Neurochemical anatomy of fetal hippocampus transplanted into large lesion cavities made in the adult rat brain

EXPERIMENTAL

NEUROLOGY

(1991)

111,36-48

Neurochemical Anatomy of Fetal Hippocampus Transplanted into Large Lesion Cavities Made in the Adult Rat Brain SUSAN WRAY,* *Laboratory

of

Neurochemistry,

RONALD H. BAISDEN,~

as nerve growth guide cholinergic

Medicine,

factor is present in the transplants innervation. 8 1991 Academic press,

to I~~.

Hippocampal anlage survive transplantation into cavities produced by bilateral ablation of the hippocampus and, once developed, ameliorate hippocampal lesion-induced deficits in learning set formation in the Rabinovitch-Rosvold series of mazes (22), acquisition of an operant DRL schedule (42), and acquisition of passive avoidance (43). Additionally, hippocampal transplants decrease open field hyperactivity produced by hippocampal damage (43). The results of these experiments involving the hippocampus are compatible with reports indicating that transplantation of fetal brain tissue can compensate behavioral deficits associated with lesions of several different brain structures (3, 7,9-11, 16,37) and support the proposal that transplantation of fetal brain tissue may ultimately be useful in promoting recovery from neurological impairments in humans. However, the results of an experiment by Woodruff et al. (41) indicate that behavioral recovery of function may not be a ubiquitous consequence of transplantation of fetal hippocampus. In this experiment hippocampal grafts made into the lesion cavities actually produced a greater impairment in one measure of acquisition of the Morris water maze than did hippocampal lesions alone. These results contrast with the reports noted above concerning hippocampal graft-induced behavioral recovery in rats with hippocampal lesions and serve to emphasize the need for development of an understanding of the mechanisms whereby neural transplants influence behavior. One approach to establishing a basis for such an understanding is to describe the anatomy, including the chemoarchitecture, of the grafted tissue. Portions of block transplants of fetal hippocampus made into lesion cavities produced in the host rat brain by aspiration of the dorsal parietal cortex and hippocampus generally develop a recognizable, albeit distorted, organotypic structure (26,40,42) and also form connections with the host brain (40-42). Further, observations by Zimmer 36

Inc. -...-~.J

of

INTRODUCTION

1 To whom reprint requests should be addressed at Department of Anatomy, P.O. Box 19960A, James H. Quillen College of Medicine, East Tennessee State University, Johnson City, TN 37614.

$3.00 0 1991 by Academic Press, ,F-...-^Ae^A:^I- ---. c_--

L. WOODRUFFS+’

NfND.9, Bethesda, Maryland 20892; and TDepartment of Anatomy, James H. Quillen College East Tennessee State University, Johnson City, Tennessee 37614

The purpose of the present study was to determine whether neurochemicals normally found within neuron somata, fibers, and terminals of the hippocampal formation would also be present in transplanted hippocampal tissue that had developed in lesion cavities made in adult rat brains by aspiration of the hippocampus and overlying dorsolateral neocortex. Embryonic Day 15 or 16 rat brain tissue containing hippocampus with some medial pallial anlage was transplanted into the site of hippocampal aspiration lesions in adult male rats. One hundred ten to one hundred thirty-five days later the brains of these rats were sectioned and processed using the avidin-biotin-horseradish peroxidase immunocytochemical procedure to visualize choline acetyltransferase, met-enkephalin (MENK), neurotensin (NT), somatostatin, substance P, tyrosine hydroxylase (TH), or vasoactive intestinal polypeptide. Sections from two brains were stained using the thiocholine technique for visualization of acetylcholinesterase. All of these substances were found within cell bodies andlor fibers in the transplants. However, several abnormalities were noted. In addition to TH-immunoreactive fibers, TH-immunoreactive cell bodies were found in the transplants. Since TH is not expressed in mature hippocampal or cortical neurons this suggests that mechanisms for suppression of manufacture of this enzyme are lacking or inhibited in the transplants. Further, although all of the peptides were present either in fibers or in both cell bodies and fibers, the density of staining for NT and MENK was less than would be expected for normal hippocampus, and none of the cell bodies or fibers reacting for the peptides exhibited any apparent organization resembling that normally observed in hippocampus or cortex. However, some histological organization was present and the cholinergic markers were associated with this organization. These data suggest that some tropic and/or trophic factor such

0014-4886/91 Copyright *,, r;“LC”

AND MICHAEL

NEUROCHEMICAL

ANATOMY

OF

TRANSPLANTED

and Sunde (46) indicate that neurons containing the peptides somatostatin, cholecystokinin, and met-enkephalin are present in grafts of Ammon’s horn or dentate gyrus. The procedure used by Zimmer and Sunde (46) to transplant fetal brain tissue to the adult rat brain differs from that employed in the experiments in which the behavioral effects of fetal hippocampal tissue have been investigated. Zimmer and Sunde made their grafts by stereotaxic injection into the otherwise undamaged host brain. This permitted exact placement of the graft in a location where it would remain and develop. Kimble et al. (22) and Woodruff et al. (41-43) placed their transplants into large lesion cavities that provided access to the ventricular system. The transplanted tissue was then free to move in a comparatively large space and the mature grafts fused with several host brain regions including the residual host hippocampus, the fimbria-fornix, the septal area, the neostriatum, the dorsal thalamus, and the neocortex. For this reason the pattern and even the type of neurotransmitters contained within the fibers and cells of the mature grafts might differ from those observed by Zimmer and Sunde. Histochemical and immunocytochemical markers were used to visualize cell bodies and fibers containing acetylcholine (acetylcholinesterase; choline acetyltransferase), dopamine, norepinephrine, epinephrine (tyrosine hydroxylase), vasoactive intestinal polypeptide, somatostatin, neurotensin, substance P, and metenkephalin. Visualization of these substances was done to determine whether some of the specific neurotransmitters and neuromodulators known to be contained in extrinsic projections into the hippocampus, or normally observed within putative hippocampal interneurons, would also appear in hippocampal transplants placed into lesion cavities produced by aspiration. An attempt was also made to determine whether these cells and fibers formed anatomical relationships similar to those normally found in the hippocampus of the rat. A brief review of these relationships is presented in the following paragraphs. The highest concentration of cholinergic fibers is found around the pyramidal cell bodies of Ammon’s horn and the granule cells of the dentate gyrus, and in the strata oriens and radiatum of Ammon’s horn and the inner third of the molecular layer of the dentate (13). Additionally, cholinergic neurons are found throughout the hippocampus with some tendency to aggregate in the molecular layer of Ammon’s horn and the subiculum (14). On the other hand, tyrosine hydroxylase immunoreactive cell bodies are not normally found in adult hippocampus or cortex, but dopaminergic neurons of the ventral tegmental area do provide a relatively minor innervation to the hippocampal formation including the subicular complex (39) and the locus coeruleus provides significant innervation to the hippocam-

