from other studies that astrocytes and other glia1 cells can produce NGF in tissue culture and that Schwann cells exhibit NGF receptors dista1 to a peripheral nerve crush’95. In the iris, a known producer of NGp7, immunohistochemistry of NGF seems to localize the NGF immunoreactivity to Schwann cells174. There have been some immunohistochemical studies which demonstrate NGF-like im-

Fig. 5. Schematic of the major cell types and inputs likely to be involved in the sympathetic sprouting response. A granule celi (GC) and hilus interneuron (HI) receive septal innervation but the association of septal terminals with hippocampal blood vessels has not been observed. Sympathetic fibers innervate the blood vessel at the top of the figure but do not accompany the penetrating branch which is surrounded by basa1 lamina. Although NGF is known to be produced in the hippocampal formation, the cells which synthesize it are unknown although much of the evidente is consistent with granule cells as one source. Three candidates are shown here, granule cells (GC), hilus interneurons (HI) and astrocytes (AST). The asterisk (*) simply indicates possible sites of release. The retrograde transport of NGF by septal fibers has been demonstrated.

munoreactivity

in both the developing4

nervous system but no systematic

and maturezol

description

of such

staining in the adult rat HF has appeared. The staining that has been reported was associated with fibers, perhaps reflecting the accumulation andlor transport of the growth factor from target tissue. Unfortunately, none of these studies provide direct evidente as to the cellular source of the NGF. This question can only be answered through the use of in situ hybridization techniques to localize the mRNA for NGF in tissue sections. A recent study17* using this approach revealed that mRNA for NGF is localized to the granule ce11layer in the dentate gyrus and, to a lesser extent, to the pyramidal ce11 layer in the hippocampus. If confirmed, such data indicate that NGF is synthesized by the major neuronal populations in the HF. The presente of mRNA for NGF in the CA, pyramidal cells is consistent with the results showing NGF in this region31s124. Table 11 summarizes the evidente supporting a role for NGF in

Fig. 6. Schematic illustrating the changes in the hippocampal formation which occur following septohippocampal denervation (dashed lines represent degenerated fibers and terminals). NGF levels increase in the tissue and sympathetic fibers are observed in association with penetrating blood vessels (BV). At least some of the perivascular fibers are separated from the neuropil by basa1 lamina (arrows). Whether sympathetic axons synapse within neuropil has not been established with certainty.

222 sympathetic sprouting. If NGF does play a role in eliciting

sympathetic

sprouting how would this be accomplished? The original hypothesis3* can be expanded and strengthened

the factor. If variations in the distribution of blood vessels between different species result in different pattems of growth factor accumulation, for example,

on the basis of recent data. Schematic summaries are presented in Figs. 5 and 6 to illustrate these ideas.

then sympathetic axons may appear in different regions based upon the distribution of the factor. If perivascular accumulation of the factor is an impor-

Basa1 forebrain neurons, many of which are cholinergic, exhibit retrograde transport of NGF180,‘83. NGF

tant aspect of initiating and directing the sprouting response then the delayed time course147 might re-

is normally synthesized and, presumably, released by neuronal and/or glia1 cells in the H~v,124,184. Three

flect the accumulation of the factor secondary to loss of retrograde transport by septal fibers. Another ap-

candidates have been suggested and are shown in Fig. 5. They are the dentate granule cells, hilus interneurons and astrocytes. Levels of hippocampal NGF may be regulated by retrograde transport by septal neurons or through activity in the septohippocampal pathway or both. Damage to the septohippocampal projection results in increased hippocampal NGF31,125 levels with no corresponding increase in mRNA for NGF’25,201. The increased leve1 of NGF is then detected by extracerebral sympathetic axons which respond by growing into the HF along penetrating blood vessels. The latter part of this hypothesis is speculative and, according to Korsching et al., incorrect’25. It is true that the tota1 increase in NGF as measured in extracts is not large following septal denervation but it is difficult to determine whether it is sufficiently great to elicit sprouting. If, for example, NGF accumulates in the perivascular region, similar to the protein accumulation observed by Rennels et al.t71, then even greater loca1 concentrations might be expected in the regions accessible to sympathetic axons. At this point we can only note the correlations between the regional and tempora1 distribution of NGF in the HF and the topography and time course of sympathetic sprouting. The results of the anti-NGF experiments of Springer and Loy lx7 also suggest that hippocampal NGF-like activity is necessary, if not sufficient, to elicit this sprouting response. In order to account for the discrete topography of sympathetic sprouting, and for variations in the topography observed in different species, it will be important to localize the distribution of the putative growth-promoting signal. It is possible, for example, that one ce11 population, such as the dentate granirle cells, represents the major source of the NGF-like activity but that the topography of sympathetic sprouting depends on the regional accumulation of

parent paradox

might be resolved

by this hypotheti-

tal mechanism. The failure to block sympathetic sprouting elicited by septal lesions made within a week after destruction of granule cells170 but not after longer intervals46,117, may reflect persistente of the sprouting signal for a period at least as long as it takes to initiate sympathetic sprouting, i.e. 7-9 days. The recent in situ hybridization data are very usefu1 since they suggest that at least one of the proposed candidates for the growth signal, the dentate granule cells, are sites of NGF synthesis. It is also possible that they, or other celi types, could accumulate NGF as has been shown for HRP injected into the ventricular system’“. Although reports 125.202indicate that mRNA for NGF does not increase in the adult HF following fimbrial transection, it will be important to confirm these results using in situ hybridization techniques. What happens to NGF levels after sympathetic axons have grown into the HF? Korsching et a1.125reported that NGF levels returned to contro1 values by 4 weeks after fimbrial transection. Our own studies suggest that greater NGF-like activity is detectable in hippocampal slice-conditioned medium for severa1 months after a media1 septal lesion3*. Superior cervital ganglionectomy does not appear to affect NGF levels, indicating that the sympathetic axons are not regulating NGF availability, either through retrograde transport or through anterograde influences. Kromer reported that septal implants would co-innervate the HF along with sympathetic fibers126 and Gage and Bjorklund” found increased survival of septal grafts in the presente of septohippocampal denervation. Gasser et a1.86 found that sympathectomy did not affect recovery of cholinergic activity in the HF following partial fimbrial transection. Such recovery is thought to be due to the sprouting of spared cholinergic septohippocampal fibersa4.

223 If septohippocampal

sprouting

on

the

part

of

lective pressure

for such a specific growth response

spared septal fibers following a fimbrial lesion occurs

nor is there any evidente

in response to NGF-like activity then the apparent absence of competition with sympathetic fibers indi-

represents reversion to a phylogenetically or ontogenetically primitive pattern of connectivity65T”2. More likely is the possibility that such sprouting reflects an unmasking of mechanisms normally operating in the

cates that sympathohippocampal fibers do not regulate NGF levels in the HF. This is not surprising if the primary means of regulating hippocampal NGF levels is via retrograde transport since the density of sympathetic fibers, assuming they enter the neuropil, is probably a small percentage of that normally provided by septal fibers. Madison and Davis’ argued that sympathetic sprouting continues for severa1 months indicating that the trophic signal may be present for an extended period after the lesion. 5.5. Non-NGF-like growth factors and sympathetic sprouting In addition to NGF-like activity there is considerable evidente for the presente of other growth factors in the CNS that are not related to NGti9,52,159, 162-164.At least one of these factors is attive on parasympathetic neurons49 and may be related to the soluble activity of heart cell conditioned medium described by Collins 29*30 . Schonfeld et al. have shown that heart ce11 conditioned medium enhances regeneration of centra1 cholinergic neurons178,179. Of particular interest is the recent demonstration that the parasympathetic growth-promoting activity does not increase after a fimbrial transection98. Recently, Ebendal et aI.@ found that similar activity in the iris did not increase following sensory, sympathetic or parasympathetic denervation. Ojika and Appe1165 found growth effects of hippocampal extracts on septal neurons in vitro that were not antagonized by antiNGF antibodies. However, this system has not been studied for possible changes in such activity following septal denervation. 6. POSSIBLE FUNCTIONS

