Cells of Origin of Entorhinal Cortical Afferents to the Hippocampus and Fascia Dentata of the Rat OSWALD STEWARD AND SHEILA A . SCOVILLE Dt,po rt t i t I I t s of Ncz t f mloq ictf I S t I rq L’f-q t i 12d P h I / 6 t o l o g , U t i I TI^' r,\ i t g ~f vfr g ; J 1 f ( f s C / f O O / Of M t ’ d i C i t f t ’ , ch ( I l / O t t C ’ , S U i / / c ’ , V i r g [ I l f t f 22901 ( 3

ABSTRACT The pathway from the entorhinal cortical region to the hippocampal formation h a s previously been shown to be comprised of two sub-systems, one of which projects predominantly to the ipsilateral fascia dentata and regio inferior of the hippocampus proper, and a second which projects bilaterally to regio superior. The goal of the present investigation was to determine if these two pathways might originate from different cell populations within the entorhinal area. The cells of origin of these entorhinal pathways were identified by retrograde labeling with horseradish peroxidase (HRP). Injections which labeled the entorhinal terminal fields i n both the fascia dentata and regio superior resulted in the retrograde labeling of two populations of cells i n the entorhinal area. Ipsilateral to the injection, HRP reaction product was found in the cells of layer I1 (predominantly stellate cells) and the cells of layer 111 (predominantly pyramidal cells). Contralateral to the injections, however, the reaction product was found almost exclusively i n the cells of layer 111. With selective injections of the entorhinal terminal field i n regio superior, only the cells of layer I11 were labeled, but these were labeled bilaterally. Selective injection of the entorhinal terminal field in the fascia dentata, however, resulted i n the labeling of cells of layer 11, but not of layer 111, and these cells of layer I1 were labeled almost exclusively ipsilaterally. A very small number of labeled cells i n layer I1 were, however, found contralateral to the injection as well. No labeled cells were found either i n the presubiculum or parasubiculum following injections of the hippocampal formation. These cell populations were found capable of retrograde transport of HRP, however, since cells in both presubiculum and parasubiculum were labeled following H R P injections into the contralateral entorhinal area. These results suggest that the projections to the fascia dentata originate from the cells of layer 11, while the projections to regio superior originate from the cells of layer 111 of the entorhinal region proper. The very slight crossed projection from the entorhinal area to the contralateral area dentata probably originates from the small population of cells in layer I1 which are labeled following HRP injections i n the contralateral area dentata

A previous analysis of the topographic organization of the entorhinal cortical projections to the hippocampal formation of the rat indicated the existence of two subsystems of entorhinal efferents which differed first by the extent of their decussation, and second by their topographic organization (Steward, ’76). The major component of the entorhinal projection system terminates in the fascia dentata and regio inferior of the hippocampus proper, and is predominantly a n ipsilateral system (although a very sparse crossed pathway to the dentate gyrus has recently been deJ.

COMP.

NEUR., 169; 347-370.

scribed (Zimmer and Hjorth-Simonsen,’75, Goldowitz et al., ’75; Steward, ’76). This projection system terminates in a laminated fashion along the dendrites of dentate granule cells and hippocampal pyramidal cells of regio inferior, with fibers from the medial entorhinal region terminating at mid proximo-distal dendritic segments, and afferents from the lateral portions of the entorhinal region terminating on distal dendritic segments (Hjorth-Simonsen and Jeune, ’72; Hjorth-Simonsen,’72; Steward, ’76; and see fig. 1). The second major sub-system of the en34 7

348

OSWALD STEWARD AND SHEILA A. SCOVILLE

torhinal pathway to the hippocampal formation projects bilaterally with approximately equivalent density on each side, at least in the rat, and terminates on the most distal dendrites of the pyramidal cells of regio superior, including the CA1 field, and the subiculum (Blackstad, '56; Raisman et al., '65; Steward et al., '73, '74; Steward, '76). A quite different organization characterizes this projection system, since the afferents from medial and lateral entorhinal regions are organized i n a longitudinal dimension across the dendritic field of regio superior. Fibers from the medial entorhinal region terminate i n the CA1 field furthest from the subiculum (near the CAl/CA2 transition), while fibers from the lateral entorhinal area, possibly including the perirhinal area, terminate i n regio superior immediately adjacent to the subiculum, and i n the subiculum proper (Steward, '76). Thus, entorhinal afferents to regio superior are topographically organized in a longitudinal dimension across the CA2-subicular axis of regio superior (fig. 1). A somewhat similar organization of entorhinal projections has recently been described in the monkey (Van Hoesen and Pandy a, '75 a). Since the entorhinal projections to the dentate gyrus and regio inferior are almost exclusively ipsilateral, and are organized along the dendrites of the recipient cells, while the projections to regio superior are bilateral, and are topographically organized across a field of dendrites, the question arose if these two entorhinal projection systems with quite different patterns of topographic organization might originate from different cell populations within the entorhinal area. The present study was designed to investigate this possibility, through the use of the horseradish peroxidase (HRP) method for the retrograde labeling of neuronal pathways.

formation under stereotaxic guidance. The enzyme solution was delivered with a Hamilton 1 pl syringe, which was hydraulically driven by a Harvard Apparatus syringe pump. Injection volumes for the present , experiments ranged from 0.3-0.5 ~ 1 and these volumes were delivered over a period of 3 0 4 5 minutes. After completing the delivery of the enzyme solution, the syringe was left in place for approximately 20 minutes, and was then removed slowly. Following surgery, the animals were given a n intramuscular injection of penicillin (Wycillin, Wyeth Laboratories), and were returned to their individual cages for the duration of the post-operative survival interval. Twenty-four to 48 hours post-injection, the animals were deeply anaesthetized with sodium pentobarbital, and were perfused transcardially with 10 % formalin in saline. The brains were post-fixed for 2-3 hours in the perfusion solution at 5 C, and were sectioned in the horizontal plane on a freezing microtome at 40 pm. HRP histochemistry was then carried out, essentially according to the method of Graham and Karnovsky ('66). The sections were incubated for five minutes at room temperature in 150 ml of Tris buffer (pH7.4) with 70 mg of 3,3'-diaminobenzidine free base (Sigma). After this pre-incubation, 0.1 ml of 30% peroxide was added, and the solution was gently agitated for a n additional 30 minutes. Following this incubation, the sections were rinsed several times, mounted from alcoholic gelatin, and permitted to dry on the slides. The sections were then stained lightly with cresylecht violet (decreasing the time in the cresyl violet solution to 45 seconds), rapidly differentiated and dehydrated to xylene, and covered. The histochemical preparations were photographed under dark-field illumination with the aid of a n Olympus Vanox microscope. O

MATERIALS AND METHODS

OBSERVATIONS

A total of 17 male Sprague-Dawley rats served as experimental animals. At the time of the experiment, these animals ranged in weight from 2 0 0 4 0 0 g. A 3050 % solution of horseradish peroxidase (HRP, Sigma Type VI) in distilled water was injected utilizing a dorsal approach into various subfields of the hippocampal

If the entorhinal projections to regio superior originate from a different population of cells in the entorhinal region than the projections to the fascia dentata, then ~

I The term subiculum is used here according to the terminology of Blackstad ('56) and refers to the prosuhiculum and subiculum (sub. area h) of Lorente de No, '34).