FETAL

HIPPOCAMPUS

37

pus. The noradrenergic innervation is generally diffuse, but there is a relative concentration of noradrenergic fibers in the dentate hilus and the stratum lucidum of subfield CA3 of Ammon’s horn (31). Many of the peptides found in the hippocampus have been associated with interneurons. Neurons showing immunoreactivity for vasoactive intestinal polypeptide (VIP-ir) are a good example. These neurons are found in the several layers of the hippocampal formation, but are concentrated within the pyramidal cell layer and stratum lacunosum-molecular of Ammon’s horn and the molecular and granule cell layers of the dentate gyrus (24,29). The size and shape of the VIP-ir neurons vary. Within the dentate gyrus some of these cells appear to correspond to the pyramidal basket cells and others are similar to granule cells and displaced granule cells (24). In Ammon’s horn both relatively large multipolar and smaller fusiform and round VIP-ir cells are found (24, 29). The distribution of somatostatin-immunoreactive cells (SS-ir) differs from that of VIP-ir cells. SS-ir cells are concentrated in the stratum oriens of Ammon’s horn and the hilus of the dentate gyrus (25,32,34), but some are also found within the pyramidal cell layer of CA1 and in stratum radiatum of CA3. The shape of the SS-ir cells varies in the different regions of the hippocampus. Not surprisingly those found in the dentate hilus are often polymorphic and some contribute to the hippocampal commissural projections. SS-ir cells in Ammon’s horn vary in shape from fusiform to multipolar. Some of the CA1 SS-ir neurons even resemble pyramidal cells, although there is no evidence that they project out of the hippocampus. The primary distribution of cells showing substance P immunoreactivity (SP-ir) overlaps that of the SS-ir cells. SP-ir cell bodies are concentrated in the stratum oriens of Ammon’s horn and in the dentate hilus, where they appear as medium- to large-sized multipolar neurons. Small- to medium-sized round and fusiform SP-ir neurons are also found in the deeper layers of the ventral subiculum (8,36). SP-ir fibers and terminals appear to be scattered throughout all strata of Ammon’s horn, with some concentration in the radiatum and molecular layers (8). SP-ir puncta, presumably representing axon terminals, are also associated with nonreactive neuron somata in all regions of the hippocampal formation, including the subiculum and entorhinal cortex (8). Neurotensin (NT) and methionine-enkephalin (MENK), the final two substances visualized in the present study, have a more restricted distribution within the hippocampal formation than do the other peptides. Cell bodies positive for NT are only found within the subiculum (19, 34). Some of these may project through the postcommisural fornix to the hypothalamus (19). Roberts et al. (34) have confirmed the presence of NT-ir fibers in the fimbria-fornix and also observed moderate

38

WRAY,

BAISDEN,

densities of NT-ir fibers around pyramidal cell bodies in CA213 and the lateral part of CAl. MENK-ir has been observed to overlap the projection zones of the mossy

AND

WOODRUFF

fibers (dentate hilus and subfield CA3) and the lateral perforant pathway (outer molecular layer of the dentate gyrus), and MENK-ir cell bodies have been observed

FIG. 1. (A) Thionin-stained section through the approximate rostral-caudal midpoint of a hippocampal lesion representative of those made in this study. The most dorsomedial part of the structure primarily composed of the dentate gyrus (d), CA3c subfield (ca) of Ammon’s horn, and dorsal subiculum (s) is spared as is the most temporal portion (at tip of arrowhead). Bar = 420 pm. (B) This photomicrograph is of a neurotensin-stained, methyl green-counterstained section from the approximate rostral-caudal midpoint of a transplant. The transplant tissue has grown to fill the lesion cavity and interfaces with the damaged parenchyma of the host cortex (c) and hippocampus (h) along the boundaries indicated by the larger black arrowheads. The methyl green-stained neurons present three patterns of arrangement. The smaller black arrowheads surround a part of the transplant that resembles Ammon’s horn. The open arrowheads indicate a cluster of cells of approximately the same size, but without appreciable laminar organization. The remainder of the transplant demonstrates little, if any, histological organization. Bar = 400 pm.

NEUROCHEMICAL

ANATOMY

OF

TRANSPLANTED

within the granule cell layer of the dentate gyrus (15,46) as well as the suhiculum and the entorhinal cortex (15). MATERIALS

AND

METHODS

Transplantation Procedures Sixteen male Long-Evans rats weighing between 225 and 360 g were anesthetized with sodium pentobarbital (60 mg/kg, ip). When the animal reached a deep surgical plane of anesthesia the scalp was shaved and the rat was placed into a stereotaxic headholder. After the scalp was incised along the midline, holes (approximately 4 mm2) were drilled through the skull about 3 mm from bregma and 3 mm from the midsagittal suture. The dura was cut and reflected. The exposed neocortex was gently aspirated. The aspiration continued through the corpus callosum to visualize the dorsolateral convexity of the hippocampus. As much of the hippocampus as possible was then removed without damaging other subneocortical structures. Hemostasis was obtained when the aspiration was completed using No. 3 size cotton pellets soaked in 0.9% saline. The cotton pellets were carefully removed after the bleeding had been arrested. Procedures for preparation and transplantation of fetal hippocampal tissue were the same as those used by Woodruff et al. (42, 43). Two female rats were bred in the colony maintained by East Tennessee State University. The presence of sperm in vaginal smears taken daily between 0900 and 1000 h was used to determine Embryonic Day 0 (EO). Fifteen and sixteen days later the pregnant female rat was anesthetized with an im injection of a ketamine hydrochloride (40 mg/kg)-acepromazine (2 mg/kg) cocktail. A cesarian section was performed and the fetuses were removed one at a time from the distal end of the uterine horn. The embryo was immediately transferred to a petri dish resting in ice and containing ice-cold Hams F-10 nutrient tissue medium (GIBCO). The brain dissection was completed for each fetus before the next fetus was removed. The fetal brain was removed under a dissecting microscope, stripped of meningeal tissue, and bissected. The ridge containing the hippocampal primordium was visualized along the inferior margin of the midline fetal pallium. This ridge was dissected away from the remainder of the fetal brain and consisted of hippocampal anlage as well as some medial neocortical primordia. The process was repeated for the contralateral hemisphere. The petri dish containing the transplant tissue was carried to the host rat and fetal tissue was placed as a solid block into each lesion cavity using microforceps. The scalp wound was then closed with sutures and a prophylatic injection of penicillin G benzathine/penicillin G procaine (25,000 IU, im) was given. Immunocytochemistry

and Histochemistry

Ninety to one hundred ten days following surgery, 14 of the rats were killed by overdose of sodium pentobar-