OF SYMPATHOHIPPOCAM-

PAL FIBERS

Whether or not there is some evolutionary significance of this sympathetic sprouting response, an issue that has been reviewed recently for other examples of neuronal plasticity 73, is difficult to determine. It seems extremely unlikely that this sprouting response represents an adaptive mechanism for recovery of function since there is no reason to suspect se-

to support the view that this

development and/or maintenance of specific populations of neurons within the CNS. In fact, one of the contributions that examples of such injury-induced plasticity provide is in pointing to the existence of specific growth-promoting factors in the mature CNS. However, the question of the possible functional significante of examples of neuronal sprouting, whether adaptive or maladaptive, is important and will be dealt with in this section. Function can be analyzed at severa1 levels. For example, the demonstration of functional innervation based on electrophysiological criteria does not necessarily mean that the projection is behaviorally relevant. In addition, such sprouting could be behaviorally relevant but maladaptive, affecting behavior in an adverse manner. Thus, the sympathohippocampal pathway, or other neuronal rearrangements for that matter, may be adaptive, neutral, or maladaptive in terms of functional recovery. 6.1. Electrophysiological and pharmacological studies Since the sympathetic fibers invading the HF are collaterals of axons that normally project to extracerebral structures, it seems reasonable to assume that they will be attive during stimulation of the sympathetic component of the autonomie nervous system. Kimble et al. simultaneously stimulated the SCG and recorded from electrodes in the HF after fimbrial lesions123. They reported no effect of such stimulation on electrical activity in the HF. Barker et al.‘, however, did find that removal of sympathetic fibers following media1 septal lesions resulted in increased spontaneous activity in the HF. Howard and Barker107 recently found the opposite effect in studies of sympathetic sprouting in the media1 habenula. Another line of investigation bears on the possible electrophysiological function of sympathohippocampal fibers. Freedman et a1.76 transplanted pieces of hippocampus to the anterior chamber of the eye to test whether the transplant would be functionally innervated by the cholinergic and noradrenergic fibers

224 normally innervating the iris. Although they were not examining this system in light of the sprouting response reviewed

here, they found that sympathetic

fibers were capable of affecting the spontaneous electrical activity of the hippocampal transplants. It seems likely,

therefore,

that sympathohippocampal

fibers elicited by septohippocampal denervation are capable of influencing HF activity. Another possible index of function is whether sympathetic fibers affect the affinity, number

or distribu-

It is clear that in order to draw some conclusions regarding the functional replacement of sympathetic fibers for septal fibers it will be necessary to identify the actual transmitters

present

in the two systems.

The existence of non-cholinergic septohippocampal neurons, for example, has already been demonstrated197. 6.2. ~etabolic studìes Even if firm evidente

cannot be obtained

support-

tion of noradrenergic receptors in the HF. In order for the sympathetic fibers to elicit effects on their tar-

ing a possible electrophysiological or pharmacologital influente of sympathetic fibers on hippocampal

gets, released transmitter, e.g. NE, must interact with noradrenergic receptors coupled to physiologital responses. The specificity of sympathetic sprouting in response to loss of cholinergic fibers could represent an example of transmitter substitution within the CNS. Thus, in this mode1 it might be possible to determine the fate of putative cholinergic and noradrenergic receptors while the transmitter rearrangement occurs. The ability to undertake such studies, however, is limited to some extent by technical considerations and is further complicated by the fact that the sympathetic fibers share a common transmitter with centra1 noradrenergic neurons. Which receptors, for example, will be activated by which noradrenergic neurons? The regional distribution of both cholinergic and

function, other interactions are possible. Since sympathetic fibers normally influente extracerebral blood vessels they could affect blood flow in the HF. This possibility has not been directly addressed but Harrell et al. have studied metabolic changes in the HF following media1 septal lesions95-97. Cholinergic denervation of the HF was found to not have long-lasting effects on glucose metabolism. Furthermore, sympathetic ingrowth was reported to affect metabolic activity in the stratum lacunosummoleculare, a region, however, not directly innervated by sympathetic fibers. These results are difficult to interpret but do suggest that sympathetic fibers may exert an influente on hippocampal glucose metabolism. Whether the effect is direct or secondary to an influente on blood flow remains to be determined.

noradrenergic receptors within the HF has been assessed with homogenate binding3’ assays and with ‘histopharmacologica1”99 methods. A study by Morrow et al.‘53 provided evidente that some putative anoradrenergic receptors in the HF increase following media1 septal lesions and the suggestion was made that this increase correlated with sympathetic sprouting137. However, in subsequent work these same investigators found that the increase was actually in serotonergic receptors and that it was not the result of septohippocampal denervation but rather, of the interruption of serotonergic fibers passing through the septum 154.As a resuh, it is stili not clear whether sympathetic sprouting is related in some way to changes in transmitter receptors which occur following media1 septal lesions. There have also been reports of some changes in cholinergic receptors after septohip~campal dene~ation but none of these have been shown to be related to sympathetic sprouting’55.

The studies by Harrell et al. are contradicted by data presented by Kelly et al.“’ who reported that fimbrial transections resulted in long-lasting (6 months) reductions in glucose utilization. These investigators did not assess the possible influente of sympathetic sprouting by performing superior cervital ganglionectomies SO it seems likely that the sympathohippocampal fibers did not exert an influente on glucose metabolism in their study. Of interest is the fact that transplants of embryonic septal tissue did result in an increase in hippocampal glucose utilization that corresponded to the innervation of the hippocampus by grafted septal neurons. 6.3. Behavioral studies The possible role of sympathetic sprouting in behavior has been addressed in severa1 studies. Kimble et ai. first tested for such a possibihty in rats sustaining lesions of the anterior pole of the hippocampal

225 formation120-‘23. pathectomy

bilatera1

sym-

sprouting

had no effect on severa1 behavioral

They

defi-

on this particular

cits that accompany duced spontaneous

found

that

hippocampal damage, e.g. realternation, poor spatial maze

learning, slower extinction of conditioned sion, and decreased open field activity.

taste aver-

may actually

inhibit

recovery

of function

task.

Kesslak and Gagen6

recently

reported

of a study in which they examined

the results

a possible role for

sympathetic fibers in recovery of performance on a forced-choice alternation task following damage to

In other studies, Chafetz et a1.24 and Gage et a1.82 studied the correlation between sympathetic sprout-

CA, and CA, pyramidal tions. After the animals

ing and changes in severa1 behavioral

this task superior cervical ganglionectomy resulted in an additional performance deficit suggesting that the

measures

after

septal lesions. Some behaviors recovered with a time course that correlated with sympathetic sprouting but other behavioral

deficits persisted.

Chafetz et al.

concluded that sympathetic sprouting is functional and that the increased NE content, as assessed with histofluorescence, is best correlated with functional recovery. Unfortunately, they did not directly determine whether sympathetic sprouting is responsible for the recovery by removing the SCG. Gage et a1.82 reported that sympathetic fibers did contribute to such behavior as heat-induced paw-licking and avoidance of bright light and that different effects were obtained depending on the lighting conditions used. The question of behavioral function was also addressed by Crutcher et al. 48, by studying a behavioral deficit which was specifically associated with cholinergic denervation of the HF and which recovered with the same time course as sympathetic sprouting. In order to eliminate the possibility that other neuronal systems could mediate the recovery they used the animals as their own controls. The task used was performance on the radia1 arm maze, a behavior which has been shown to be dependent on an intact HF. They found that media1 septal lesions result in a deficit on this task which is highly correlated with the extent of cholinergic denervation of the HF. The recovery of the behavior was found to occur with a time course closely matching the time course of sympathetic sprouting. However, subsequent bilatera1 superior cervical ganglionectomy had no effect on the recovered behavior. Thus, sympathetic sprouting did not appear to mediate or interfere with behavioral recovery. In fact, Harrell and Davis undertook a study using a similar paradigm except that the effect of prior sympathectomy was assessed on the ability of the animals to learn the radia1 arm maze task. They found greater rates of learning in the sympathectomized animals and concluded that sympathetic

sympathetic

cells with kainic acid injecrecovered performance on

fibers may be involved

in the initial re-

covery. Ganglionectomy did not, however, prevent re-acquisition of the behavior indicating that sympathetic fibers do not have an obligatory role in recovery of function on this task. Collectively, these studies suggest that sympathetic sprouting may mediate recovery of behavioral function following septal lesions or hippocampal damage, on some tasks and may, on the other hand, interfere with such recovery for other tasks. The number of possible effects of sympathetic sprouting on hippocampal function is vast and the feasibility of testing them will only increase as our understanding of the norma1 function of the HF progresses. The importance of documenting possible functional influences of sympathohippocampal sprouting is emphasized by the possibility that such sprouting could occur in humans (see below). 7. CONCLUSIONS AND FUTURE DIRECTIONS

7.1. Are sympathetic axons replacing septal terminals? In some ways sympathetic sprouting in response to septohippocampal denervation is a valuable mode1 for studying the mechanisms, as well as functional significante, of neuronal plasticity in the mature CNS. Much of the value of this system stems from the isolation of the sprouting neurons in a peripheral ganglion, thus making it possible to selectively remove, stimulate, or labe1 the sprouted pathway without interrupting other projections to the target. In addition, the ability to visualize the sympathetic fibers with histofluorescent and anterograde tracing methods allows for histochemical analysis as well as ultrastructural studies of synaptic relationships. With additional electron microscopie studies it should be possible to determine whether classica1 synaptic con-

226 tacts are formed

and, if SO, what hippocampal

ele-

ments are contacted. Such information is necessary in order to determine whether synaptic substitution is actually occurring. Based on the evidente currently available, this seems not to be the case since the vast majority

of sympathetic

fibers

appear

to be asso-

ciated with blood vessels yet septohippocampal

ter-

sprouting

response.