CELLS OF ORIGIN OF ENTORHINAL EFFERENTS

at least two organizational differences should be apparent in HRP histochemical procedures. (1) The cell bodies of origin of the projections to regio superior should be labeled bilaterally while those to the fascia dentata should be labeled almost exclusively ipsilaterally. (2) Since the length of the entorhinal terminal zone along the granule cell dendrites is small relative to a n HRP injection site, the cells of origin of' the projections to the fascia dentata should be labeled throughout the medio-lateral extent of the entorhinal cortex (Steward, '76). However, since a n injection would be expected to label only a limited extent of the CA2-subicular axis of regio superior, (fig. l), the cells of origin of this projection system should be labeled in only a portion of the entorhinal area, corresponding to the position of the injection site in regio superior (Steward, '76). The cell bodies of origin of entorhinal

349

afferents to the hippocampal formation have previously been investigated with horseradish peroxidase (HRP) histochemical methods (Lavail et al., '73; Segal and Landis, '74),but in these studies, no attempt was made to differentiate between the various components of the entorhinal projection system. Lavail et al. ('73), report that injections which label primarily the dentate gyrus and regio inferior result i n the retrograde labeling of cells in layers I and I1 of the entorhinal cortex, but primarily layer 11. Segal and Landis ('74), however, report that injections which label the dentate gyrus and regio superior (their fig. 1) result in the labeling of cells in layers 11, and to a lesser extent layer 111. In the present study, when the entorhina1 terminal fields in both regio superior and the fascia dentata were included i n the injection site, the pattern of retrograde labeling of cells in the entorhinal area was

Fig 1 Organization of entorhinal afferents along the dendrites of granule cells of the area dentata (AD), and the pyramidal cells of regio superior (RS) and regio inferior (RI) Summary from Steward ('76)

350

OSWALD STEWARD AND SHEILA A. SCOVILLE

contralateral

1m m

ipsilateral

A B

C

Fig. 2 The pattern of retrograde labeling of cells i n the entorhinal area is diagramatically illustrated for animal H-D-I at four dorso-ventral levels (encompassing approximately the dorsal half of the entorhinal region). For a complete cytoarchitectonic description of these levels, see Steward ('76), figure 1. The border between layer I1 a n d layer 111 is indicated by the line. In the lower portion of the figure. the injection site i s reconstructed for this case as it would appear i n coronal section. The lines a t A, B, and C indicate the levels of the horizontal sections of figure 4. aepm. area entorhinalis pars medialis, aepi, area entorhinalis p a r s intermedialis. aepl. area entorhinalis p a r s lateralis; FD, fascia dentata: RS, regio superior; RI, regio inferior. Filled circles and triangles indicate heavily labeled cells, while open circles and tri. angles indicate lightly labeled cells.

CELLS OF ORIGIN OF ENTORHINAL EFFERENTS

35 1

A

Fig. 3 The injection site i n the entorhinal terminal zones i n the fascia d e n t a t a a n d reglo superior is illustrated for a n i m a l H-D-1. FD, fascia d e n t a t a ; RS, regio superior of the hippoc a m p u s ; HF, hippocampal fissure. A schematically illustrates t h e position of the injection site. while B provides a photographic enlargement of t h e center of the injection. The dorso-ventral extent of t h e injection site is shown i n figure 4.

similar to that described by Segal and Landis ('74). Figures 2 4 illustrate one of a total of eight cases with injections of this type (animal H-D-2, 48 hours post-injection survival). The injection in the case illustrated was centered in the stratum moleculare of the fascia dentata, but spread into the adjacent stratum lacunosum-mo-

leculare of regio superior (see fig. 2 for a diagramatic reconstruction of the injection site as it would probably appear in coronal section, and figs. 3 and 4 for its actual appearance in the horizontal sections which were utilized). Although i t is difficult in dorsal horizontal sections to determine the exact position of the injection in regio su-

352

OSWALD STEWARD AND SHEILA A. SCOVILLE

5

i

Y

CELLS O F ORIGIN OF E N T O R H I N A L E F F E R E N T S

perior (owing to the fact that regio superior is sharply folded near this point, as illustrated in figs. 3 and 4) the coronal reconstruction of figure 2 indicates that the injection is centered i n a portion of regio superior relatively distant from the CAlCA2 transition, and should thus result i n preferential labeling of the cells of origin of the projections to regio superior i n the lateral entorhinal region (fig. 1). The pattern of retrograde labeling of cells i n the entorhinal region following this injection is illustrated i n figures 2 and 5. Ipsilateral to the injection, the HRP reaction product is evident in cells in layer I1 (which appear to be the stellate cells which inhabit this layer), and cells i n layer I11 (which appear to be of pyramidal form, and are probably the medium sized pyramids described in this layer by Lorente de NO, '34). The layering scheme is as defined by Lorente de NO ('34). The cells of layer I1 are labeled throughout the mediolateral extent of the entorhinal area, while the cells of layer I11 are preferentially labeled in the lateral portions of the entorhinal region, particularly in ventral segments (fig. 2). In the entorhinal area contralateral to the injection (figs. 2, 5 , 6), the reaction product was found almost exclusively in cells of layer 111 (but see exceptions discussed below). Thus, cells of layer I1 were labeled almost exclusively ipsilaterally, and were found throughout the medio-lateral axis of layer 11, while the cells of layer I11 were labeled bilaterally, and in the present case, were preferentially located in the lateral entorhinal region. Labeled cells were not observed beyond the boundary of the entorhinal area with the adjacent neocortex (the boundary coinciding largely with the rhinal fissure) although labeled cells of layer 111 did extend into the depths of the rhinal fissure, into an area which might partially correspond to what Van Hoesen and Pandya ('75a) have defined as the prorhinal area in the monkey (Steward, '76). In addition to these major populations of cells, a very few labeled cells were found which did not conform to these general patterns. For example, a very few lightly labeled cells were found in layer I1 of the entorhinal region contralateral to the injection (figs. 2, 5, 6). These rare cells which were labeled i n the entorhinal region con-