FETAL

HIPPOCAMPUS

39

bital and transcardially perfused with 300 ml of 0.9% saline followed by 400 ml of cacodylate-buffered (pH 7.2) 4% paraformaldehyde/0.2% picric acid fix. The brains were removed and postfixed in the same fixative for 1 to 2 h. The tissue was blocked into 3- to 4-mmthick slices. The slices were embedded in 8% calf gelatin and serially sectioned (50 pm coronal sections) on a vibrating microtome into a bath of phosphate-buffered saline (PBS; pH 7.6). Alternate sections through each transplant were immunocytochemically stained for either vasoactive intestinal polypeptide (VIP), somatostatin (SS), neurotensin (NT), substance P (SP), methionine-enkephalin (met-enkephalin; MENK), tyrosine hydroxylase (TH), or choline acetyltransferase (ChAT) using the avidin-biotin-horseradish peroxidase procedure (21). (The hyphenated suffix “like” following each of these substances is omitted for the sake of brevity.) Each section was stained for only one of the listed substances and any particular brain was stained for from three to five of the substances. The staining procedure was similar to that previously described (44). Briefly, the free-floating sections were rinsed in PBS to remove the fix and placed in 10% normal goat serum (NGS)/0.2% Triton X-100 for 1 h. Following two rinses in 10% NGS (30 minirinse) the sections were incubated in primary antibody diluted in a goat serum/sodium azide solution for 12 to 14 h at 4°C. After incubation in primary antibody the sections were rinsed in PBS (6 X 10 min) and then incubated in biotinylated secondary antiserum (goat anti-rabbit IgG; 1:300; Vector Laboratories, Burlingame, CA) for 1 to 4 h. The sections were then rinsed in PBS (2 X 15 min) followed by two rinses (15 minirinse) in Tris-buffered saline (TBS, pH 7.6), and transferred to Vectastain avidin-biotin complex (Vector Laboratories) for 1 to 4 h. Following rinses in TBS and PBS the final reaction enabling visualization of the biotinylated horseradish peroxidase was performed using 3,3’-diaminobenzidine tetrahydrochloride (Sigma) and glucose oxidase (Boehringer-Mannheim). The sections were then slide mounted and air-dried. Most of the slides were counterstained with methyl green. The concentrations of the primary antibodies were 1:lOOO (VIP, SP, TH, ChAT) or 1:3000 (MENK, SS, NT). The primary antibodies for VIP, SP, NT, MENK, and SS were obtained from Immunonuclear Corp. (Stillwater, MN) and that for TH was obtained from Eugene Tech International (Allendale, NJ). The primary antibody for ChAT was raised in the laboratory of Dr. Louis B. Hersh and has been thoroughly characterized (6, 17). As a control for specificity some sections from each brain were incubated at 4°C in NGS without the primary antibody. Omission of the primary antibody resulted in no immunopositive staining. The two remaining rats were also killed by pentobarbital overdose 90 to 110 days after surgery. After trans-

40

WRAY,

BAISDEN,

AND

WOODRUFF

FIG. 2. (A) Thionin-stained section through a portion of a transplant that had grown in an ectopic position external to the pia and arachnoid. The arrows indicate a cluster of similarly sized neurons. Bar = 250 am. (B) In addition to the neurons that are present in such ectopic grafts, fibers that stain for AChE can also be observed. Bar = 250 pm. (C) This is a thionin-stained section taken more caudally through

NEUROCHEMICAL

ANATOMY

OF

TRANSPLANTED

FETAL

HIPPOCAMPUS

41

FIG. 3. (A) Clusters of methyl green-stained cells demonstrate ChAT immunoreactivity along their borders. Four of these are indicated by arrowheads. Bar = 50 pm. (B) Occasional neurons within the transplants exhibited substantial ChAT immunoreactivity throughout their cell body and into their processes. Bar = 50 pm. (C) ChAT- immunoreactive fibers were seen to enter the transplants from any source of choline& axons that merged with the transplant. In C many ChAT-immunoreactive axons (below arrowhead) cross from the host striatum (St) into the transplant (tp; bar = 100 pm), while in D the source of ChAT-immunoreactive fibers (examples between arrowheads) is the fimbria (f; bar = 50 pm).

cardial perfusion with 300 ml of 0.9% saline followed by 400 ml of 1% glutaraldehyde/l.25% paraformaldehyde in 0.1 M phosphate buffer containing 10% sucrose, the brains of these rats were sectioned frozen at 60 pm in the coronal plane. Alternate sections were mounted on slides and stained with thionin or processed using the

procedure for visualization of AChE described by Koelle (23). Pseudocholinesterase activity was inhibited by addition of 1 &i N,N’-bis(l-methylethyl)pyrophosphorodiamidic anhydride (ISO-OMPA) to the preincubation and incubation media. ISO-OMPA was omitted for one set of slides from each brain.

the transplant presented in A and B. The transplant/host boundary transplant containing neuron cell bodies adjacent to fiber bundles transplant was stained for AChE. The arrows point to AChE-stained indicate the host/transplant interface. The host brain demonstrates higher power photomicrograph of a thionin-stained section through hippocampus. The smaller arrowheads indicate the transplant/host pared to the CA1 region (at open arrowhead) of the host hippocampus larger arrowhead) and (F) the AChE-positive fibers (open arrowhead) arrow) within the transplant in the vicinity of these cell bodies. Bars

is indicated by the arrowheads. The arrows indicate an area of the that are unstained. Bar = 500 pm. (D) The next section through this fibers that are a part of the clear areas indicated in C. The arrowheads AChE staining denser than that of the transplant. Bar = 500 pm. (E) A this transplant at a more caudal level as it fuses with residual host boundary. While not presenting a genuine cytoarchitecture when com(h), the neurons within the transplant (tp) form clusters (arrow and that follow the CA neurons appear to concentrate (larger arrowhead and = 100 pm.