7.3. IS sympathetic sprouting functional? Although the biologica1 mechanism accounting such growth in the mature

for

CNS is reason enough for

studying this particular sprouting response, other aspects of this phenomenon are also of interest. The

minals have not been shown to exhibit such vascular associations. Negative evidente is never definitive,

question of functional significante, if any, has not been definitively answered. This aspect may also be

however,

and in light of the recent data presented

by

relevant

Armstrong3, the possibility of centra1 cholinergic nervation of parenchymal blood vessels remains.

in-

mans

If synaptic

substitution

does not occur, and since

the sympathetic axons are apparently completely different in terms of transmitter, how do we account for the specificity of this sprouting response? The answer must lie in the affinity and responsiveness of both septohippocampal and sympathetic neurons for a growth factor such as NGF. Without this link it is difficult to understand the specificity of the sprouting response. 7.2. IS NGF involved in sympathetic sprouting? The presente of NGF as well as mRNA for NGF in the rat HF is now beyond doubt. In situ hybridization data”* further suggest that dentate granule cells, and possibly hippocampal pyramidal cells, contain the NGF message. Thus, this well-characterized growth factor is present in the regions where sympathetic sprouting occurs. Furthermore, NGF levels increase in the HF following septohippocampal denervation and injection of anti-NGF antiserum appears to block the sympathetic sprouting response. Al1 of the evidente is therefore consistent with a role for NGF in eliciting sympathetic ingrowth. In order to determine whether an increase in hippocampal NGF alone is sufficient to elicit sympathetic sprouting we have undertaken preliminary experiments in which NGF is injected directly into the ventricle or HF via a chronically implanted cannula”‘. We have not observed sympathetic sprouting in any brain region, including the HF, following the injection of NGF. However, injection of NGF does appear to result in enhancement of the sympathetic sprouting elicited by septohippocampal denervation. These results suggest that some other aspects of septohippocampal denervation, possibly in addition to the increase in NGF levels, is involved in eliciting the

if a similar sprouting (see

below).

response

Unfortunately,

occurs in huas alluded

to

above, there is no simple way of assaying the potential function

of this system, partly because our under-

standing of the function of the HF is stili in its infancy. Yet, as interest in this area grows, we may yet uncover possible influences the sprouted sympathetic system may have on hippocampal function, whether adaptive or maladaptive. 7.4. Possible clinica1 significante There are two areas in which this example of neuronal plasticity may shed light on human neuropathology. The first is in the general area of the constraints on neuronal growth in the mature brain and spina1 cord. That such growth is possible even in the human brain*‘,*’ is now evident but the specific requirements of such growth remain elusive. The results obtained from studies of this particular example of neuronal growth dovetail nicely with the results obtained in other systems. Thus, the requirements of growing axons for both diffusible growth factors as well as substrate-bound matrix factors seem to be a recurring theme both in developmental studies of neuronal growth as well as in the cases of successful growth in the mature CNS’,‘*. This leads to the general conclusion that stimulation of such growth in the mature brain or spina1 cord will require attention to at least these two general classes of growth factors. The other area of possible clinica1 application is more direct. Alzheimer’s disease is characterized, in part, by a degeneration of basa1 forebrain cholinergic neurons200 many of which have recently been shown to contain NGF receptors’a’. Since it appears that the loss of just these cells in the rodent brain accounts for sympathetic sprouting, we may well ask the question whether a similar growth response occurs in Alzheimer’s disease. The answer to this question, in turn, depends on what the causative agent in Alz-

227 heimer’s

disease is. For example,

if interferente

the norma1 trophic support for basa1 forebrain ergic neurons

is involved

of

cholin-

in the etiology of Alzheim-

themselves

to solve the mysteries

ity but progress

of neuronal

plastic-

can be made when attention

rected to specific growth responses

is di-

such as the one

er’s disease2*9y~103,then sympathetic ingrowth would not be expected. If, on the other hand, some other aspects of Alzheimer’s disease account for the loss of

reviewed here. In spite of the many unanswered questions, the past 10 years have witnessed the study of this isolated example of anomalous neuronal

centra1 cholinergic neurons and NGF-like activity increases as a result of loss of these cells, then sympa-

sprouting grow into an area of intensive ing investigation.

thetic sprouting

might well be expected

to occur. Re-

cent data suggest that NGF or its message changed in Alzheimer’s tissuegO. How would sympathetic sprouting human autopsy material? Identifying

and expand-

are not

be identified sympathetic

in fi-

bers through the use of standard histofluorescent techniques would not be feasible, although immunohistochemical procedures might be of use since the noradrenergic synrhetic enzymes persist for severa1 hours. In fact, Booze et al.16 recently reported that tyrosine-hydroxylase-positive fibers exhibiting a ‘peripheral sympathetic morphology’ are present in autopsy materia1 taken from victims of Alzheimer’s disease, suggesting that sympathetic ingrowth may be occurring. Recently, plaques in aged subhuman primates have been shown to contain neurites positive for severa1 different neurotransmitters including noradrenergic synthetic enzymes. It is possible that sympathetic sprouting occurs in cholinergic-denervated brain regions in Alzheimer’s disease and that some of these fibers contribute to the formation of senile neuritic plaques. Another way of addressing this hypothesis might be to determine whether plaque-like structures develop in animals with chronic septal lesions. Unfortunately, if plaque formation requires severa1 years, as seems likely, this approach could only be undertaken in long-lived species. Focused studies of relatively simple models of neuronal growth in the mature brain are not sufficient in REFERENCES Aguayo, A.J., Axonal regeneration from injured neurons in the adult mammalian centra1 nervous system. In C.W. Cotman (Ed.), Synuptic Plusriciry, Guilford, New York, 1985, pp. 457-484. Appel, S.H., A unifying hypothesis for the cause of amyotrophic lateral sclerosis, parkinsonism, and Alzheimer’s disease, Am Neurol., 10 (1981) 499-505. Armstrong, D.M., Ultrastructural characterization of choline acetyltransferase-containing neurons in the basa1 forebrain of rat: evidente for a cholinergic innervation of intracerebral blood vessels, J. Comp. Neuro/., 250 (1986)

SUMMARY Sympathetic

sues

but

are

fibers

innervate

normally

many

confined

peripheral

tis-

to extracerebral

structures within the cranial cavity, e.g. blood vessels. The invasion of the centra1 nervous system by vascular sympathetic axons is a unique example of neuronal plasticity which provides new information concerning the regulation and mechanisms of neuronal sprouting in both the peripheral and centra1 nervous systems. In this paper, the principal findings concerning the conditions under which such sprouting occurs, the mechanisms which may be involved, and the question of its possible function are reviewed. Of special interest is the fact that a nerve growth factor-like brain factor may be involved in this growth response. The principles gleaned from studies of this sprouting phenomenon may be applicable to other models of neuronal plasticity and may have clinica1 relevance. ACKNOWLEDGEMENTS Preparation of this review and much of the work cited was supported by grants from the National Science Foundation (BNS-8501269) and the National Institutes of Health (NS-17131). Special thanks to Nancy Peixhot for secretarial assistance. 81-92. 4 Ayer-Lelievre, C.S., Ebendal, T., Olson, L. and Seiger, A., Localization of nerve growth factor-like immunoreactivity in rat nervous tissue, Med. Biel., 61 (1983) 296-304. 5 Barker, D.J., Howard, A.J. and Gage, F.H., Functional significante of sympathohippocampal sprouting: changes in single celi spontaneous activity, Brain Ra., 291 (1984) 357-363. 6 Bayer, S.A., Yackel, J.W. and Puri, P.S., Neurons in the

rat dentate gyrus granular layer substantially increase during juvenile and adult life, Science, 216 (1982) 890-892. 7 Bevan, R.D. and Tsuru, H., Functional and structural changes in the rabbit ear artery after sympathetic denerva-

