Commissural and intrinsic connections of the rat hippocampus.

Cornrnissural and Intrinsic Connections of the Rat Hippocampus S. LAURBERG Institute of Anatomy, University of Aarhus, DK-8000 Aarhus C, Denmark

ABSTRACT The commissural and intrinsic connections of the hippocampus were studied using the Fink-Heimer method and the horseradish peroxidase (HRP) uptake technique. A conspicuous septo-temporal gradient was found of the density of the commissural projection that passes through the psalterium ventrale to the Ammon’s horn. The degeneration resulting from transection of the psalterium ventrale was most dense in the septal tip and decreased towards the temporal tip. The commissural and ipsilateral connections from the hilus fasciae dentatae (CA4) and regio inferior (CA3EA2) were found to terminate in different parts of the hippocampus. The hilus fasciae dentatae gave rise to ipsilateral and commissural projections to the dentate area only. The regio inferior has ipsilateral and commissural projections to the Ammon’s horn. A specific termination pattern was found of the projection from regio inferior to stratum radiatum of both the ipsilateral and contralateral regio superior (CA1) and regio inferior (CA2/CA3). At levels temporal to the lesion, the projection is primarily to the superficial part of stratum radiatum, while a t levels septal to the lesion the terminal zone occupies the deep part of the layer. This pattern was not related to the position of the cells of origin, along the septotemporal or subiculo-dentate axes. In general, the commissural projections showed the same degree of septotemporal divergence as the ipsilateral projections. The only major difference in the terminal fields of the two sets of projections to the Ammon’s horn was that the terminal zone of the commissural projection to stratum oriens was always more dense than that of the ipsilateral projection to this layer, while an inverse gradient was seen in stratum radiatum. The projections from the septal and middle dorso-ventral parts of regio inferior differed. The temporal spread of the projections from the septal part was large while that from the projections arising a t middle dorso-ventral levels was more restricted. Moreover, a longitudinal association path interconnecting different parts of the regio inferior along the septo-temporal axis was seen to arise only from the cells in the septal parts of the regio inferior. Each part of the regio inferior projected to all parts of stratum radiatum and oriens of the contralateral Ammon’s horn. However, the projection to the contralateral regio inferior was most dense at the site homotopic to that lesioned. The ventricular part of regio inferior projected primarily to the contralateral stratum oriens of the Ammon’s horn, while the part adjacent to the dentate area mostly supplied stratum radiatum. The commissural fibers to the hippocampus have been described in many studies. As a result of the works by Blackstad (‘56),Raisman et al. (‘651, Laatsch and Cowan (‘671, Blackstad et al. (‘70), Gottlieb and Cowan (‘731, AnJ. COMP. NEUR. (1979)184: 685-708.

dersen et al. (‘73), and Segal and Landis (‘74), i t is well established that the commissural fibers from the hippocampus originate in regio inferior and hilus fasciae dentatae and that the fascia dentata and the regio superior do

685

686

S. LAURBERG

not project to the contralateral hippocampus. The fibers cross through the psalterium ventrale and terminate in approximately the inner one third of the molecular layer of t h e fascia dentata, in the hilus fasciae dentatae, and in stratum oriens and radiatum of Ammon’s horn. It has recently been shown by Hjorth-Simonsen and Laurberg (’77) t h a t the hilus fasciae dentatae projects only to the contralateral area dentata. Very few details are known about the rrganization of the commissural projections from regio inferior. Gottlieb and Cowan (’73) have shown that regio inferior projects to the contralateral Ammon’s horn and they suggested that the CA3c also gave rise to commissural fibers to the molecular layer of the fascia dentata. Furthermore, little is known about differences between the projections from the septal and temporal parts, as well as about the possible differences between the projections from individual parts of regio inferior. Zimmer (’71) and Gottlieb and Cowan (‘72) have described a n ipsilateral projection to the inner part of the stratum moleculare of the fascia dentata. It originates in the hilar region, but the exact origin of these fibers has not been known. It is well established from the work by Hjorth-Simonsen (’73) t h a t the regio inferior projects ipsilaterally to the stratum oriens and stratum radiatum of the Ammon’s horn. However, little information is available about the organization of the projections along the septo-temporal and subiculo-dentate axes of the region. This paper dealt with some of the remaining problems regarding the origin of these hippocampal fiber systems and the variations in their projections along the septo-temporal and subiculo-dentate axes of the hippocampus. MATERIALS AND METHODS