42

WRAY,

BAISDEN,

AND

WOODRUFF

FIG. 4. (A) The transplant (tp)/host cortex (c) interface is indicated by the row of smaller arrowheads. Two neurons within the transplant demonstrating TH immunoreactivity are indicated by larger arrowheads (methyl green counterstain). Bar = 100 km. (B) The appearance of one of these cells suggests that of a pyramidal cell. Bar = 50 pm. (C) This photomicrograph presents a varicose SP-immunoreactive fiber (arrowheads) from one of the transplants. Bar = 50 pm. (D) Cells (arrowheads) within the transplants stained for SP exhibited SP immunoreactivity along the edges of their somata and along their dendrites. Bar = 50 pm.

RESULTS

Histology

of the Lesion Sites and Transplants

In the brains that had not developed transplants the lesion cavity resembled those presented as serial reconstructions in our previous publication (42). A photomicrograph of a cross section taken through the largest extent of a typical lesion is presented in Fig. 1A. A portion of the dorsolateral neocortex is destroyed. The lesion includes the dorsolateral and lateral part of the hippocampus. The most dorsal and ventral (temporal) parts of the hippocampus are spared (Fig. 1A) as are the most rostra1 (septal) and caudal portions (42). Because most of the brains examined in the present study were sectioned with a Vibratome, thickness presumably varied slightly from section to section. Additionally, some tissue was destroyed as the sections were

transferred from solution to solution. For these reasons calculation of transplant size was not attempted. However, as can be seen in Fig. lB, the transplants typically filled a large part of the lesion cavity. They also occasionally extended from the lesion cavity and grew above the pial surface of the cortex (Figs. 2A and 2B). Previous volumetric studies (33,43) of the growth of transplants made using the procedures employed in this experiment indicate that they do not increase in size after Posttransplantation Day 45. Therefore, in terms of gross size the present experiment involved observation of mature transplant tissue. Figures 1 and 2 present examples of typical transplant histology. Restricted areas of each transplant contained cytoarchitecture reminiscent of hippocampal histology (Fig. 1B). Other segments of each transplant contained aggregations of similar-sized neurons that formed loose rows (Fig. 2E) or simply occupied aparticu-

NEUROCHEMICAL

ANATOMY

OF

TRANSPLANTED

FETAL

HIPPOCAMPUS

43

FIG. 6. (A) SS-immunoreactive cells were shaped somewhat more irregularly than neuron cell bodies that reacted for the other substances and the immunoreactivity was spread in patches throughout the cell body. As can be seen in this oil immersion photomicrograph processes arising from these cells were difficult to trace for more than 25 to 30 pm from the cell body. Bar = 25 pm. (B) MENKimmunoreactive fibers (smaller arrowheads) tended to be fine and only sparsely scattered within the transplants. MENK-immunoreactive cell bodies (larger arrowhead) were relatively small, but processes could be followed away from the cell body (open arrowhead). Bar = 100 pm.

FIG. 5. (A) This photomicrograph of a section of transplant presents a VIP-immunoreactive cell with a round-to-oval cell body and a beaded process. Bar = 20 pm. (B) Beaded VIP-positive fibers (arrowheads) were also found unattached to neighboring VIP-immunoreactive cells within the transplants (Bar = 30 pm). (C) This oil immersion photomicrograph indicates that the VIP immunoreactivity was fairly heavily distributed throughout the somata of stained neurons. Bar = 50 pm.

lar area of the transplant (Figs. 1B and 2A). As represented in Figs. lB, 2A, and 2C, the histology of the remaining and largest part of each transplant had no apparent cytoarchitectural organization. The degree to which cytoarchitecture developed within a given area of the transplant did not appear to depend on the area of host brain with which it interfaced. Choline AcetyltransferaselAcetylcholinesterase The brains of eight rats were processed using the primary antibody directed against ChAT. Of these eight

44

WRAY,

BAISDEN,

brains six developed bilateral transplants while two had only unilateral transplants. Both of the brains processed to visualize AChE contained bilateral transplants. Several observations concerning choline&c innervation of the transplants were made from this material. First, ChAT-reactive or AChE-positive fibers were present throughout the grafts, even, as is shown in Figs. 2A and 2B, in sections taken from a part of the transplant that had grown in an ectopic position dorsal to the neocortex and did not interface with the host brain parenchyma at the level of the section. Second, although putative cholinergic fibers could be seen throughout each transplant, their density was higher at the transplant/ host interface and in areas that had relatively high concentrations of neuron cell bodies (Figs. 2D-2F). Third, ChAT-immunoreactive or AChE-positive fibers crossed the transplant/host interface wherever the transplant merged with the host brain parenchyma near a source of cholinergic input (Figs. 2 and 3). Fourth, putative cholinergic innervation of the grafts was generally less dense than that of surrounding host brain (Fig. 2). Finally, in addition to neurons that stained with methyl green and had punctate ChAT-positive staining around their somata and dendrites, a few ChAT- or AChE-positive neurons were found in the transplants (Fig. 3). These were relatively small and round to oval in shape. They resemble ChAT-immunoreactive neurons found in normal hippocampus as reported by Frotscher et al. (14).

AND

WOODRUFF

cose axons. Second, punctate SP immunoreactivity was observed between cells within each transplant. Third, numerous neuron somata were ringed by SP-positive puncta which often extended along the surface of the dendrites (Fig. 4D). Because this SP immunoreactivity is limited to the edges of the cells it is interpreted as being terminals opposing the soma and dendrites rather than as indicating that the neuron itself is SP positive. SP-positive cell bodies could not be unequivocally demonstrated. This is compatible with previous reports for normal hippocampus in which SP-immunoreactive neuron somata could only be visualized in rats pretreated with colchicine (8,36) and does not preclude the existence of potential SP-immunoreactive cell bodies within our transplants.