228 tion, Circ. Res., 49 (1981) 478-485. 8 Bjerre, B., Bjorklund, A. and Stenevi, U., Stimulation of growth of new axonal sprouts from lesioned monoamine neurons in adult rat brain by nerve growth factor, Brain Res., 60 (1973) 161-176. 9 Bjorklund, A., Owman, C. and West, K.A., Peripheral sympathetic innervation and serotonin cells in the habenular region of the rat brain, Z. Zellforsch., 127 (1972) 570-579. 10 Bjorklund, A. and Stenevi, U., Nerve growth factor: stimulation of regenerative growth of centra1 noradrenergic neurons, Science, 175 (1972) 1251-1253. 11 Bjorklund, A. and Stenevi, U., Experimental reinnervation of the rat hippocampus by grafted sympathetic ganglia. 1. Axonal regeneration along the hippocampal fimbria, Brain Res., 138 (1977) 259-270. 12 Bjorklund, A., Kromer, L.F. and Stenevi, U., Cholinergic reinnervation of the rat hippocampus by septal implants is stimulated by perforant path Iesion, Brain Res., 173 (1979) 57-64. 13 Bjorklund, A. and Stenevi, U., Regeneration of monoaminergic and cholinergic neurons in the mammalian centra1 nervous system, Physiol. Rev., 59 (1979) 62-100. 14 Bjorklund, A. and Stenevi, U., In vivo evidente for a hippocampal adrenergic neuronotrophic factor specifically released on septal deafferentiation, Brain Res., 229 (1981) 403-428. 15 Bjorklund, H., Hokfelt, T., Goldstein, M., Terenius, L. and Olson, L., Appearance of the noradrenergic markers tyrosine hydroxylase and neuropeptide Y in cholinergic nerves of the iris following sympathectomy, J. Neurosci., 5 (1985) 1633-1643. 16 Booze, R.M., Rases, A., Cook, G.M. and Davis, J.N., Evidente for anomalous sympathetic ingrowth fibers in patients with Alzheimer’s disease, Soc. Neurosci. Absrr., 12 (1986) 101. 17 Booze, R.M., Laforet, G. and Davis, J.N., Hippocampal sympathetic ingrowth in rats and guinea pigs: quantitative morphometry and topographical differences, Brain Res., 375 (1986) 251-258. 18 Booze, R.M. and Davis, J.N., Maintenance of sympathetic ingrowth fibers in aged rat hippocampus, Neurobiol. Aging, submitted. 19 Cardinali, D.P., Vacas, MI. and Gejman, P.V., The sympathetic superior cervical ganglia as peripheral neuroendocrine centers, J. Neural Transm., 52 (1981) 1-21. 20 Casell, M.D. and Brown, M.W., The distribution of Timm’s stain in the nonsulphide-perfused human hippocampa1 formation, J. Comp. Neurol., 222 (1984) 461-471. 21 Cervos-Navarro, J. and Matakas, F., Electron microscopic evidente for innervation of intracerebral arterioles in the cat, Neurology, 24 (1974) 282-286. 22 Chafetz, M.D. and Gage, F.H., Fluorescente measurement of lesion-induced fiber growth. 1. A microfluorometric technique to measure density and intensity of variscosities and fibers, Brain Res., 247 (1982) 209-216. 23 Chafetz, M.D., Evans, S. and Gage, F.H., Fluorescente measurement of lesion-induced fiber growth. 11. Measurement of sympathohippocampal sprouting using a new microfluorometric method, Brain Res., 247 (1982) 217-226. 24 Chafetz, M.D., Evans, S. and Gage, F.H., Recovery of function from septal damage and the growth of sympathohippocampal fibers, Physiol. Psychol., 10 (1982) 391-398. 25 Chafetz, M.D. and Gage, F.H., Pineal and habenula in-

26

27

28 29 30

31

32

33

34

35

36 37 38

39

40

41

42

43

44

nervation altered by septal lesion, Brain Res. Bull., 10 (1983) 27-31. Chandler, J.P. and Crutcher, K.A., The septohippocampal projection in the rat: an electron microscopie horseradish peroxidase study, Neuroscience, 3 (1983) 685-696. Clarke, D.J., Cholinergic innervation of the rat dentate gyrus: an immunocytochemical and electron microscopital study, Brain Res., 360 (1985) 349-354. Clemente, C.D., Regeneration in the vertebrate centrai nervous system, Int. Rev. Neurobiol., 6 (1964) 257-301. Collins, F., Axon initiation by ciliary neurons in culture, Dev. Biol., 65 (1978) 50-57. Collins, F., Induction of neurite growth by a conditioned medium factor bound to the culture substratum, Proc. Nari. Acad. Sci. U.S.A., 75 (1978) 5210-5213. Collins, F. and Crutcher, K.A., Neurotrophic activity in the aduh rat hippocampal formation: regional distribution and increase after septal lesion, J. Neurosci., 5 (1985) 2809-2814. Collins, F. and Crutcher, K.A., Sustained increase in NGF-like activity in medium conditioned by hippocampal slices from septal-lesioned rats, Soc. Neurosci. Abstr., 12 (1986) 787. Cook, T.M. and Crutcher, K.A., Extensive target ce11loss during development results in mossy fibers in the regio superior (CA,) of the rat hippocampal formation, Dev. Brain Res., 21 (1985) 19-30. Cotman, C.W., Matthews, D.A., Taylor, D. and Lynch, G., Synaptic rearrangement in the dentate gyrus: histochemical evidente of adjustments after lesions in immature and adult rats, Proc. Natl. Acad. Sci. U.S.A., 70 (1973) 3473-3477. Cotman, C.W., Nieto-Sampedro, M. and Hairis, E.W., Synapse replacement in the nervous system of adult vertebrates, Physiol. Rev., 61 (1981) 684-784. Coyle, P., Vascular patterns of the rat hippocampal formation, EXP. Neurol., 52 (1976) 447-458. Coyle, P., Spatial features of the rat hippocampal vascular system, EXP. Neurol., 58 (1978) 549-561. Crutcher, K.A., Brothers, L. and Davis, J.N., Sprouting of sympathetic nerves in the absence of afferent input, EXP. Neurol., 66 (1979) 778-783. Crutcher, K.A. and Davis, J.N., Hippocampal alpha- and beta-adrenergic receptors: comparison of (3H)-dihydroalprenolol and (3H)-WB4101 binding with noradrenergic innervation in the rat, Brain Res., 182 (1980) 107-117. Crutcher, K.A. and Davis, J.N., Noradrenergic sprouting in response to cholinergic denervation: the sympathohabenular connection, EXP. Neurol., 70 (1980) 187-191. Crutcher, K.A., Brothers, L. and Davis, J.N., Sympathetic noradrenergic sprouting in response to centra1 cholinergic denervation: a histochemical study of neuronal sprouting in the rat hippocampal formation, Brain Res., 210 (1981) 115-12X. Crutcher, K.A., Madison, R. and Davis, J.N., A study of the rat septohippocampal pathway using anterograde transport of horseradish peroxidase, Neuroscience, 6 (1981) 1961-1973. Crutcher, K.A. and Davis, J.N., Sympathohippocampal sprouting is directed by a target tropic factor, Brain Res., 204 (1981) 410-414. Crutcher, K.A., Cholinergic denervation of rat neocortex results in sympathetic innervation, EXP. Neurol., 74 (1981) 324-329.

229 45 Crutcher, K.A. and Davis, J.N., Sympathetic noradrenergic sprouting in response to centra1 cholinergic denervation, Tre& Neurosci., 4 (1981) 70-72. 46 Crutcher, K.A. and Davis, J.N., Target regulation of sympathetic sprouting in the rat hippocampal formation, Exp. Neurol., 75 (1982) 347-359. 47 Crutcher, K.A., Neonata1 septal lesions result in sympathohippocampal innervation in the adult rat, EXP. Neurol., 76(1982) l-11. 48 Crutcher, K.A., Kesner, R.P. and Novak, J.M., Media1 septal lesions, radia1 arm maze performance, and sympathetic sprouting: a study of recovery of function, Brain Res., 262 (1983) 91-98. 49 Crutcher, K.A. and Collins, F., In vitro evidente for two distinct hippocampal growth factors: basis of neuronal plasticity?, Science, 217 (1982) 67-68. 50 Crutcher, K.A. and Chandler, J.P., Association of basa1 lamina with peripheral axons elongating within the rat centra1 nervous system, Brain Res., 308 (1984) 177-181. 51 Crutcher, K.A. and Chandler, J.P., Evidente for sprouting specificity following media1 septal lesions in the rat, J. Comp. Neurol., 237 (1985) 116-126. 52 Crutcher, K.A., The role of growth factors in neuronal development and plasticity, Cr&. Rev. Clin. Neurobiol., 2 (1986) 297-333. 53 Crutcher, K.A., Anatomica1 correlates of neuronal plasticity. In J.L. Martinez and R.P. Kesner (Eds.), Learning and Memory: A Biologica1 View, Academic, New York, 1986, pp. 83-123. 54 Crutcher, K.A. and Collins, F., Entorhinal lesions result in increased nerve growth factor-like growth-promoting activity in medium conditioned by hippocampal slices,

ry changes in contralateral sympathetic neurons of the superior cervical ganglion and in their terminals in the pineal gland following unilateral ganglionectomy, J. Neurosci., 5 (1985) 1522-1526. 64 Dreyfus, C., Peterson,