Adult male Wistar rats weighing between 200 and 300 g were used. Operations, injections and perfusions were made under Nembutal anaesthesia. Lesions Lesions in t h e hippocampus were made either electrolytically by a n anodal current or by means of an iridectomy knife. In a preliminary study using 1,2, 3,5, and 8 day survival periods after transections of t h e ventral psalterium, the degeneration was maximal after two days, and consequently

this survival period was used for the remaining part of the study. The rats were transcardially perfused with a 4% unbuffered formaldehyde solution; the brains were dissected out and left in the fixative for a t least seven days. They were then soaked in a 30% (w/v) sucrose solution for three days, frozen with C02-snow, and sectioned in the horizontal plane a t 20 pm with a Dittes Duspiva cryostat. Every fifth section was mounted in serial order on gelatinized slides, and the sections were impregnated with a modification of the Fink-Heimer silver impregnation method for mounted cryostat sections (Hjorth-Simonsen, ’70). Injection of HRP 0.05-0.1p1 of a 30%solution of HRP (Grad 1, Boehringer Mannheim GmbH) in distilled water was injected into the hippocampus. The animals were perfused 48 hours after the injection of HRP with a fixative containing 1% formaldehyde and 1.25% glutaraldehyde dissolved in 0.1 mmol phosphate buffer (pH 7.3). The brains were removed, kept in the fixative for two to three hours and then stored in a mixture of 0.1 mmol phosphate buffer (pH 7.3) and 5% sucrose for three days. The 20 p m thick horizontal sections were cut in the cryostat and every fifth section was mounted on gelatinized slides and further processed according to Mesulam (‘76, procedure 8 in his table 1). Two parallel series were cut from each brain; one of the series was counterstained with neutral red. OBSERVATIONS

The nomenchture (fig. 1) used is largely t h a t of Blackstad (‘56). The term hippocampus is used to denote the entity consisting of the two major parts, Ammon’s horn and area dentata. The Ammon’s horn has two regions, regio superior (CA1 of Lorente de NO, ’34) and regio inferior (CA2 and CA3 of Lorente de NO I). The terminology for the stratification of Ammon’s horn follows Blackstad (‘56) except t h a t t h e term stratum moleculare is used instead of s t r a t u m lacunosum-moleculare (Hjorth-Simonsen, ’77). The regio inferior is divided into a n intrahilar part, a n extraventricular part, and a ventricular part (fig. 1) roughly corresponding t o Blackstad’s I CA2 was defined by Lorente de N O (‘34) as a region with large pyramidal cells but without mossy fiber terminals. This region is in the rat very small, and the regio inferior therefore in the rat largely corresponds to CA3 (Haug, ’74).

COMMISSURAL AND INTRINSIC HIPPOCAMPAL FIBERS

“lower end,” “middle portion” and “upper end” of regio inferior. This system of division is used instead of Lorente de No’s (‘34) subdivisions CA2, CA3a, CA3b and CA3c, which are based on criteria available only in Golgi preparations. The area dentata is composed of the fascia dentata and hilus fasciae dentatae (CA4 of Lorente de No, ’34). The border between the hilus and regio inferior is defined by a curved Ah hreuiations a, ventricular part of the regio inferior alv, alveus b, extraventricular part of the regio inferior c, intrahilar part of the regio inferior g, granule cell layer (stratum granulare) of the fascia dentata hil, hilus fasciae dentatae Mol, molecular layer (stratum moleculare) of the Ammon’s horn mol, molecular layer (stratum moleculare) of the fascia dentata mos, mossy fiber layer (stratum lucidum) or, stratum oriens pyr, pyramidal cell layer (stratum pyramidale) rad, stratum radiatum REG INF, regio inferior of the Ammon’s horn REG SUP, regio superior of the Ammon’s horn -

~

Fig. 1 Fink-Heimer impregnated section from the middle dorso-ventral part of the hippocampus cut in the horizontal plane. The section is from case 386 in which the psalterium ventrale was transected two days before sacrifice. The large arrowhead points to the border between regio superior and regio inferior. The small arrows delimit the ventricular (a), extraventricular (b) and intrahilar (c) subdivisions of regio inferior. X 24.

687

line with its concavity towards regio inferior. The line connects the tips of the medial and lateral blade of the granule cell layer and passes through the hilar end of the ammonic stratum pyramidale. Directions along the longitudinal axis of the hippocampus are referred to as “septal” and “temporal” according to Blackstad et al. (’70). Positions in the plane perpendicular to the septo-temporal axis are described with reference to the subiculo-dentate axis defined by Gottlieb and Cowan (’73).