Vasoactive Intestinal

Polypeptide

Sections from five rat brains were processed to identify TH immunoreactivity. TH-positive fibers within the transplants and crossing the transplant/host interface were identified. Cell bodies and processes containing dopamine, norepinephrine, or epinephrine will be visualized with this antibody. However, because adult hippocampal and neocortical neurons do not produce dopamine, norepinephrine, or epinephrine, TH-positive somata should not have been found, but they were (Figs. 4A and 4B). TH-positive neurons appeared in groups of 2 to 8 and as many as 30 were found in a section. The shapes of the cell bodies suggest that they may be cortical neurons.

Sections from the brains of six rats were processed to visualize VIP immunoreactivity. All of these six had developed bilateral transplants. VIP-immunoreactive cell bodies and fibers were observed in all of the transplants. The shape of the cell bodies varied (Fig. 5). Some somata were round, some were fusiform, and a few were irregular and gave off several processes. The dendrites arising from these cell bodies appeared to be aspinous. In addition to the dendrites that could be followed from cell bodies, VIP-positive axons were observed. These axons appeared slightly varicose (Fig. 5A) and resembled disconnected fibers found scattered throughout the transplant (Fig. 5B). These disconnected fibers did not form appreciable plexuses as have been observed in normal hippocampus (29), nor were VIP-positive somata consistently associated with groups of nonimmunoreactive methyl green-stained neurons within the transplant as might be expected from their location in the pyramidal and granule cell layers of the normal Ammon’s horn and dentate gyrus. The VIP-ir cells were not arranged in any detectable pattern or orientation. Although many VIP-positive cells and fibers were observed in parts of the host cortex that merged with the transplants, VIP-positive fibers were not seen to cross the host/transplant interface. This observation suggests that the VIP-immunoreactive fibers found within the transplants participate in local circuit interactions as they are believed to do in normal hippocampus (27).

Substance P

Somatostatin

Sections from the brains of seven rats were processed to demonstrate substance P immunoreactivity. Two of these had only unilateral transplants, resulting in a total of 12 transplants that could be examined. SP immunoreactivity was found in all of these transplants and appeared in three forms. First, fibers were found (Fig. 4C), particularly in the periphery of the transplants near the graft/host interface. These tended to be vari-

Sections from the brains of five rats, each having bilateral transplants, were processed to demonstrate SS immunoreactivity. SS-immunoreactive cell bodies were observed scattered within each transplant. The somata were irregular in shape, as opposed to the fusiform and round somata noted for cells containing other peptides, and the SS immunoreactivity appeared scattered throughout the cell body (Fig. 6A). Neither dendrites

Tyrosine Hydroxyluse

NEUROCHEMICAL

ANATOMY

OF

TRANSPLANTED

nor axons could be followed for more than a few micrometers past their points of origin at the cell body and the SS-immunoreactive fibers that could be observed within the grafts were not traceable to any particular cells. Neurotensin The brains from four rats with bilateral transplants and one rat that had developed only a unilateral transplant had sections stained for NT immunoreactivity. NT immunoreactivity was observed in six of nine transplants and was restricted to fibers. Further, the NT-immunoreactive fibers were few in number and they were scattered in areas of the transplant where an interface with damaged host dorsal hippocampus occurred. NT-ir neuron somata were observed in residual host subiculum. Met-Enkephalin Sections from the brains of the rats also treated to reveal NT immunoreactivity were reacted for MENK. Although sparsely distributed, MENK-positive fibers could be observed within all of the grafts (Fig. 6B), but could not be seen to surround any particular size of cell or to exhibit any pattern. MENK-immunoreactive fibers were not observed to cross the host/transplant interface. In addition to the MENK-immunoreactive fibers, an occasional lightly stained MENK-positive cell body could also be observed in each transplant (Fig. 6B). DISCUSSION Accepted markers for one “classical” neurotransmitter (acetylcholine), the cell bodies of which are primarily extrinsic to the hippocampus; a marker that may indicate the presence of any of three additional neurotransmitters (dopamine, norepinephrine, epinephrine); and markers for five neuropeptides that represent both intrinsic and extrinsic systems of the hippocampal formation were chosen as a sample for this study. The results were not intended to present an exhaustive catalog of all the neurotransmitters and neuromodulators potentially associated with the hippocampal formation. However, the data collected in the present study can be generalized to address two questions regarding the neurochemical anatomy of transplants of medial pallial tissue made into large lesion cavities. The first question is whether neurotransmitters and neuromodulators appropriate to the hippocampus are present in the mature hippocampal graft. The second question is whether an organization of cell bodies and fibers that resembles that observed in normal hippocampus develops in the grafts. All of the substances sampled were visualized in the transplants. This finding is compatible with previous

FETAL

HIPPOCAMPUS

45

reports. For example, it has been known since the earlier reports concerning successful transplantation of fetal dopaminergic ventral mesencephalon or cholinergic septal and basal forebrain neurons (1,5,10,30) into the rat brain that grafted neurons could express appropriate neurotransmitters. The results of a few additional experiments have indicated that neuropeptides will also be expressed within grafts of fetal neocortical tissue made into adult rat neocortex (12,20,35), subcortical brain regions (20), or spinal cord (4). Further, Zimmer and his colleagues have observed the presence of somatostatin-, cholescystokinin-, enkephalin-, and acetylcholinesterase-positive cells and fibers in rat-rat allografts (46) and in mouse-rat xenografts (45). The results of the present study extend these observations to transplants made into large lesion cavities where they may receive inputs from a variety of sources not available to the transplants made in the cited studies. The fact that these transplants still contain many of the appropriate neurochemicals provides additional support for the conclusion that most, if not all, of the neurotransmitters and neuromodulators that should be expressed by neuron somata and fibers typical of the transplanted tissue are indeed expressed. This conclusion does not entail that these transmitters and modulators were present in either a normal pattern or a normal amount. Of the peptides, SP and VIP appeared to be the most robust. The density of SP fibers and of VIP cell bodies and fibers within the transplants was as heavy as within residual host hippocampal tissue and also compared to the density of these peptides as represented in previous publications (8, 22, 24, 27). Because preinjections of colchicine were not given SP-ir cell bodies were not visualized in the present study so no comment can be made concerning their size or distribution. However, as indicated above, all regions of the normal hippocampus appear to contain SP-ir fibers and terminals. The terminals are observed as SP-positive puncta around neurons of varying sizes (8) as they were in the present study, suggesting that the developmental signals that govern at least the ultimate positioning of SP terminals on transplant neurons may have functioned properly. However, this may not be the case for VIP-containing axons. While VIP-ir cell bodies did exhibit the range of sizes and shapes reported for the normal hippocampus (24,27) and the VIP-ir fibers were varicose as has been described (24, 27), the cell bodies were scattered throughout the transplants and the fibers did not form the plexuses that are seen in normal hippocampus (24). These observations suggest that migration patterns for these neurons were disrupted during development and that subsequent fiber outgrowth also failed to follow normal patterns. That VIP-ir fibers did not cross the host/transplant interface may be considered an exception to this statement. This suggests