E.R. and Crain, S.M., Failure of nerve growth factor to affect fetal mouse brain stem catecholaminergic neurons in culture, Bruin Res., 194 (1980) 540-547. 65 Ebbesson, S.O.E., The parcellation theory and its relation to interspecific variability in brain organization, evolutionary and ontogenetic development, and neuronal plasticity, Ce11Tissue Res., 213 (1980) 179-212. 66 Ebendal, T., Stromberg, I., Olson, L. and Seiger, A., Parasympathetic neurotrophic activity in the rat iris: determination after different denervations, Neuroscience, 16 (1985) 181-185. 67 Ebendal, T., Olson, L., Seiger, A. and Hedlund, K.-O., Nerve growth factors in the rat iris, Nature (London), 286 (1980) 25-28. 68 Edvinsson, L. and MacKenzie, E.T., Amine mechanisms in the cerebral circulation, Pharmacof. Rev., 28 (1977) 275-348. 69 Edwards, R.H., Selby, M.J. and Rutter, W.J., Differ-

ential RNA splicing predicts two distinct nerve growth factor precursors, Nature (London), 319 (1986) 784-787. 70 Erzurumlu, R.S., Rose, G., Lynch, G.S. and Killackey, H.P., Selettive uptake and anterograde transport of horseradish peroxidase by hippocampal granule cells, Neuroscience, 6 (1981) 897-902. 71 Falck, B., Hillarp, N.A., Thieme, G. and Torp, A., Fluo-

Bruin Res., 399 (1986) 383-389.

55 Crutcher, K.A. and Marfurt, C., Ultrastructural identification of sympathohippocampal axons using anterograde transport of WGA-HRP, Soc. Neurosci. Abstr., 12 (1986) 1505. 56 Davis, J.N., Target regulation of neuronal sprouting. In F.J. Sei1 (Ed.), Nerve, Organ, and Tissue Regeneration: Research Perspecrives, Academic, New York, 1983, pp. 157-169. 57 Davis, J.N. and Martin, B., Sympathetic ingrowth in the hippocampus: evidente for regulation by mossy fibers in thyroxin treated rats, Bruin Res., 247 (1982) 145-148. 58 DeKosky, S.T., Scheff, S.W. and Cotman, C.W., Elevated corticosterone levels. A possible cause of reduced axon sprouting in aged animals, Neuroendocrinology, 38

72

73

74

75

(1984) 33-38.

59 Descarries, L. and Lapierre, Y., Noradrenergic axon terminals in the cerebral cortex of rat. 1. Radioautographic visualization after topica1 application of oL-(3H]norepinephrine, Bruin Res., 51 (1973) 141-160. 60 Descarries, L., Watkins, K.C. and Lapierre, Y., Noradrenergic axon terminals in the cerebral cortex of the rat. 111. Topometric ultrastructural analysis, Brain Res., 133

76

77

(1977) 197-222.

61 De la Torre, J.C. and Surgeon, J.W., A methodological approach to rapid and sensitive monoamine histofluorescence using a modified glyoxylic acid technique: the SPG method, Histochemisrry, 49 (1976) 81-93. 62 De la Torre, J.C., An improved approach to histofluorescence using the SPG method for tissue monoamines, J. Neurosci. Meth., 3 (1980) l-5.

63 Dornay, M., Gilad, V.H. and Gilad, G.M., Compensato-

78

79

rescence of catecholamines and related compounds condensed with formaldehyde, J. Histochem., 10 (1962) 348-354. Falck, B., Mchedlishvili, G.I. and Owman, C., Histochemica1 demonstration of adrenergic nerves in cortex-pia of rabbit,Actu Pharmacol., 23 (1965) 133-142. Finger, S. and Almli, C.R., Brain damage and neuroplasticity: mechanisms of recovery or development?, Bruin Res. Rev., 10 (1985) 177-186. Frederickson, C.J., Gage, F.H., Howell, G.A., Stewart, G.R., Kesslak, J.P., Stuart, P.R. and Klitenick, M.A., A possible role of mossy fiber zinc in sympathetic sprouting. In C.J. Frederickson, G.A. Howell and E.J. Kasarkis (Ed%), The Neurobiology of Zinc, Alan R. Liss, New York, 1984, pp. 173-187. Freed, W.J., The role of nerve growth factor (NGF) in the centra1 nervous system, Bruin Res. Bull., 1 (1976) 393-412. Freedman, R., Taylor, D.A., Seizer, A., Olson, L. and Hoffer, B.J., Seizures and related epileptiform activity in hippocampus transplanted to the anterior chamber of the eye: modulation by cholinergic and adrenergic input, Ann. Neurol., 6 (1979) 281-295. Frotscher, M. and Zimmer, J., Lesion-induced mossy fibers to the molecular layer of the rat fascia dentata: identification of the postsynaptic granule cells by the Golgi-EM technique, J. Comp. Neurol., 215 (1983) 299-311. Frotscher, M. and Leranth, C., Cholinergic innervation of the rat hippocampus as revealed by choline acetyltransferase immunocytochemistry: a combined light and electron microscopie study, J. Comp. Neurol., 239 (1985) 237-246. Frotscher, M. and Leranth, C., The cholinergic innervation of the rat fascia dentata: identification of target structures on granule cells by combining choline acetyltransfer-

230

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

ase immunocytochemistry and Golgi impregnation, J. Comp. Neurol., 243 (1986) 58-70. Furukawa, S., Kamo, I., Furukawa, Y., Akazawa, S., Satoyoshi, E., Itoh, K. and Hayashi, K., A highly sensitive enzyme immunoassay for mouse nerve growth factor. J. Neurochem., 40 (1983) 734-744. Gaarskjaer, F.B., Danscher. G. and West, M.J., Hippocampa1 mossy fibers in the regio superior of the European hedgehog, Br& Res., 237 (1982) 79-90. Gage, F.H., Stenevi, U., Moes, P. and Bjorklund, A., Anatomicai, biochemical, and behavioral factors related to synaptogenesis of the superior cervical ganglion following damage to the centra1 nervous system. In E. Endroczi (Ed.), Integrative Neurohumoral Mechanisms: Physiologica1 and Clinica1 Aspects, Elsevier, Amsterdam, 1982. pp. 51-62. Gage, F.H., Bjorklund, A. and Stenevi, U.. Denervation releases a neuronal survival factor in adult rat hippocampus, Nature (London), 308 (1984) 637-639. Gage, F.H., Bjorklund, A. and Stenevi, U., Cells of origin of the ventral cholinergic septohippocampal pathway undergoing compensatory collateral sprouting following fimbria-fornix transection, Neurosci. Lett., 44 (1984) 211-216. Gage, F.H. and Bjorklund, A., Enhanced graft survival in the hippocampus following selettive denervation. Neuroscience, 17 (1986) 89-98. Gasser, U.E., Van Deusen, E.B. and Dravid,A.R., Homologous cholinergic efferents spared by partial fimbrial lesions contribute to the recovery of hippocampal cholinergic enzymes in adult rats, Brain Res., 367 (1986) 368-373. Geddes. J.W., Monaghan, D.T., Cotman, C.W., Lott. I.R., Kim, R.C. and Chui, H.C., Plasticity of hippocampal circuitry in Alzheimer’s disease, Science, 230 (1985) 1179-1181. Geneser-Jensen, F.A., Distribution of acetylcholinesterase in the hippocampal region of the guinea pig. 11. Subiculum and hippocampus, Z. Zellforsch., 124 (1972) 546-560. Gibby. W.A and Crutcher, K.A., Absence of sympathetic sprouting in the rat olfactory bulb after cholinergic denervation, EXP. Neuroi., 84 (1984) 386-395. Goedert, M., Fine, A., Hunt, S.P. and Ullrich. A., Nerve growth mRNA in peripheral and centra1 rat tissues and in the human centra1 nervous system: lesion effects in the rat brain and levels in Alzheimer’s disease, Mal. Brain Res., 1 (1986) 85-92. Gottesfeld, Z. and Jacobowitz, D.M., Cholinergic projections from the septal-diagonal band area to the habenular nuclei, Brain Res., 176 (1979) 391-394. Gottesfeld, Z., Sprouting of noradrenergic terminals in the partially deafferented habenula, Bruin Res., 191 (1980) 559-563. Handelmann, G.E., Olten, D.S., O’Donohue, T.L.. Beinfeld, M.C., Jacobowitz, D.M. and Cummins, C.J., Effects of time and experience on hippocampal neurochemistry after damage to the CA, subfield. Pharmacol. Biochem. Behav., 18 (1983) 551-561. Harrell, L.E. and Davis. J.N., Sympathetic sprouting and recovery of a spatial behavior, EXP. Neurol., 82 (1983) 379-390. Harrell, L.E. and Davis, J.N., Cholinergic denervation of the hippocampal formation does not produce long-term