Degeneration in the Ammon k horn after complete transection of the ventral hippocampal commissure These observations are based upon ten rats, in which the psalterium ventrale was completely transected close to the midline 48 hours before sacrifice. The commissural projections to Ammon’s horn have been described in considerable detail before, and therefore only features concerning the differential distribution of the fibers along the septo-temporal axis will be presented here. Stratum moleculare of Ammon’s horn and the layer of mossy fibers were virtually without degeneration, and only a few particles were seen in stratum pyramidale. A region of dense degeneration was found in the stratum oriens and the stratum radiatum. The amount of degeneration in both stratum radiatum and stratum oriens of the regio superior (figs. 2 , 3) was greatest in the septal end and decreased gradually towards the temporal tip. In the septal part, the degeneration appeared almost as dense in stratum radiatum as in stratum oriens (fig. 3a); however, the evaluation of the relative densities was difficult due to the heavy degeneration. In the temporal half (fig. 2b) the degeneration was clearly more pronounced in stratum oriens than in stratum radiatum. On the whole, the density of the degeneration did not, in both strata, change along the subiculo-dentate axis. In both straturn radiatum and oriens of regio inferior the degeneration was most dense in the septal end and decreased continuously in the temporal direction (figs. 2, 3). At all septo-temporal levels the degeneration in these layers was weaker than that in the same layers of regio superior. In the septal end (figs. 2a, 3c), the degeneration was equally pronounced in the two layers both along the sub-

688

S. LAURBERG

Fig. 2 Three photomicrographs of Fink-Heimer impregnated sections from different dorso-ventral levels of a horizontally sectioned hippocampus. The ventral psalterium had been transected two days before sacrifice. The insert in figure 2a shows the approximate dorso-ventral levels of the sections shown in figures 2a-c. The two arrowheads in figure 2a point t o two transition zones between regio superior and regio inferior. Note that the right part of figure 2a represents a more septa1 level than the left part. Areas outlined by rectangles in figures 2a and 2c are shown at higher magnification in figure 3. Case 386.2a X 28; 2b X 23; 2c X 23.


689

Fig. 3 Photomicrographs at higher magnification of boxed areas in figures 2a and 2c. a. Photomicrograph of the area of regio superior outlined by broken line rectangle in figure 2a. A dense degeneration is found throughout stratum radiatum and oriens contrasting with the small number of particles in the pyramidal and molecular layers. The degeneration in stratum radiatum appears nearly aa dense aa in stratum oriens. b. Photomicrograph of the area of regio superior outlined by broken line rectangle in figure 2c. The degeneration in stratum radiatum and oriens is much weaker than in the same layers in figure 3a. The density of the degeneration is slightly greater in stratum oriens than in stratum radiatum. c. Photomicrograph of the area of the ventricular part of regio inferior outlined by solid line rectangle in figure 2a. The degeneration in stratum radiatum is almost as dense aa in stratum oriens. The amount of degeneration differs only slightly between the superficial and the deep parts of each layer. d. Photomicrograph of the area of the ventricular part of regio inferior outlined by solid line rectangle in figure 2c. In both stratum radiatum and oriens the degeneration is weaker than in figure 2c. In both stratum radiatum and oriens the degeneration is more dense near the pyramidal cell layer and diminishes gradually with increasing distance from the pyramidal cell layer. 3a-d X 198.

690

S. LAURBERG

iculo-dentate axis and from proximal to distal along the pyramidal cell dendrites. In the middle and temporal parts of regio inferior the number of impregnated particles was greater in stratum oriens than in stratum radiatum. In both strata the degeneration was heaviest towards the dentate (figs. 2b,c) and decreased gradually in the subicular direction until just before regio superior where it increased again. In both strata the degeneration decreased gradually towards the distal parts of the apical and basal dendrites of the pyramidal cells (fig. 3d). The gradients described did not depend upon a specific survival time, since identical gradients were observed in animals with survival periods of 1, 3, 5, and 8 days.