46

WRAY,

BAJSDEN,

that VIP neurons retain their identity as interneurons even in this abnormal situation. The remaining three neuropeptides did not appear to be as well represented in the transplants as they were in the remaining host brain hippocampus or as densely as may be inferred from the reports of others. For example, Roberts et al. (34) reported that SS-ir cell bodies were more numerous in rat hippocampus than VIP-ir cell bodies and they were at least as numerous as VIP-ir somata in residual host hippocampus in the present study, but in the transplants SS-ir cell bodies were noticeably less numerous than VIP-ir cell bodies. The apparent difference between density of SS-ir somata in the transplants and that reported by Roberts et al. (34) for normal hippocampus might reflect differences in the antibody or technique (e.g., PAP vs ABC). The apparently greater proportion of SS-ir cell bodies in residual host hippocampus compared to that in transplant should be interpreted with caution because the lesion may have affected SS synthesis or transport. But it may also be that potential intrinsic SS-ir neurons do not survive the procedures we employ for transplantation or that SS-ir is not expressed in detectable amounts. The same comment may be made for MENK. Although many small neurons were found in methyl green-stained sections and some even formed dentatelike configurations, very few showed MENK-ir. This stands in contrast to previous reports (l&46) and may well be due to the lack of cholchicine pretreatment in the present study. However, we also failed to find the large numbers of MENK-ir fibers reported in normal hippocampus (15, 46). Once again, either antibody or procedural differences may account for this, but the possibility that MENK-ir is not expressed in transplants made using the procedures we have used cannot be dismissed. Roberts et al. (34) reported that NT-ir fibers are comparatively few within the hippocampus and that NT-ir cell bodies are restricted to the subiculum. We failed to observe NT-ir cell bodies within the transplants although the subiculum was included. Roberts et al. (34) also did not use colchicine and we did find NT-ir cells in the subiculum of our hosts. Thus the failure to observe NT-positive cells within the transplants is unlikely to be due to omission of colchicine pretreatment. An indication that the antibody was sensitive to the presence of NT within the transplants is the observation of NT-ir fibers spatially associated with the transplant/host hippocampus. The most likely source of these fibers would seem to be the damaged host brain. These observations concerning the peptides suggest that, with the possible exception of SP, the development of these substances was abnormal within the transplant. Either the pattern of fiber growth did not mimic that seen in normal hippocampus (VIP), most likely due to the disorganization of much of the transplant histol-

AND

WOODRUFF

ogy, or, as in the case of SS, MENK, and NT, the expression of these peptides by transplant neurons was reduced compared to normal hippocampus. Since these peptides are associated with various parts of the internal circuitry of the hippocampus, their lack may contribute in part to deficient information processing that would relate to the inability of these transplants to compensate the effects of hippocampal damage on behaviors such as the Morris water maze (41) that require more complex stimulus-response coding than do tasks such as DRL and passive avoidance which are improved by postlesion transplants of hippocampal anlage (41-43). Another indication that not all of the mechanisms leading to the development of the neurochemical anatomy of the normal hippocampus or neocortex are operating properly in transplants of anlage of these structures is the appearance of TH-ir neurons. This observation extends that of Herman et al. (20) who saw neuron cell bodies exhibiting TH positivity in mature (40 days posttransplantation) grafts of neocortex. There is no evidence that the presence of TH immunoreactivity in transplanted pallial neurons means that these neurons necessarily produce dopamine (20). However, the existence of this enzyme indicates that at least some transplanted neurons are in an abnormal state. Further, the observation that expression of some of the peptides may be retarded in the transplanted tissue, combined with finding TH-ir neuron somata within the grafts, indicates that transplantation using the procedures we employ places the developing neurons in an environment that may lack the mechanisms necessary to regulate both expression and suppression of some genes as these are normally regulated in uiuo. Although the putative peptidergic neurons and fibers did not have an obvious organization, attention to the pattern of cell organization seen with methyl green staining or in thionin-stained sections taken from the series prepared for AChE histochemistry indicated that, as we reported previously (40-43), neuron cell bodies within the transplants do form clusters and rows which in some parts of the transplant resemble the dentate gyrus or Ammon’s horn. Of particular interest is the observation that while immunopositive peptidergic fibers did not form obvious bundles or plexuses associated with groupings of neurons, innervation by AChEstained fibers was denser in the vicinity of clusters of neuron cell bodies within the transplant than in areas where the cells were more widely scattered. This observation suggests that tropic and trophic substances, such as nerve growth factor (2,38), to which host choline@ fibers respond are associated with the transplant neurons. These substances apparently guide and support growth of choline@ fibers into the transplants from any of several sources. This observation is compatible with experiments using what might be considered an

NEUROCHEMICAL

ANATOMY

OF

TRANSPLANTED

experimental design obverse to that employed in this study. That is, transplanted fetal choline@ cells derived from any one of three sites have been shown to innervate the denervated adult host hippocampus (18, 28). In conclusion, the results of this study indicate that transplants of hippocampal formation anlage taken from El6 fetal rat brain and placed immediately into large lesions of the adult hippocampus express several of the peptides normally observed associated with Ammon’s horn and the dentate gyrus. However, the expression of some of these peptides may be reduced compared to normal hippocampus and the neurochemical anatomy of the somata and fibers reacting for these peptides is not normal. This suggests that the transplanted tissue may lack appropriate epigenetic factors to guide histogenesis and fiber outgrowth of these neurons. Further, at least one substance, tyrosine hydroxylase, that should not appear in adult pallium is found in neurons within the grafts. This indicates that some factor necessary for suppression of the gene for TH expression is lacking from the transplants. However, the molecular mechanisms for the expression of factors supporting ingrowth of choline@ factors appear to be active within the graft tissue and lead to innervation from sources of cholinergic fibers that would not normally supply the hippocampus.