changes in glucose metabolism, EXP. Neuro/.. 85 (1984) 128-138. 96 Harrell, L.E. and Davis, J.N., Cholinergic influences on hippocampal glucose metabolism. Neuroscience, 15 (1985) 359-369. 97 Harrell, L.E.. Haring, J.H. and Davis, J.N., Peripheral sympathetic ingrowth can alter metabolic activity within the hippocampal formation, EXP. Neurol., 91 (1986) 622-627. 98 Heacock, A.M., Schonfeld. A.R. and Katzman. R.. Hippocampal neurotrophic factor: characterization and response to denervation. Brain Res., 363 (1986) 299-306. 99 Hefti. F.. IS Alzheimer’s disease caused by lack of nervr growth factor?, Ann. Neurol., 13 (1983) 110. 100 Hefti, F., Dravid. A. and Hartikka. J., Chronic intraventricular injections of nerve growth factor elevate hipp»campa1 choline acetyltransferase activity in adult rats with partial septo-hippocampal lesions, Brain Res., 293 (lY84) 305-311. 101 Hefti, F., Nerve growth factor promotes survival of septal cholinergic neurons after fimbrial transection. J. Neurosci.. 6 (1986) 2155-2162. 102 Hefti, F., Hartikka. J., Salvatierra, A.. Weiner, W.J. and Mash, D.C., Localization of nerve growth factor receptors in cholinergic neurons of the human basa1 forebrain, Neurosci. Lett., 69 (1986) 37-4 1. 103 Hefti, F. and Weiner, W.J.. Nerve growth factor and Alzheimer’s disease, Atm. Nrurol., 20 (1986). 104 Hefti. F.. Hartikka. J.. Eckenatein. F.. Gnahn, H., Heumann, R. and Schwab. M.. Nerve growth factor increases choline acetyltransferase but not survival or fiber outgrowth of cultured fetal septal cholinergic neurons. Neuroscience. 14 (1985) SS-68. 105 Hendry. I.A., Developmental changes in tissue and plasma concentrations of the biologically attive species of nerve growth factor in the mouse. by using a two-site radioimmunoassay, Biochem J.. 128 (1972) 1265- 1272. 106 Honegger, P. and Lenoir. D., Nerve growth factor (NGF) stimulation of cholinergic telencephalic neurons in aggregating cell cultures, Dev. Brain Res., 3 (1982) 229-238. 107 Howard. A.J. and Barker. D.J., Functional significante of sympathetic fiber ingrowth in the habenula. Soc. Neurosci. Ahstr., 12 (1986) 511. 108 Johnson. D.G.. Gorden. P. and Kopin, I.J.. A sensitive radioimmunoassay for 7S nerve growth factor antigens in serum and tissues. 1. Neurochem.. 18 (1971) 2355-2362. 109 Johnston, M.V., McKinney. M. and Coyle, J.T., Cholinergic projection to ncocortex from neurons in basa1 forebrain, Proc. Natl. Acad. Sci. U.S.A., 76 (1979) 5392-5398. 110 Jones, B.E.. Relationship between catecholamine neurons and cerebral blood vessels studied by their simultaneous fluorescent revelation in the rat brainstem, Brain Res. Bull., 9 (1982) 33-44. 111 Kappers, J.A., The development. topographical relations and innervation of the epiphysis cerebri in the albino rat, Z. Zellforsch., 52 (1960) 163-215. 112 Katz, M.J.. Lasek, R.J. and Kaiserman-Abramof. I.R.. Ontophyletics of the nervous system: eyeless mutants illustrate how ontogenetic buffer mechanisms channel evolution. Proc. Nati. Acad. Sci. U.S.A., 78 (1981) 397-401. 113 Katzman. R.. Bjorklund. A.. Owman, C., Stenevi, U. and West. K.A., Evidente for regenerative axon sprouting of centra1 catecholamine neurons in the rat mesencephalon

231 following electrolytic lesions, Brain Res., 25 (1971) 579-596. 114 Kelly, P.A. and Moore, K.E., Decrease of neocortical choline acetyltransferase after lesion of the globus pallidus in the rat, EXP. Neurol., 12 (1978) 479-484. 115 Kelly, P.A.T., Gage, F.H., Ingvar, M., Lindvall, O., Stenevi, U. and Bjorklund, A., Functional reactivation of the deafferented hippocampus by embryonic septal grafts as assessed by measurements of loca1 glucose utilization, Exp. Brain Res., 58 (1985) 570-579. 116 Kesslak, J.P. and Gage, F.H., Recovery of spatial alternation deficits following selettive hippocampal destruction with kainic acid, Behav. Neurosci., 100 (1986) 280-283. 117 Kesslak, J.P., Frederickson, C.J. and Gage, F.H., Quantification of hippocampal noradrenaline and zinc changes after selettive ce11destruction, Exp. Br& Res., in Press. 118 Kessler, J.A., Parasympathetic, sympathetic, and sensory interactions in the iris: nerve growth factor regulates cholinergic ciliary ganglion innervation in vivo, J. Neurosci., 5 (1985) 2719-2725. 119 Kiernan, J.A., Hypotheses concerned with axonal regeneration in the mammalian nervous system, Biel. Rev., 54 (1979) 155-197. 120 Kimble, D.P., Anderson, S., Bremiller, R. and Dannen, E., Hippocampal lesions, superior cervical ganglia removal, and behavior in rats, Physiol. Behav., 22 (1979) 461-466. 121 Kimble, D.P., Bremiller, R. and Schroeder, L., Hippocampa1 lesions slow extinction of a conditioned taste aversion in rats, Physiol. Behav., 23 (1979) 217-222. 122 Kimble, D.P., Bremiller, R. and Perez-Polo, J.R., Nerve growth factor applications fail to alter behavior of hippocampa1 lesioned rats, Physiol. Rev., 23 (1979) 653-657. 123 Kimble, D.P., Bremiller, R. and Stickford, G., Failure to find a behavioral role for anomalous sympathetic innervation of the hippocampus in male rats, Physiol. Behav., 25 (1980) 675-681. 124 Korsching, S., Auburger, G., Heumann, R., Scott, J. and Thoenen, H., Levels of nerve growth factor and its mRNA in the centra1 nervous system of the rat correlate with chohnergic innervation, EMBO J., 4 (1985) 1389-1393. 125 Korsching, S., Heumann, R., Thoenen, H. and Hefti, F., Cholinergic denervation of the rat hippocampus by fimbrial transection leads to a transient accumulation of nerve growth factor (NGF) without change in mRNA NGF content, Neurosci. Lett., 66 (1986) 175-180. 126 Kromer, L.F., Cholinergic axons from delayed septal implants and sympathetic fibers co-exist in the denervated dentate gyrus, Bruin Res. Bull., 9 (1982) 539-544. 127 Langley, J.N., On the regeneration of pre-ganglionic and of Post-ganglionic viscera1 nerve fibers, J. Physiol. (London), 22 (1897) 215-230. 128 Leblanc, G. and Landis, S., Development of choline acetyltransferase (CAT) in the sympathetic innervation of sweat glands, J. Neurosci., 6 (1986) 260-265. 129 Lehman, J., Nagy, S., Atmadja, S. and Fibiger, H.C., The nucleus basalis magnocellularis: the origin of a cholinergic projection to the neocortex of the rat, Neuroscience, 5 (1980) 1161-1174. 130 Lindsay, R.M., Adult rat brain astrocytes support survival of both NGF-dependent and NGF-insensitive neurons, Nature (London), 282 (1979) 80-82. 131 Lindvall, 0. and Bjorklund, A., The glyoxylic acid fluo-