353

tralateral to the injection could be divided into two populations. One was localized in the same medio-lateral position as the deeper cells of layer I11 (fig. 5C, arrows). These tended to lie near the border of layers I1 and 111, and were labeled to approximately the same extent as the deeper cells. A second very rare population of lightly labeled cells was found predominantly in the most medial portion of layer 11, and was concentrated in the most dorsal portion of the entorhinal area (figs. 2 , 6, arrows). This general pattern of retrograde labeling was observed i n eight animals with injections similar to the one illustrated i n figures 2-4 for animal H-D-1. In two animals however, with apparently adequate injections, there was no evidence of retrograde transport of HRP to a n y of the cells of origin of hippocampal afferents (including septa1 and commissural). In these cases, however, the HRP reaction product was evident at the site of the injection, discounting the possibility that the histochemical method failed. The reasons for the absence of retrograde transport in these cases are unclear at this time. Thus, HRP injections which label the entorhinal synaptic fields in both the fascia dentata and regio superior result in bilateral labeling of cells i n layer 111, but almost exclusively ipsilateral labeling of cells i n layer I1 (with the exceptions discussed above). Since the entorhinal projections to regio superior are bilateral (Steward et al., '73, '74; Steward, '76), while the entorhinal projections to the fascia dentata are almost exclusively ipsilateral (Zimmer and Hjorth-Simonsen, '75; Steward, '76), the pattern of retrograde labeling suggests that the entorhinal projections to regio superior originate from cells in layer 111, while those to the fascia dentata originate from layer 11. The few labeled cells i n layer I1 of the dorsal portion of the entorhinal region contralateral to the injections could represent the cells of origin of the sparse crossed projection from the entorhinal area to the rostra1 fascia dentata (Zimmer and Hjorth-Simonsen, '75; Goldowitz et al., '75; Steward, '76), or could be related to the cells of origin of the crossed projections to regio superior in the deeper strata. This interpretation of the differential origin of the projections to the fascia den-

354

OSWALD STEWARD A N D SHEILA A. SCOVILLE

CELLS O F ORIGIN OF E N T O R HI N A L EF F EREN TS

355

tata and regio superior gains support from tions of figures 3 and 9C, it is relatively additional cases with selective injections. restricted to a portion of the fascia dentata In two cases (H-1, 24 hours survival, and relatively distant from regio superior. In H-2, 48 hours survival) the injections were fact, while it is not evident in the horizoncentered i n the C A I region of regio supe- tal sections, the coronal reconstruction ilrior, relatively distant from the CAl/CA2 lustrates that the injection is actually centransition (see fig. 7 for a coronal recon- tered in the ventral blade of the rostral struction of the injection site, and figs. 8 fascia dentata (or the free blade according and 9C for its appearance i n the horizontal to the terminology of Hjorth-Simonsen and sections). In both of these cases, there was Zimmer, '75). Thus, the center of the inno evidence of spread of HRP into the jection site is well separated (by approxistratum moleculare of the fascia dentata, mately 750 pm) from the nearest entosince the hippocampal fissure (which had rhinal terminal field in regio superior not been punctured by the microsyringe) (although the syringe did pass through the seemed to serve as a n effective barrier to entorhinal terminal field in regio superior, the diffusion of HRP into the fascia den- and there could thus have been some slight tata. In these cases, as illustrated for H-1 spread of HRP along the track of the by figures 7 and 9A, the HRP reaction prod- syringe). In these cases, with relatively selective uct was found almost exclusively i n the cells of layer 111, particularly in the lateral injections of the fascia dentata, the HRP entorhinal area, and these were labeled reaction product was found almost exclubilaterally. There were a very few cells sively i n cells of layer I1 (figs. 9B, lo), and labeled in layer I1 (near the border with these cells were labeled almost exclusively layer 111, figs. 7, 9). These appeared bilat- ipsilaterally (fig. 10). Careful searching erally, but appeared to have the same me- also revealed a very few lightly labeled cells dio-lateral distribution as the labeled cells in layer I1 of the entorhinal region contrain layer 111. No labeled cells were found in lateral to the injection, but this rare populayer I1 in the most medial portion of the lation was confined to the most medial entorhinal cortex immediately adjacent to portions of layer I1 in the dorsal most segthe parasubiculum. These observations sug- ments of the entorhinal region. Since the gest that the small population of cells i n very sparse population of cells i n the melayer I1 which were labeled following the dial most portion of layer I1 (adjacent to selective injections of regio superior, and the parasubiculum) was labeled (but very which have the same medio-lateral distri- lightly) following injections which included bution as the cells of layer 111, are perhaps the entorhinal synaptic fields in both the atypically located cells of origin of the pro- fascia dentata and regic superior (fig. 2 ) and following injections which were relajections to regio superior. In two cases (D-1, and D-2, 48 hours tively selective for the fascia dentata, it is survival) the injections were centered i n probably this very sparse population which the stratum moleculare of the fascia den- gives rise to the sparse crossed pathway tata, with minimal spread into the stratum from the entorhinal area to the contralatlacunosum-moleculare of regic superior eral fascia dentata. The present observations also provide (see fig. 10 for a coronal reconstruction of the injection site, and figs. 11 and 9D for additional information on the topographic its appearance i n the horizontal sections). organization of the two projection systems Note that while the injection site in this in the dorso-ventral domain. Hjorth-Simoncase is much more diffuse than the injec- sen and Jeune ('72) have reported that projections from the medial entorhinal area Fig. 5 Labeled cells in layer 111 of the entoto the hippocampal formation are organized rhinal region contralateral to t h e injection shown in a dorso-ventral dimension, with dorsal in figures 2 4 , C is a x 100 photograph of the segments of the entorhinal area projecting labeled cell laminae shown in A. For the purpose to more rostral portions of the hippocampal of comparison, B a n d D illustrate the pattern of retrograde labeling ipsilateral to the injection. The formation, and ventral sites in the entoarrows in C indicate the very few cells which are rhinal area projecting to the caudal and found in layer I1 contralateral to the injection, ventral hippocampal formation. A topowhich have the s a m e medio-lateral distribution as graphic organization of this sort was apthe labeled cells of layer 111.