0407003-03) and the Frato M.L.W. Preparation of Grant AGO 5893 to Louis to us for use in this study.

plantation performance

2.

3.

4.

5.

6.

magnocellularis cerebral

1987. Transplantation of choline@ neurons into the cortex. Ann. N. Y. Acad. Sci.

BATCHELOR, P. E., D. M. ARMSTRONG, S. N. BLAKER, AND F. H. GAGE. 1989. Nerve growth factor receptor and choline acetyltransferase colocalization in neurons within the rat forebrain: Response to fimbria-fornix transection. J. Comp. Neurol. 284: 187-204. BERM~EZ-RA?TONI, F., J. FERNANDEZ, M. ANTONIO-SANCHEZ, R. AGUILAR-ROBLJZRO, AND R. DRUCKER-COLfN. 1987. Fetal brain transplants induce recuperation of taste aversion learning. Bruin Res. 416: 147-152. BERNSTEIN, J. J., AND J. P. CONNOR. 1986. Somatostatin-like neurons are expressed in fetal neocortical homografts in adult rat spinal cord. Brain Res. 374: 147-152. BJBRKLUND, A., U. STENEVI, R. H. SCHMIDT, S. B. DUNNETT, AND F. H. GAGE. 1983. Intracerebral grafting of neuronal cell suspensions implanted in different brain sites. Actu Physiol. Stand. (Suppl.) 522: 9-18. BRUCE, G., B. H. WAINER, AND L. B. HERSH. 1985. Immunoaffinity purification of human choline acetyltransferase: Com-

enzymes.

of norepinephrine of a learned

task.

neurons Brain

J. Neurochem.

into aged rats improves Res. 448: 77-87.

DAVIES, S., AND C. K~HLER. 1985. The substance of the rat hippocampal DECKEL,

A. W., R. G.

region.

Anat.

Emixyol.

P innervation 173: 45-52.

ROBINSON, J. T. COYLE, AND P. R. SAN-

BERG. 1983. Reversal

of long-term locomotor abnormalities in the kainic acid model of Huntington’s disease by day 18 fetal striatal implants. Eur. J. Pharmacol. 93: 287-288.

10.

DIJNNETT, S. B. 1987. Anatomical of cholinergic-rich of the nucleus 495: 415-430.

11.

and behavioral consequences grafts to the neocortex of rats with lesions basalis magnocellularis. Ann. N. Y. Acad. Sci.

DLJNNETT, S. B., W. C. Low, BJ~RKLUND. in rats with

1982. Septal fornix-fimbria

S. D. IVERSEN, U. STENEVI, AND A. transplants restore maze learning lesions. Bruin Res. 251: 335-348.

12.

EBNER, F. F., J. A. OLSCHOWSKA, AND D. M. JACOBOWIT~. The development of peptide-containing neurons within cortical transplants in adult mice. Peptides 5: 103-113.

13.

FROTSCHER, M., R. NITSCH, AND C. LBRANTH. 1989. Choline@ innervation of identified neurons in the hippocampus: Electron microscopic double labeling studies. In The Hippocampus: New Vistas (V. Chan-Palay and C. Kohler, Eds.), pp. 8596. A. R. Liss, New York.

14.

FROTSCHER, M., M. SCHLANDER, AND C. LI?RANTH. linergic neurons in the hippocampus: A combined electron-microscopic immunocytochemical study Cell Tissue. Res. 246: 293-301.

15.

GALL,C.,N.

16.

GASH, D. M., T. J. COLLIER, AND J. R. SLADEK, JR. 1985. Neural transplantation: A review of recent developments and potential applications to the aged brain. Neurobiol. Aging 6: 131~ 150.

17.

GERMAN, D. C., G. BRUCE, AND L. B. HERSH. 1985. Immunohistochemical staining of cholinergic neurons in the human brain using a polyclonal antibody to human choline acetyltransferase. Neurosci. Lett. 61: l-5.

18.

GIBBS, R. B., K. ANDERSON, AND C. W. COTMAN. 1986. Factors affecting innervation in the CNS: Comparison of the three cholinergic cell types transplanted to the hippocampus of adult rats. Brain Res. 383: 362-366.

19.

HARA, Y., S. SHIOSAKA, E. SENBA, M. SAKANAKA, S. INAGAKI, H. TAKAGI, Y. KAWAI, K. TAKATSUKI, T. MATSUZAKI, AND M. TOHYAMA. 1982. Ontogeny of the neurotensin-containing neuron system of the rat: Immunohistochemical analysis. I. Forebrain and diencephalon. J. Comp. Neural. 208: 177-195.

20.

HERMAN, J. P., N. ABROUS, A. VIGNY, J. DULLUC, AND M. LEMOAL. 1988. Distorted development of intracerebral grafts: Long-term maintenance of tyrosine hydroxylase-containing neurons in grafts of cortical tissue. Deu. Bruin Res. 40: 81-88.

21.

Hsu, S.-M., L. RAINE, AND H. FANGER. 1981. Use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase techniques: A comparison between ABC and unlabelled antibody (PAP) procedures. J. Histochem. Cytochem. 29: 577-580.

22.

KIMBLE, D. P., R. BREMILLER, AND G. STICKROD. 1986.Fetal

ARENDASH, G. W., AND P. R. MOUTON. nucleus basalis choline+-depleted 495: 431-443.

and placental

COLLIER, T. J., D. M. GASH, AND J. R. SLADEK, JR. 1988. Trans-

REFERENCES 1.

47

HIPPOCAMPUS

parison of the brain 45: 611-620.

ACKNOWLEDGMENTS Supported by grants from the NIH (ES ternal Order of Eagles Golden Eagle Fund the CbAT antibody was supported by NIH B. Hersh, Ph.D., who generously donated it

FETAL

1986. Cholight- and in the rat.

BRECHA,H.J.KARTEN,AND K.-J. CHANG.~~~~. Localization of enkephalin-like immunoreactivity to identified axonal and neuronal populations of the rat hippocampus. J. Comp. Neurol. 198: 335-350.

brain implants improve maze performance sioned rats. Brain Res. 363: 356-363. 23.

1984. neo-

KOELLE,

G. B. 1955. The

histochemical

in hippocampal-leidentification

of acetyl-

48

24.