132

133

134

135

136

137

138

139

rescence histochemical method: a detailed account of the methodology for visualization of centra1 catecholamine neurons, Htktochemistry, 39 (1974) 97- 127. Lindvall, M., Edvinsson, L. and Owman, C., Sympathetic nervous contro1 of cerebrospinal fluid production from the choroid plexus, Science, 201 (1978) 176-178. Loy, R. and Moore, R.Y., Anomalous innervation of the hippocampal formation by peripheral sympathetic axons following mechanical injury, EXP. Neurol., 57 (1977) 645-650. Loy, R. and Moore, R.Y., Regenerative growth of sympathetic fibers into rat hippocampus. In E. Usdin, E.J. Kopin and J. Barchas (Eds.), Cutecholumines: Busic und Clinica1 Frontiers, Pergamon, New York, 1979, pp. 1336-1338. Loy, R., Milner, T.A. and Moore, R.Y., Sprouting of sympathetic axons in the hippocampal formation: conditions necessary to elicit ingrowth, EXP. Neurol., 67 (1980) 399-411. Loy, R. and Milner, T.A., Sexual dimorphism in extent of axonal sprouting in hippocampus, Science, 208 (1980) 1282-1284. Loy, R., Morrow, A.L. and Creese, I., Cholinergic denervation induces adrenergic receptor increases: correlation with sympathetic axon sprouting in hippocampus. In B. Haber, J.R. Perez-Polo, G.A. Hashim and A.M.G. Stella (Eds.), Nervous System Regenerution, Alari R. Liss, New York, 1983, pp. 409-415. Loy, R. and Milner, T.A., Neonata1 steroid treatment alters axonal sprouting in adult hippocampus: sexually dimorphic development of cholinergic target neurons. In B. Haber, J.R. Perez-Polo, G.A. Hashim and A.M.G. Stella (Eds.), Nervous System Regeneration, Alan R. Liss, New York, 1983, pp. 417-423. Lundberg, J.M., Hokfelt, T., Anggard, A., Terenius, L., Elde, R., Markey, K., Goldstein, M. and Kimmel, J., Organizational principles in the peripheral sympathetic nervous system: subdivision by coexisting peptides (somatostatin-, avian pancreatic polypeptide-, and vasoactive intestina] polypeptide-like immunoreactive materials), Proc. Nutl. Acud. Sci. U.S.A.,

140

79 (1982) 1303-1307.

Lynch, G., Matthews, D.A., Mosko, S., Parks, T. and Cotman, C., Induced acetylcholinesterase-rich layer in the rat dentate gyrus following entorhinal lesions, Brain

Res., 42 (1972) 311-318. 141 Madison, R., Crutcher, K.A. and Davis, J.N., Sympatho-

hippocampal neurons are inside the blood-brain

barrier,

Bruin Res., 213 (1981) 183-189. 142 Madison, R. and Davis, J.N., Regulation of hippocampal sympathetic ingrowth: role of afferent input, Bruin Res., 270 (1983) l-9. 143 Madison, R. and Davis, J.N., Sprouting of noradrenergic

fibers in hippocampus after media1 septal lesions: contributions of the centra1 and peripheral nervous systems, EXP. Neurol., 80 (1983) 167-177. 144 McGinty, J., Milner, T.A. and Loy, R., Association of sympathetic axons in denervated hippocampus to intracerebral vasculature. 1. Fluorescente histochemistry combining glyoxylic acid and pontamine sky blue, Anat. Embryol., 164 (1982) 95-100. 145 McNicholas, L.F., Martin, W.R., Sloan, J.W. and Nozaki, M., Innervation of the spina1 cord by sympathetic fibers, EXP. Neurol., 69 (1980) 383-394. 146 Menesini-Chen, M.G., Chen, J.S. and Levi-Montalcini,

232

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

R.. Sympathetic nerve fibers ingrowth in the centra1 nervous system of neonata1 rodents upon intracerebral NGF injections, Arch. Ital. Biel., 116 (1978) 53-84. Milner, T.A. and Loy, R., A delayed sprouting response to partial hippocampal deafferentation: time course of sympathetic ingrowth following fimbrial lesions, Brain RU., 197 (1980) 391-399. Milner, T.A. and Loy. R., Interaction of age and sex in sympathetic axon ingrowth into the hippocampus following septal afferent damage. Anat. Embryol., 161 (1980) 159-168. Milner, T.A. and Loy, R., Hormonal regulation of axonal sprouting in the hippocampus, Brain Res., 243 (1982) 180-185. Mobley. W.C., Rutkowski, J.L., Tennekoon, G.I., Buchanan, K. and Johnston, M.V., Choline acetyltransferase activity in striatum of neonata1 rats increased by nerve growth factor, Science, 229 (1985) 284-287. Mobley, W.C., Rutkowski, J.L., Tennekoon, G.I., Gemski, J., Buchanan. K. and Johnston, M.V., Nerve growth factor increases choline acetyltransferase activity in developing basa1 forebrain neurons, Mo/. Brain Res., 1 (1986) 53-62. Moore, R.Y. and Bloom. F.E., Centra1 catecholamine neuron systems: anatomy and physiology of the norepinephrine and epinephrine systems, Annu. Rev. Neurosci., 2 (1979) 113-168. Morrow, A.L., Loy. R. and Creese, I., Septal deafferentation increases hippocampal adrenergic receptors: correlation with sympathetic axon sprouting, Proc. Nati. Acad. Sci. U.S.A., 80 (1983) 6718-6722. Morrow, A.L., Norman. A.B., Battaglia, G., Loy, R. and Creese, I., Up-regulation of serotonergic binding sites labeled by [‘H]WB4101 following fimbrial transection and 5,7-dihydroxytryptamine-induced lesions, Life Sci., 37 (1985) 1913-1922. Morrow, A.L.. Loy, R. and Creese, I., Alteration of nicotinic cholinergic agonist binding sites in hippocampus after fimbria transection, Brain Res., 334 (1985) 309-314. Nadler, J .V., Shelton, D.L. and Cotman, C.W., Lesioninduced plasticity of high affinity choline uptake in the developing rat fascia dentata, Brain Res., 42 (1972) 311-318. Nadler, J.V., Cotman, C.W. and Lynch, G.S., Altered distribution of choline acetyltransferase and acetylcholinesterase activities in the developing rat dentate gyrus following entorhinal lesion, Brain Res., 63 (1973) 215-230. Nadler, J.V., Cotman, C.W. and Lynch, G.S., Histochemica1 evidente of altered development of cholinergic fibers in the rat dentate gyrus following lesions. 1. Time course after complete entorhinal lesion at various ages, J. Comp. Neurol., 171(1977) 561-588. Needels, D.L.. Nieto-Sampedro, M., Whittemore, S.R. and Cotman, C.W., Neuronotrophic activity for ciliary ganglion neurons. Induction following injury to the brain of neonatal, adult. and aged rat, Dev. Brain Res.. 18 (1985) 275-284. Neild, T.O. and Zelcer, E., Noradrenergic neuromuscular transmission with special referente to arteria1 smooth muscle, Prog. Neurobiol., 19 (1982) 141-158. Nielsen, K.C. and Owman, Ch., Adrenergic innervation of pial arteries related to the circle of Willis in the cat, Brain Res., 6 (1967) 773-776. Nieto-Sampedro, M., Lewis, E.R., Cotman, C.W., Man-