356

O S W A I B S T E W A R D A N D S H E I L A A. SCOVILLE

Fig. 6 ' l h c ~ second small population of labeled cells i n l a y e r I1 c o n t r a l a t e r a l to t h e injection I S i l l u s t r a t c d 'These a r e found 111 t h e most m e d i a l portions of l a y e r 11. i m m e d i a t e l y a d j a cent to t h e parasubiculuiii. T h e X 400 p h o t o g r a p h of B i s t a k e n from t h e a r e a circumscribed by t h e rectangle i l l A T h e a r r o w s i n d i c a t e t h e few lightly labeled cells w h i c h m a y be found in t h i s zone

parent in the present material. First, all injections of the present series were restricted to approximately the rostral half of the hippocampal formation, and labeled cells were found in approximately the dorsal half of the entorhinal region (fig. 2). In addition, within the series of injections of the present study, there was also evidence for a dorso-ventral organization. For example, in the case illustrated in figure 10, where the HRP injection was localized to the most rostral (septal) tip of the hippocampal formation, labeled cells were found only in the dorsal three levels illustrated in the present series (fig. 10). With injections centered slightly more caudally, however, the labeled cells extended more ventrally into the entorhinal area (figs. 2, 7). All of the preparations indicate, however, that while this organization is present, it is not highly discrete, since the relatively limited injections illustrated in figures 2 and 7 resulted in the retrograde labeling

of cells throughout approximately the dorsal half of the entorhinal area. With the available cases, the dorso-ventral organization seemed comparable for cells in both layers I1 and 111, and in addition, the dorsoventral organization of' the crossed projections to regio superior seemed comparable to the ipsilateral projection systems, since cells in layer I11 were labeled at similar dorso-ventral levels both ipsilateral and contralateral to the injections. In addition, there was a tendency for cells in layer I1 to be labeled somewhat further ventrally in the lateral entorhinal area than in the medial, which is consistent with the observations of Segal and Landis ('74). The very lightly labeled cells in the medial portion of layer I1 contralateral to injections which included the fascia dentata were concentrated in the dorsal most portion of the entorhinal area (fig. 2). This fact is consistent with the known organization of the very sparse crossed projection from the en-

'

1

. i

CELLS OF ORIGIN OF ENTORHINAL EFFERENTS

contralateral

1 mm

35 7

ipsi lateral

\ -

Fig. 7 The pattern of retrograde labeling of cells i n the entorhinal area is diagramaticaliy illustrated for a n i m a l H - I . Conventions a n d abbreviations a r e a s for figure 2. The injection site for H - l is reconstructed as it would appear i n coronal section i n the lower portion of the figure. A, B, and C indicate the levels of the horizontal sections of figure 8.

torhinal region to the fascia dentata (Zimmer and Hjorth-Simonsen, '74; Goldowitz et al., '75). Since the amount of reaction product in these cells was always very

slight, and since cells with this type o f l a beling were quite rare, negative evidence is of questionable value, however. The absence of labeled cells in the pre-

358

OSWALD STEWARD AND SHEILA A . SCOVILLE

d Y

1 *

c

CELLS OF ORIGIN OF ENTORHINAL EFFERENTS

subiculum and parasubiculum following injections into the hippocampal formation suggests that these regions do not give rise to hippocampal afferents. However, it is possible that these two regions are incapable of exhibiting retrograde transport of HRP. To test this hypothesis, two animals received injections of 0 . 5 4 . 7 pl of HRP into the entorhinal region of one hemisphere, a known target area for presubicular and parasubicular fibers (Blackstad, '56; Shipley, '74, '75). In both animals with unilateral HRP injections into the entorhinal region, cells in both the contralateral presubiculum and parasubiculum clearly exhibited retrograde transport of HRP (fig. 12). In addition, labeled cells were also found in layer 111 of the entorhinal region proper. Since the cells in both the presubiculum and parasubiculum are clearly capable of exhibiting retrograde transport of HRP following injections into the contralateral entorhinal region, the failure to label these cell populations following injections into the ipsilateral hippocampal formation provides strong evidence for the absence of any significant projection from the presubiculum or parasubiculum to the hippocampal formation. Despite the fact that the entorhinal region has been reported to send a substantial projection to the dorsal thalamus (Ramon y Cajal, '1 1) possibly from the presubiculum or the deep cell layers (IV-V) of the entorhinal region proper, no labeled cells were found in the entorhinal region following injections which either included, or were restricted to various portions of the dorsal thalamus. It should be noted, however, that no systematic effort was made to inject thalamic sites, and that the available cases resulted from accidental labeling. If the deep cells of layers IV-V do project to the thalamus, their labeling could have been obscured by the diffuse band of HRP reaction product in layers IV and V (fig. 5). Alternatively, the post-injection survival time could be inappropriate for this connection. The diffuse band of labeling in layers IV-V is itself a peculiarity, since it has the appearance of a band of HRP containing terminals, yet has a medio-lateral distribution which corresponds quite closely with the labeled cells of layer I11 (fig. 5). Whether this labeled zone might represent orthograde transport by a pro-

359

jection from the hippocampal formation back to the entorhinal region (Hjorth-Simonsen, '72; Shipley and Sprrensen, '75), or whether the labeling is due to HRP i n the proximal axonal or dendritic segments of the labeled cells of layer 111 remains to be seen. DISCUSSION

The pattern of retrograde labeling of cells in the entorhinal area following injections of horseradish peroxidase (HRP) into the hippocampal formation suggests that the entorhinal projections to the fascia dentata and to regio superior originate from two quite distinct populations of cells. Injections of HRP which label the entorhinal terminal fields i n both the fascia dentata and regio superior result in hilateral labeling of cells in layer 111, but almost exclusively ipsilateral labeling of cells in layer 11. Since the entorhinal projections to regio superior are bilateral (Steward et al., '73, '74; Steward, '76) while the projections to the fascia dentata are almost exclusively ipsilateral, these results strongly suggest that the bilateral projections to regio superior originate from cells in layer 111, while the almost exclusively ipsilateral projections to the fascia dentata originate from cells in layer 11. This organization is schematically illustrated in figure 13. A very small number of the total population of cells in layer I1 are also lightly labeled contralateral to an injection which includes the fascia dentata. This small population of cells may represent the cells of origin of the sparse crossed projection to the fascia dentata (Goldowitz et al., '75; Steward, '76). It should be noted that the presumed cells of origin of this sparse normal crossed pathway to the fascia dentata are so rare that they would occasion little comment save for the fact that this projection is of considerable interest, owing to its marked capacity for proliferation i n response to unilateral entorhinal lesions (Steward et al., '73, '74, '76; Zimmer and Hjorth-Simonsen, '75). The above interpretations are reinforced by the fact that the topographic patterns of retrograde labeling correspond to the predicted topographic organization of the two entorhinal projection systems. As indicated previously, since the length of the entorhinal terminal zone along the granule