25. 26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

WRAY,

BAISDEN,

cholinesterase in choline@, adrenergic, and sensory neurons. J. Phnrmacol. Exp. Ther. 114: 167-184. KBHLER, C. 1983. A morphological analysis of vasoactive intestinal polypeptide (VIP)-like immunoreactive neurons in the area dentata of the rat brain. J. Comp. Neural. 221: 247-262, KBHLER, C., AND V. CHAN-PALAY. 1982. Somatostatin-like immunoreactive neurons in the hippocampus: An immunocytochemical study in the rat. Neurosci. Lett. 34: 259-264. KROMER, L. F., A. BJBRKLUND, ANLI U. STENEVI. 1983. Intracephalic embryonic neural implants in the adult rat brain. I. Growth and mature organization of brainstem, cerebellar, and hippocampal implants. J. Comp. Neural. 218: 433-459. LI%R&NTH, Cs., M. FROTSHER, T. T~MBOL, AND M. PALKOVITS. 1984. Ultrastructural and synaptic connections of vasoactive intestinal polypeptide-like immunoreactive non-pyramidal neurons and axon terminals in the rat hippocampus. Neuroscience12:531-542. LEWIS, E. R., AND C. W. COTMAN. 1983. Neurotransmitter characteristics of brain grafts: Striatal and septal tissues form the same laminated input to the hippocampus. Neuroscience 8: 57-66. LOREN, I., P. C. EMSON, J. FAHRENKRUG, A. BJ~RKLUND, J. ALUMETS, R. HAXANSON, AND F. SUNDLER. 1979. Distribution of vasoactive intestinal polypeptide in the rat and mouse brain. Neuroscience 4: 1953-1976. Low, W. C., P. R. LEWIS, S. T. BUNCH, S. B. DUNNED, S. R. THOMAS, S. D. IVERSEN, A. BJBRKLLJND, AND U. STENEVI. 1982. Function recovery following neural transplantation of embryonic septal nuclei in adult rats with septohippocampal lesions. Nature (London) 300: 260-262. LOY, R., D. A. KOZIELL, J. D. LINDSEY, AND R. Y. MOORE. 1980. Noradrenergic innervation of the adult rat hippocampal formation. J. Comp. Neurol. 189: 699-710. MORRISON, J. H., R. BENOIT, P. J. MAGISTRETTI, N. LING, AND F. BLOOM. 1982. Immunohistochemical distribution of prosomatostatin-related peptides in hippocampus. Neurosci. L&t. 34: 137-142. PAUL, D. J., R. H. BAISDEN, AND M. L. WOODRUFF. 1989. Developmental time course of grafts of fetal hippocampus placed into aspiration lesion sites in adult rats. Sot. Neurosci. Abstr. 15: 1219. ROBERTS, G. W., P. L. WOODHAMS, J. M. POLAK, AND T. J. CROWE. 1984. Distribution of neuropeptides in the limbic system of the rat: The hippocampus. Neuroscience 11: 35-77. SHARP, F. R., M. F. GONZALEZ, AND S. M. SAGAF~. 1987. Fetal frontal cortex transplanted to injured motor/sensory cortex of adult rats. II. VIP-, somatostatin-, and NPY-immunoreactive neurons. J. Neurosci. 10: 3002-3015.

AND WOODRUFF 36. SHULTS, C. W., P. JOHNSTON, AND F. H. GAGE. 1987. Comparison of substance K-like and substance P-like fibers and cells in the rat hippocampus. Brain Res. 426: 290-296. 37. STEIN, D. G., R. LABBE, A. FIRL, JR., AND E. J. MUFSON. 1985. Behavioral recovery following implantation of fetal brain tissue into mature rats with bilateral, cortical lesions. In Neural Grafting in the Mammalian CNS (A. Bjorklund and U. Stenevi, Eds.), pp. 605-614. Elsevier, Amsterdam. 38. TONIOLO, G., S. B. DUNNETT, F. HEF-~I, AND B. WILL. 1985. Acetylcholine-rich transplants in the hippocampus: Influence of intrinsic growth factors and application of nerve growth factor on choline acetyltransferase activity. Brain Res. 345: 141146. 39. VERNEY, C., M. BAIJLEC, B. BERGER, C. ALVAREZ, A. VIGNY, AND K. B. HELLE. 1985. Morphological evidence for a dopaminergic terminal field in the hippocampal formation of young and adult rat. Neuroscience 14: 1039-1052. 40. WOODRUFF, M. L., AND R. H. BAISDEN. 1988. Structure and function of intracerebral grafts of fetal neocortical or hippocampal tissue. In Neural Plasticity: A Lifespan Approach (T. L. Petit and G. 0. Ivy, Eds.), pp. 69-104. A. R. Liss, New York. 41. WOODRUFF, M. L., R. H. BAISDEN, AND A. J. NONNEMAN. 1990. Transplantation of fetal hippocampus may prevent or produce behavioral recovery from hippocampal ablation and recovery persists after removal of the transplant. In Neural Transplantation: From Molecular Bask to Clinical Application. Progress in Brain Research (S. B. Dunnett and S.-J. Richards, Eds.), Vol. 82, pp. 367-376. Elsevier, Amsterdam. 42. WOODRUFF, M. L., R. H. BAISDEN, D. L. WHI~INGTON, AND A. E. BENSON. 1987. Embryonic hippocampal grafts ameliorate the deficit in DRL acquisition produced by hippocampectomy. Brain Res. 408: 97-117. 43. WOODRUFF, M. L., R. H. BAISDEN, D. L. WHI~INGTON, N. L. SHELTON, AND S. WR.AY. 1988. Grafts containing fetal hippocampal tissue reduce activity and improve passive avoidance in hippocampectomized or trimethyltin-exposed rats. Erp. Neural. 102: 130-143. 44. WRAY, S., A. NIEBURGS, AND S. ELKABES. 1989. Spatiotemporal cell expression of luteinizing hormone-releasing hormone in the prenatal mouse: Evidence for an embryonic origin in the olfactory placode. Dev. Brain Res. 46: 309-318. 45. ZIMMER, J., B. FINSEN, T. SORENSEN, AND P. H. POULSEN. 1988. Xenografts of mouse hippocampal tissue. Exchange of laminar and neuropeptide specific nerve connections with the host rat brain. Brain Res. 20: 369-379. 46. ZIMMER, J., AND N. SUNDE. 1984. Neuropeptides and astroglia in intracerebral hippocampal transplants: An immunohistochemical study in the rat. J. Comp. Neurol. 227: 331-347.