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

thorpe, M., Skaper, S.D., Barbin, G., Longo, F.M. and Varon, S., Brain injury causes a time-dependent increase in neuronotrophic activity at the lesion site, Science, 217 (1982) 860-861. Nieto-Sampedro, M., Manthorpe, M., Barbin, G., Varon, S. and Cotman, C.W.. Injury-induced neuronotrophic activity in adult brain. Correlation with survival of delayed implants in the wound cavity, J. Neurosci., 3 (1983) 2219-2229. Nieto-Sampedro, M., Whittemore, SR., Needels, D.L., Larsen, J. and Cotman, C.W., The survival of brain transplants is enhanced by extracts of injured brain, Proc. Natl. Acad. Sci. U.S.A., 81 (1984) 6250-6254. Ojika, K. and Appel, S.H., Neurotrophic effects of hippocampa1 extracts on medial septal nucleus in vitro, Proc. Natl. Acad. Sci. U.S.A., Sl(l984) 2567-2571. Olson, L. and Malmfors, T., Growth characteristics of adrenergic nerves in the adult. Fluorescente histochemical and 3H-noradrenaline uptake studies using tissue transplantations to the anterior chamber on the eye, Atta Physiol. Stand., Suppl. 348 (1970) l-112. Olson, L., Ebendal, T. and Seiger, A., NGF and antiNGF: evidente against effects on fiber growth in locus coeruleus from cultures of perinatal CNS tissue, Dev. Neurosci., 2 (1979) 160-176. Otten, U., Weskamp, G., Keller, H.H. and Lorez, H.P.. Lesion of cholinergic but not aminergic innervation increases nerve growth factor content in adult rat hippocampus, Soc. Neurosci. Abstr.. 12 (1986) 587. Patterson, P.H., Environmental determination of autonomic neurotransmitter functions, Annu. Rev. Neurosci., 1(1978) l-17. Peterson, G.M. and Loy, R., Sprouting of sympathetic fibers in the hippocampus in the absence of major target cell candidates, Brain Res., 264 (1983) 21-29. Rennels, M.L., Gregory, T.F., Blaumanis, O.R., Fujimoto, K. and Grady, P.A.. Evidente for a ‘paravascular’ fluid circulation in the mammalian centra1 nervous system, provided by the rapid distribution of tracer protein throughout the brain from subarachnoid space, Brain Res., 326 (1985) 47-63. Rennert, P.D. and Heinrich, G., Nerve growth factor messenger RNA in brain localization by in situ hybridization, Biochem. Biophys. Res. Commun., 138 (1986) 813-818. Rosenberg, M.B., Grossman, M.H. and Breakefield, X.0., Artifactual presente of nerve growth factor in adult mouse brain, 295 (1984) 35-40. Rush, R.A.. Immunohistochemical localization of endogenous nerve growth factor, Nature (London), 312 (1984) 364-367. Saffran, B.N. and Crutcher, K.A., Exogenous NGF enhances but does not elicit sympathohippocampal sprouting, in preparation. Scheff, S.W., Bernardo, L.S. and Cotman. C.W., Decrease in adrenergic axon sprouting in the senescent rat, Science, 202 (1978) 775-778. Scheff, S. W., Anderson. K. and DeKosky, S.T., Morphologica1 aspects of brain damage in aging. In W.W. Scheff (Ed.), Aging and Recovery of Function in the Centra1 Nervous System, Plenum, New York, 1984, pp. 57-85. Schonfeld, A.R., Heacock, A.M. and Katzman, R., Enhancement of centra1 cholinergic sprouting by prior injury: correlation with endogenous trophic content of hip-

233 pocampus, Brain Res., 321 (1984) 377-380. 179 Schonfeld, A.R., Heacock, A.M. and Katzman, R., Neurotrophic factors: effects on centrai cholinergic regeneration in vivo, Bruin Rex, 336 (1985) 297-301. 180 Schwab, M.E., Otten, U., Agid, Y. and Thoenen, H., Nerve growth factor (NGF) in the rat CNS: absence of specific retrograde axonal transport and tyrosine hydroxylase induction in locus coeruleus and substantia nigra, Brain Rex, 168 (1979) 473-483. 181 Schwab, M.E. and Thoenen, H., Dissociated neurons regenerate into sciatic but not optic nerve explants in culture irrespective of neurotrophic factors, J. Neurosci., 5 (1985) 2415-2423. 182 Scott, S.M., Tarris, R., Eveleth, D., Mansfield, H., Weichsel Jr., M.E. and Fisher, D.A., Bioassay detection of mouse nerve growth factor (mNGF) in the brain of adult mice, J. Neurosci. Res., 6 (1981) 653-658. 183 Seiler, M. and Schwab, M.E., Specific retrograde transport of nerve growth factor (NGF) from neocortex to nucleus basalis in the rat, Br& Res., 300 (1984) 33-39. 184 Shelton, D.L. and Reichardt, L.F., Expression of the /3nerve growth factor gene correlates with the density of sympathetic innervation in effector organs, Proc. Nari. Acad. Sci. U.S.A., 81 (1984) 7951-7955. 185 Shelton, D.L. and Reichardt, L.F., Studies on the expression of the beta nerve growth factor (NGF) gene in the centra1 nervous system: leve1 and regional distribution of NGF mRNA suggest that NGF functions as a trophic factor for severa1 distinct populations of neurons, Proc. Nurl. Acad. Sci. U.S.A., 83 (1986) 2714-2718. 186 Shine, H.D. and Perez-Polo, R.G., 7S nerve growth factor protein in the golden hamster, J. Neurochem., 27 (1976) 1315-1318. 187 Springer, J.E. and Loy, R., Intrahippocampal injections of antiserum to nerve growth factor inhibit sympathohippocampal sprouting, Brain Res. Bull., 15 (1985) 629-634. 188 Stanfield, B.B. and Cowan, W.M., The sprouting of septal afferents to the dentate gyrus after lesions of the entorhinal cortex in adult rats, Bruin Res., 232 (1982) 162-170. 189 Stenevi, U. and Bjorklund, A., Growth of vascular sympathetic axons into the hippocampus after lesions of the septo-hippocampal pathway: a pitfall in brain lesion studies, Neurosci. Letr., 7 (1978) 219-224. 190 Steward, O., Assessing the functional significante of lesion-induced neuronal plasticity, Int. Rev. Neurobiol., 23 (1982) 197-253. 191 Stewart, G.R., Frederickson, C.J., Howell, G.A. and Gage, F.H., Cholinergic denervation-induced increase of chelatable zinc in mossy-fiber region of the hippocampal formation, Brain Res., 290 (1984) 43-51. 192 Storm-Mathisen, J., Choline acetyltransferase and acetylcholinesterase in fascia dentata following lesion of the entorhinal afferents, Brain Res., 80 (1974) 181-197. 193 Suda, K., Barde, Y.A. and Thoenen, H., Nerve growth factor in mouse and rat serum: correlation between bioassay and radioimmunoassay determinations, Proc. Nari.

Acad. Sci. U.S.A., 75 (1978) 4042-4046. 194 Taniuchi, M., Schweitzer, J.B. and Johnson Jr., E.M., Nerve growth factor receptor molecules in rat brain, Proc. Natl. Acad. Sci. U.S.A., 83 (1986) 1950-1954. 195 Taniuchi, M., Clark, H.B. and Johnson Jr., E.M., Induction of nerve growth factor receptor in Schwann cells in vivo after axotomy of peripheral nerves: possible implications for nerve regeneration, Soc. Neurosci. Abstr., 12 (1986) 392. 196 Toniolo, G., Dunnett, S.B., Hefti, F. and Will, B., Acetylcholine-rich transplants in the hippocampus: influente of intrinsic growth factors and application of nerve growth factor on choline acetyltransferase activity, Brain Res., 345 (1985) 141-146. 197 Wainer, B.H., Levey, AI., Rye, D.B., Mesulam, M.-M. and Mufson, E.J., Cholinergic and non-cholinergic septohippocampal pathways, Neurosci. Lem., 54 (1985) 45-52. 198 Walker, P., Weichsel Jr., M.E., Fisher, D.A., Guo, S.M. and Fisher, D.A., Thyroxine increases nerve growth factor concentration in adult mouse brain, Science, 204 (1979) 427-429. 199 Wamsley, J.K. and Palacios, J.M., Receptor mapping by histochemistry. In A. Lajtha (Ed.), Handbook of Neurochemimy, Vol. 2, Plenum, New York, 1983, p. 93. 200 Whitehouse, P.J., Price, D.L., Struble, R.G., Clark, A.W., Coyle, J.T. and Delong, M.R., Alzheimer’s disease and senile dementia: loss of neurons in the basa1 forebrain, Science, 215 (1982) 1237-1239. 201 Whittemore, SR., Ebendal, T., Larkfors, L., Olson, L., Seiger, A., Stromberg, 1. and Persson, H., Development and regional expression of beta nerve growth factor messenger RNA and protein in the rat centra1 nervous system, Proc. Nati Acad. Sci. U.S.A., 83 (1986) 817-821. 202 Whittemore, S.R., Larkfors, L., Ebendal, T., Holets, V.R., Ericsson, A. and Persson, H., Increased beta-nerve growth factor messenger RNA and protein levels in neonata1 rat hippocampus following specific cholinergic lesions, J. Neurosci., 7 (1987) 244-251. 203 Wiklund, L., Development of serotonin-containing cells and the sympathetic innervation of the habenular region in the rat brain, Ce11 Tissue Res., 155 (1974) 231-243. 204 Will, B. and Hefti, F., Behavioral and neurochemical effects of chronic intraventricular injections of nerve growth factor in adult rats with fimbria lesions, Behav. Brain Res., 17 (1985) 17-24. 205 Williams, L.R., Varon, S., Peterson,

G.M., Wictorin, K., Fischer, W., Bjorklund, A. and Gage, F.H., Continuous infusion of nerve growth factor prevents basa1 forebrain neuronal death after fimbria-fornix transection, Proc.

Natl. Acad. Sci. U.S.A., 83 (1986) 9231-9235. 206 Yoshida, K., Kohsaka, S., Idei, T., Otani, M., Toya, S.

and Tsukada, Y., Septal deafferentation enhances the neurotrophic effects of rat hippocampus on cultured neurai cells from the centra1 nervous system, Neurosci. Lett., 66 (1986) 181-186.