360

OSWALD STEWARD AND SHEILA A. SCOVILLE

CELLS OF ORIGIN O F ENTORHINAL EFFERENTS

cell dendrites is quite small (on the order of 200 Fm) i n relation to the diameter of a typical HRP injection site, the cells of origin of the projections to the fascia dentata should be labeled throughout the medio-lateral extent of the entorhinal area (Steward, '76). As is evident from figures 4 and 5 , the cells of layer I1 (which on the basis of the above evidence are the cells of' origin of dentate bound afferents) are indeed labeled throughout the medio-lateral extent of the entorhinal area. Labeled cells of layer I1 are, however, occasionally found further ventrally in the lateral portion of' the entorhinal area than i n the medial, which may be consistent with the notion that the medio-lateral axis of the entorhinal region is tilted slightly from a perfect horizontal plane (Steward, '76). In the case of the projections to regio superior, the injections of the present series labeled only a limited extent of the CA2-subicular axis of regio superior (fig. I ) , and the cells of origin of this projection system were labeled in only a portion of the entorhinal area, corresponding to the position of the injection site i n regio superior (Steward, '76). Thus, the injections of the present series resulted i n a topographically specific pattern of labeling of cells in layer I11 which corresponded to the distribution which would be expected for the cells of origin of the projections to regio superior. These interpretations are still further reinforced by the cases with selective injections. While it is clearly difficult, if not impossible to define the effective limits of a n injection site (beyond which retrograde transport is impossible) the fact that apparently selective injections result i n exactly the pattern of retrograde labeling predicted by the more extensive injections lends additional credence to the above in-

__

-

Fig. 9 A illustrates the injection site for case H - I , which i s centered i n regio superior. The p a t tern of retrograde labeling of cells i n the entorhinal area for case H - l is illustrated i n C. Note that the labeled cells a r e found predominantly i n layer 111. B illustrates the injection site for r a s e D-1, which is centered i n the stratum moleculare of the fascia dentata. The pattern of retrograde labeling of cells i n the entorhinal area for case D-1 i s illustrated i n D. Note that i n this case, t h e labeled cells i n layer I1 i n the left h a n d portion of the photograph of D represents the expansion of layer I1 a t its point of intersection with the parasubiculum (figs. 2, 5 ) . In all photographs, the left h a n d side is the more medial.

36 1

terpretations. It should be noted that our injections appear somewhat more restricted (figs. 2, 3, 9) than would be expected on the basis of previous studies in the hippocampal formation of the rat (Lavail et al., '73, Segal and Landis, '74). We are unable at present to account for this fact, and can only point out two procedural differences between our methods and those of previous investigators, (1) the use of distilled water as the vehicle for the HRP injections, and (2) the use of 10% formalin as a fixative. While the above interpretations are completely consistent with the observations, some confusion could result if damaged fibers of passage transported HRP in a retrograde fashion. For example, the track of the syringe passes through the dorsal psalterium (the dorsal hippocampal commissure) which carries the commissural connections between the entorhinal areas of each hemisphere (Blackstad, '56). If these damaged fibers transported HRP in a retrograde fashion, the interpretation of the bilateral labeling of cells in layer 111 could be confused, since these give rise to at least some commissural projections (fig. 12). This interpretation receives no support from the observations, however, since the major commissural component of the dorsal psalterium arises from the presubiculum and parasubiculum (Blackstad, '56, and fig. 12), and these are not labeled following HRP injections into the hippocampal formation. Similarly, the cells of origin of other projections which travel through the hippocampal formation (such as the presubicular projections to the thalamus, Shipley, '74, and the subicular projections to the mammillary bodies, Swanson and Cowan, '75, both of which travel in the fornix) remain unlabeled following HRP injections into the hippocampal formation. Thus, it is extremely unlikely that transport of HRP by damaged fibers of passage could confuse the interpretation of the present observations i n any way. Some other reservations also exist regarding the interpretation of the results. For example, the differential labeling of cells ipsilateral and contralateral to the injection could conceivably result from some peculiar selectivity in the ability to transport HRP in a retrograde fashion, and the interpretation of the apparently selective injections could be confused by slight

362

OSWALD STEWARD AND SHEILA A. SCOVILLE

contralateral

1mm

ipsilateral

Fig. 10 The pattern of retrograde labeling of cells i n t h e entorhinal area is diagramatically illustrated for animal D-1. Conventions and abbreviations are a s for figure 2. The injection site for D - l is reconstructed a s it would appear i n coronal section i n the lower portion of the figure. A , B, a n d C indicate t h e levels of the horizontal sections of figure 1 1 .

CELLS OF ORIGIN OF E N T O R H I N A L EFFERENTS

, . 3

a

364

OSWALL) STEWARD AND SHEILA A . SCOVILLE

Fig. 12 1 j R P labeled cells are illustrated i n the presubiculum (PrS) a n d parasubiculum ( P a s ) following a n injrction of HRP into t h e contralateral entorhinal region. Labeled cells are also present i n layer I11 of area entorhinalis ( a e ) proper.

spread of HRP at the injection site into nearby terminal fields. In addition, it is also possible that other cells in the entorhinal area (perhaps in the deeper strata) also give rise to hippocampal afferents, but fail to transport HRP in a retrograde fashion. This latter hypothesis is improbable, however, in view of the reported absence

of such projections from classical Golgi studies (Ramon y Cajal, '11; Lorente de NO, '33). Thus, while these possibilities cannot be completely excluded, all of the present observations strongly suggest the same conclusion, that the entorhinal projections to the fascia dentata originate from cells in layer 11, while the projections to

CELLS OF ORIGIN O F E N T O R H I N A L E F F E R E N T S

365

Fig. 13 Organization of t h e entorhinal projection systems a s suggested by the present results, and those of previous studies (Steward, '76). a e p m , area entorhinalis p a r s medialis, aepi, area entorhinalis pars intermedialis; aepl, area entorhinalis pars lateralis; FD, fascia dentata, RS, regio superior.

regio superior originate from cells in layer 111. Since it was not possible to selectively label the entorhinal terminal field in regio inferior, the cells of origin of this component of the entorhinal projection system remain in doubt. Because, however, the entorhinal projections to regio inferior have a topographic organization comparable to the dentate projection system, it may be that the afferents to regio inferior originate from the same population of cells (in layer 11) which give rise to dentate bound projections. To summarize the characteristics of the two entorhinal projection systems which have been elucidated (fig. 13), the entorhinal projections to the fascia dentata apparently originate predominantly or exclusively from cells in layer 11, which on the basis of their position i n the entorhinal area, and their appearance in HRP histochemical preparations, are probably the population of stellate cells described i n this layer by Ramon y Cajal ( ' l l ) , and Lorente de No ('33). These cells send their axons

almost exclusively ipsilaterally , and terminate in proximo-distally organized laminae along the dendrites of dentate granule cells (Hjorth-Simonsen, '72; Hjorth-Simonsen and Jeune, '72; Steward, '76). Entorhinal projections to regio superior on the other hand, apparently originate predominantly or exclusively from cells in layer 111, which on the basis of their position in the entorhinal area, and their appearance in HRP histochemical preparations, are probably the medium sized pyramidal cells of layer 111 (Ramon y Cajal, '11; Lorente de No, '33). This cell population projects bilaterally, and terminates not in proximodistally organized laminae, but rather i n a topographically organized fashion across the distal dendritic field of regio superior (fig. 1 , Steward, '76). Thus, these two pathways originate from different cells, are topographically organized i n the mediolateral dimension in a different fashion, and terminate on different types of cells in the hippocampal formation (fig. 13). Their topographic organization in the dorso-ventral dimension is, however, similar.

366

OSWALD STEWARD A N D SHEILA A

The fact that the projections to regio superior originate from a different population of cells in the entorhinal area than the projections to the fascia dentata may help to explain why this projection has been inconsistently observed in experimental investigations. While some authors report heavy degeneration within the stratum lacunosum-moleculare of regio superior following entorhinal lesions (Blackstad, '56, '58; Raisman et al., '65; Steward et al., '73, '74; Steward, '76) other investigators report a negligible projection (Nafstad, '67). A part of the reason for this discrepancy may arise from the differential topographic organization of the projection system to regio superior (Steward, '76). The present results suggest a second possible reason. Since the population of' cells giving rise to the projections to regio superior lie deeper in the entorhinal area than the cells of origin of the projections to the fascia dentata, the pathway to regio superior may be spared following superficial lesions such as those utilized by Nafstad for his electron microscopic investigations (see Nafstad, '67, his fig. 1). These observations on the differential origin of entorhinal afferents to the fascia dentata and regio superior of the hippocampal formation also provide several new insights into the significance of the laminated pattern of afferent input to the entorhinal region itself. Lorente de NO ('33) demonstrated that the dendrites of the stellate cells of layer I1 ramify within layers I and 11, but do not extend into layer 111. The dendrites of layer I11 pyramidal cells, on the other hand, ramify within layers I and 111, but give off few branches within layer 11, suggesting few synaptic contacts within this layer (see fig. 1 3 for a diagramatic illustration of this organization). In fact, on the basis of Golgi material, Lorente de NO ('33) suggested that the pyramidal cells probably receive 1,000-fold more synaptic inputs within layers I and I11 t h a n in layer 11. These Golgi observations thus suggest (1) that afferents terminating in layer I could innervate both cell types, ( 2 ) afferents terminating in layer I1 terminate preferentially or exclusively on stellate cells, and, (3) afferents terminating i n layer 111 would innervate the basal and proximal apical dendritic segments of the pyramidal cells of this layer.

SCOVILLE

Entorhinal a f f i r e n t s f r o m retrohippocampnl structures The pattern of termination of several afferents to the entorhinal region can be correlated with this lamination of dendritic arborization. Blackstad ('56) demonstrated that presumably homotypical commissural afferents from the contralateral entorhind area terminated preferentially within layer I, particularly within pars medialis of the entorhinal area, while a heavy projection to layer I1 was evident when the lesions involved non-homotypic areas of the contralateral entorhinal region, particularly the parasubiculum. This projection of the parasubiculum is apparently bilateral (Blackstad, '56). More recently, a rather heavy projection to layers I11 and I of the medial entorhinal area from the presubiculum has been described in the cat (Karten, '63), in the guinea pig (Shipley, '74, '75), and in the monkey (Van Hoesen and Pandya, '75a), although in the monkey, no discrimination was made between presubicular and parasubicular projections. A similar projection has been observed from the presubiculum to the entorhinal area i n the rat (E. Geisert, personal communication). In all species which were investigated for crossed projections, including the guinea pig (Shipley, '74, ,759, monkey (Van Hoesen and Pandya, '75a), and rat (E. Geisert, personal communication), the presubicular projections are bilateral, although the ipsilateral component seems to be somewhat larger than the crossed projection. Thus, in the rat at least, the entorhinal cortical region receives at least three inputs from neighboring retro-hippocampal structures, (1) a homotypical commissural projection from the contralateral entorhinal region which terminates in layer I (a lamina which contains the distal dendritic arborizations of both the stellate cells of layer I1 and the pyramidal cells of layer 111), (2) a bilateral projection from the parasubiculum which terminates predominantly in layer I1 (a distribution which is coincident with the main dendritic processes of the stellate cells of layer II), and (3) a bilateral projection from the presubiculum, which terminates primarily in layers I and I1 of the medial portion of the entorhinal region (a distribution which is coincident with the

CELLS OF ORIGIN OF E N T O R H I N A L E F F E R E N T S

367

possibly including the prepyriform cortex and the amygdala itself. This projection has been demonstrated in the rat (Cragg, '61; Powell et al., '65; and Krettek and Price, '74), rabbit (Cragg, '61), cat (Cragg, '61; and Krettek and Price, '74), and monkey (Van Hoesen and Pandya, '75a). This projection seems to distribute primarily to layer I11 in the lateral entorhinal region (Krettek and Price, '74), although a component of this projection is also present in layer I, particularly in the medial entorhinal region (Krettek and Price, '74). Since the projections from the periamygdaloid region distribute preferentially in layers I and 111, they may terminate predominantly on the pyramidal cells which give rise to the projections to regio superior. In turn, the periamygdaloid region has been considered as a relay for olfactory input into the hippocampal formation (Cragg, '61). In addition, the periamygdaloid region apparently gives rise to a direct projection to the subiculum (Cragg, '61; Krettek and Price, '74) thus bypassing the indirect projection via the layer I11 pyramidal cells of the entorhinal region. A second source for non-retrohippocampal input to the pyramidal cells of layer I11 is the orbito-frontal cortex, at least in the monkey (Van Hoesen et al., '75). This cortical region, including Broadman's areas 12, 13, and 14 terminates predominantly in layer I11 of the lateral entorhinal region (area 28b) and also provides a significant source of input for the periamygdaloid regions. Whether this orbito-frontal projection to the entorhinal area also exists in species other than the monkey is not clear at this time. Finally, a large but diffuse projection apparently reaches the entorhinal area from the neighboring transition areas between the entorhinal and neocortices (the prorhinal and perirhinal regions, Van Hoesen and Pandya, '75a). These do not terminate with a distinctive pattern of lamination in the entorhinal area proper, however. These results suggest that the pyramidal cells of layer I11 of the entorhinal region, Other entorhinal afferents which give rise to the multi-synaptic proConsidering the afferents to the ento- jections to the hypothalamus and the Papez rhinal region other than those from neigh- circuit, via regio superior, receive their boring retrohippocampal structures, per- input primarily from the olfactory system haps the most well-documented projection (via the periamygdaloid region), the orbitooriginates in the periamygdaloid region, frontal region (at least in the monkey), and

dendritic arborizations of the pyramidal cells of layer 111). Interestingly, the rather strong homotypic commissural projection from the contralateral entorhinal area is not, apparently, present in the monkey (Van Hoesen and Pandya, '75a) although the presubicular and parasubicular projections are present and are bilateral. If this correlation between the pattern of afferent organization, and the pattern of dendritic arborization does signify the relative contribution of the various afferents to the two cell types, then this would suggest that the stellate cells of layer 11, which give rise to the afferents to the fascia dentata receive a major input from the parasubiculum, but if fibers of presubicular or contralateral entorhinal origin terminate on stellate cells, this input is restricted to the most distal dendritic regions (in layer I). The pyramidal cells of layer 111, however, receive a large projection from the presubiculum, possibly a small projection from the contralateral entorhinal area, and only a very slight input, if any, from the parasubiculum. If the afferents from the presubiculum do terminate predominantly on the pyramidal cells of layer 111, then it is possible that the loop of the Papez circuit (from the hippocampus -+ mammillary bodies + anterior thalamus + presubiculum + entorhinal cortex ---$ hippocampus) may bypass the temporo-dentate circuit and the internal hippocampal circuitry to return directly to the pyramidal cells of regio superior, including CA1 and the subiculum (Steward, '76; and see figs. 1, 13, 14). Since it is apparently the subiculum which in turn gives rise to the projections to the mammillary bodies (Swanson and Cowan, '75; Meiback and Siegel, '75), and since recent anatomical (Hjorth-Simonsen, '73) and electrophysiological evidence suggests that a major output of the CA1 region is to the subiculum (Andersen et al., '72) it may be that the limbic feedback loop of the Papez circuit operates quite independently of the dentate gyrus and regio inferior.

368

OSWALD STEWARD AND SHEILA A. SCOVILLE

lhnsl t

(Olf)

Layer I I I aepl aepm

7------J (Pm4

(OF0

Fig. 14 Thc circuit mediated priiiiarily vial the cells of laver 111 of the entorhinal region is illustrated. Thc striict tires bounded by rectangles represent the Papez circuit. The structures enclosed within the civals represent the presumed inputs to the cells of layer 111. aepl. area entorhinalis pars latt%riilis,at.pm. area entorhirialis p a r s medialis. Sub, subiculum. C A l . CAI field of regio supcrior 01 t h e hippocarnpus, MB. mammillary bodies, AT. anterior t h a l a m u s , Cing. cingulurn: PrS. presubiculum: PRhin. perirhinal a n d prorhinal areas: ObtF. orbitofrontal regions. 011. olfactory regions; PAmg. periamygdaloid regions.

from the Papez circuit itself, via the projections to the entorhinal area from the presubiculum. This entire circuit, as summarized in figure 14 could operate quite independently of the circuit from layer I1 stellate cells to the dentate gyrus. The question remains, from where do the stellate cells of laver I1 receive their afferent input? One source of input to the stellate cells of layer I1 has previously been discussed (the parasubiculum). Unfortunately, it is not clear at this time what the inputs are to the parasubiculum. A second possible source for inputs to the stellate cells has, however, recently been described in the monkey (Van Hoesen and Pandya, '75a) These investigators report that following lesions of areas TF and TH (as defined by Bonin and Bailey, '47) terminal degeneration was apparent in layers I and I1 of the entorhinal region proper. This distribution thus coincides with the dendritic ramifications of the stellate cells of layer I1 Areas TF and TH in turn receive inputs from area 7 of the parietal lobe (a presumed somatosensory association area), area TA of the temporal lobe (a presumed auditory association area), and area OA of the occipital lobe (a presumed visual association area) (Jones and Powell, '70, Seltzer and Pandya, '74, Van Hoesen and Pandya, '75b).

Thus, these inputs from areas TF and TH could form a n important relay between non-olfactory cortical association areas and the granule cells of the fascia dentata, via the stellate cells of layer I1 of the entorhinal region. These relationships are summarized i n figure 15. This circuit, which outputs from the hippocampal formation via the projections from the fascia dentata-regio inferior-septum could operate quite independently of the olfacto-hippocampo-hypothalamic circuit through regio superior. However, opportunities are provided at several synaptic stations for interaction between the two systems (such as the multisynaptic relay from the fascia dentata-regio inferior-regio superior via the Schaffer collateral system). Needless to say, these speculations are based on incomplete evidence at best, and have been developed through the assimilation of data from a variety of species which may well prove to be incomparable. Nevertheless, the concept of two separate and potentially independent pathways from the entorhinal region through the hippocampal formation (one primarily involved in olfactory and hypothalamic circuitry, and one receiving input indirectly from non-olfactory cortical association areas) deserves further investigation.

CELLS OF ORIGIN OF ENTORHINAL EFFERENTS

Fig. 1 5 The circuit mediated primarily via the cells of layer I1 of the entorhinal region is illustrated. PRhin, perirhinal and prorhinal areas; TFTH, areas T F and TH a s defined by Bonin a n d Bailey ('47); aepl, area entorhinalis pars lateralis, aepm, area entorhinalis pars medialis; P a s , parasubiculum, FD, fascia dentata; CA3, CA3 field of regio inferior of the hippocampus; CAI, CA1 field of regio superior of the hippocampus; Sep, septum; SC, Schaffer collateral system. The laminated pattern of termination of entorhinal afferents to the fascia dentata is illustrated by the schematic granule cell. The dotted arrow from CA3 to CA1 indicates one potential site for interaction between the olfacto-hippocampo-hypothalamic pathways, (fig. 14) a n d the pathways illustrated i n the present figure. ACKNOWLEDGMENTS

Supported in part by an Alfred Sloan Foundation Grant (72-11-4), and in part by USPHS Research Grant 1 R 0 1 NS1233301 to 0. Steward. LITERATURE CITED Andersen, P., B. H. Bland a n d J. D. Dudar 1972 Organization of the hippocampal output. Exp. Brain Res., 17: 152-168. Blackstad, T. 1956 Commissural connections of the hippocampal region i n the rat, with special reference to their mode of termination. J. Comp. Neur., 105: 417-537. 1958 On the termination of some afferents to the hippocampus a n d fascia dentata. An experimental study i n the rat. Acta Anat., 35: 202-214. Bonin, G . von, a n d P. Bailey 1947 The Neocortex of Macaca mulatta. Univ. of Ill. Press, Urbana, 111.. 136 pp.

369

Cragg, B. G. 1961 Olfactory a n d other afferent connections of the hippocampus i n the rabbit, rat, a n d cat. Exp. Neur., 3: 5 8 8 4 0 0 . Goldowitz, D., W. F. White, 0. Steward, C. W. Cotman a n d G . Lynch 1975 Anatomical evidence for a projection from t h e entorhinal cortex to the contralateral dentate gyrus of the rat. Exp. Neur., 47: 4 3 3 4 4 1 . Graham, R. C., a n d M. J. Karnovsky 1966 The early stages of absorbtion of injected horseradish peroxidase i n the proximal tubules of mouse kidney: ultrastructural cytochemistry by a new technique. J. Histochem. Cytochem., 14: 291299. worth-Simonsen, A . 1971 Hippocampal efferents to t h e ipsilateral entorhinal a r e a : An experimental study i n the rat. J. Comp. Neur., 1 4 2 : 41 7-438. 1972 Projection of the lateral part of the entorhinal area to the hippocampus a n d fascia dentata. J. Comp. Neur., 146: 219-232. 1973 Some intrinsic connections of the hippocampus in the r a t : An experimental analysis. J . Comp. Neur., 147: 145-162. worth-Simonsen, A,, a n d B. Jeune 1972 Origin a n d termination of the hippocampal perforant path in the rat studied by silver impregnation. J . Comp. Neur., 144: 215-232. Karten, H. J. 1963 Projections of the parahippocampal gyrus of the cat. Anat. Rec., 145: 247248. Krettek, J . E., a n d J . L. Price 1974 Projections from the amygdala to the perirhinal a n d entorhinal cortices a n d the subiculum. Brain Res., 71 : 150-154. Lavail, J. H., K . R. Winston a n d A. Tish 1973 A method based o n retrograde intra-axonal transport of protein for identification of cell bodies of origin of axons terminating within the CNS. Brain Res., 58: 4 7 0 4 7 7 . Lorente de NO, R. 1933 Studies on the structure of the cerebral cortex. I . The area entorhinalis. J. Psychol. Neurol. (Leipzig), 45: 381438. 1934 Studies o n the structure of the cerebral cortex. 11. Continuation of the study of the Ammonic System. J. Psychol. Neurol. (Leipzig), 46: 113-177. Maiback, R. C., a n d A. Siege1 1975 The origin of fornix fibers which project to the mammillary bodies in the rat; a Horseradish Peroxidase study. Brain Res., 88: 508-512. Nafstad, P. H. J . 1967 An electron microscope study o n t h e termination of t h e perforant path i n the hippocampus a n d fascia dentata. 2.Zellforsch., 76: 5 3 2 5 4 2 . Powell, T. P. S . , W. M. Cowan a n d G. Raisman 1965 The central olfactory connections. J. Anat. (London), 99: 791-813. Raisman, G., W. M. Cowan a n d T. P. S. Powell 1965 ' The extrinsic afferent commissural a n d association fibers of the hippocampus. Brain, 88: 963-996. Ramon y Cajal, S. 1911 Histologie du Systeme Nerveus de l'homme et des vertebres. T. 2. I n stituto Ramon y Cajal, Madrid 1955, 995 pp. Segal, M., a n d S. Landis 1974 Afferents to the hippocampus of the rat studied with the method

370

OSWALD STEWARD AND SHEILA A. SCOVILLE

of retrograde transport of horseradish peroxidase. Brain Res., 78: 1-15. Shipley, M. T. 1974 Presubiculum afferents to the entorhinal area and the Papez circuit. Brain Res., 67: 162-168. 1975 The topographical and l a m i n a r organization of the presubiculum’s projection to the ipsi- and contralateral entorhinal cortex i n the guinea pig. J. Comp. Neur., 160: 127-146. Shipley, M. T., and K. E. S ~ r e n s e n 1975 Evidence for a n ipsilateral projection from the subiculum to the deep layers of the presubiculum and entorhinal cortices i n the guinea pig. Exp. Brain Res. Suppl., 23: 190 (abstract). Steward, 0. 1976 Topographic organization of the projections from the entorhinal area to the hippocampal formation of the rat. J. Comp. Neur.. 167: 285-314. Steward, O . , C. W. Cotman and G . Lynch 1973 Re-establishment of electrophysiologically functional entorhinal cortical input to the dentate gyrus deafferented by ipsilateral entorhinal lesions. Exp. Brain Res., 18: 3 9 6 4 1 4 . - 1974 Growth of a new fiber projection in the brain of adult rats: re-innervation of the dentate gyrus by the contralateral entorhinal cortex following ipsilateral entorhinal lesions. Exp. Brain Res., 20: 4 5 4 6 .

Steward, O . , C. W. Cotman and G. Lynch 1976 A quantitative autoradiographic a n d electrophysiological study of the reinnervation of the dentate gyrus by the contralateral entorhinal cortex following ipsilateral entorhinal lesions. Brain Res., i n press. Swanson, L. W., and W. M. Cowan 1975 Hippocampo-hypothalamic connections: origin i n subicular cortex, not Ammon’s horn. Sci., 189; 303-304. Van Hoesen, G. W., and D. N. Pandya 1974a Some connections of the entorhinal (area 28) a n d perirhinal (area 35) cortices of the rhesus monkey. I. Temporal lobe afferents. Brain Res., 95: 1-24. 1975b Some connections of the entorhinal (area 28) a n d perirhinal (area 35) cortices of the rhesus monkey. 111. Efferent connections. Brain Res., 95: 39-59. Van Hoesen, G. W., D. N. Pandya and N . Butters 1975 Some connections of the entorhinal (area 28) and perirhinal (area 35) cortices of the rhes u s monkey. 11. Frontal lobe afferents. Brain Res., 95: 25-38. Zimmer, J., and A. Hjorth-Simonsen 1975 Crossed pathways from the entorhinal area to the fascia dentata: 11. Provokable in rats. J . Comp. Neur., 161: 71-102.