Origin of the intrinsic and commissural projections to the Ammon's horn Lesions of the hilus fasciae dentatae (fig. 4) This part of the study was based on five ani-

-

309

-

368

_ _ _ _ _ 370

mals (309, 368, 370,401,433) that had lesions a t the middle septo-temporal level of the hilus fasciae dentatae. In case 309 the hilus was approached from the lateral side with the electrode penetrating the subiculum and regio superior, and the overlying neocortex. In the remaining cases the area was approached from a postero-medial side to avoid regio superior. Degeneration in the contralateral hippocampus. Regio superior and regio inferior were devoid of degeneration in all animals. In the contralateral area dentata, terminal degeneration was found throughout t h e hilus fasciae dentatae and in approximately the inner third of the molecular layer of the fascia dentata. Degeneration in the ipsilateral A m m o n 's horn. Regio superior: In cases 309 and 401, in which there was a small concomitant lesion of regio superior close to the subiculum a weak degeneration was found in stratum radiatum and oriens at the level of the lesion. No degeneration was found in these layers in the remaining cases. Regio inferior: Stratum oriens and radiatum were free of degeneration except for a few cases in which a small number of large silver particles were observed a t the level of the lesion in the intrahilar part of the regio inferior. These did not resemble the terminal degeneration seen elsewhere in the hippocampal region and were not assumed to be terminals. In animals 368 and 433, the stratum moleculare of regio superior and inferior were devoid of degeneration. In the other cases (309, 370, and 401) it contained massive degeneration as to be expected since fibers from the entorhinal area to the ipsilateral hippocampus were interrupted. In conclusion, no evidence was obtained of ipsilateral or commissural projections t o the Ammon's horn from the hilus fasciae dentatae. Injection of HRP

Fig. 4 Diagram of the hippocampal region showing the approximate location of lesions of the hilus fasciae dentatae in five cases, described in the text. X 23.

HRP was injected into the dorsal part of the hippocampus in two animals (cases 464, 465). The center of injection was the regio inferior, but in both cases there was a small amount of diffusion into the area dentata. Only a few HRP containing cells were found in both the ipsilateral and contralateral hilus fasciae dentatae compared to several hundred cells in regio inferior (figs. 7a,b). These observations thus support the conclusion based upon the lesions of the hilus fasciae dentatae made at


b

Granule cell layer

......

Reglo interior pyramidal layer

trijection

:'.

Regio superior pyramidal layer

691

Fig. 5 a. Lateral profiles of four hippocampi in which HRP was injected into the area dentata. Shaded areas show the locations of the injections. Solid lines across the hippocampi indicate the most temporal level at which HRP-containing cells were found in the hilus fasciae dentatae. The arrows on profiles from cases 458,459and 460 show the temporal limit of HRP-containing cells in regio inferior. Broken lines marked a and b in the profile of case 461 indicate the level of the first and last camera lucida drawing in figure 5b.b. Four camera lucida drawings from case 461 in which the HRP injection was restricted to the area dentata. The distance between the horizontal sections drawn is 200 pm. x 55.

more temporal levels in that they also indicate that the hilus has neither ipsilateral nor commissural projections to the Ammon's horn.

tributed throughout the cross section of the region and were found along a considerable extent of the septo-temporal axis, viz., in the septal one-half to four-fifths of the septo-temOrigin of the intrinsic and commissural poral extent of the hilus (fig. 5a). In all cases, projections to the area dentata cells were present in the most septal part. HRP was injected into the septal one fifth The HRP-stained axons that project to the of the hippocampus in four animals (458,459, contralateral hippocampus were not diffusely 460,461) (fig. 5a). In case 461,the hippocam- distributed in the fimbria and the psalterium pal injection was restricted to the area den- ventrale. In both the ipsilateral and contralattata (fig. 5b), while in the other animals (458, eral fimbria, HRP-positive fibers were con459,and 460)there was a minor diffusion into fined to the part of the fimbria that was the part of regio inferior adjacent to the hilus farthest away from the lateral ventricle. In fasciae dentatae. the contralateral fimbria, fibers could be HRP containing cells and fibers in the con- traced into the alveus of the extraventricular tralateral hippocampus. In case 461,in which part of regio inferior and further on to the the injection was confined to area dentata, the area dentata. HRP-positive cells in the ipsilateral hiponly hippocampal cells that had taken up the HRP were those in the hilus fasciae dentatae. pocampus. The distribution of HRP containIn the remaining cases a few pyramidal cells ing cells in the hilus fasciae dentatae (fig. 6) in regio inferior contained HRP. These cells and the regio inferior was similar to that in were found a t the same level as the contralat- the contralateral hemisphere and the cells eral injection (fig. 5a) and were distributed were present a t the same septo-temporal diffusely along the subiculo-dentate extent of levels as there. Additional experiments indicate that the the pyramidal cell layer of regio inferior. HRP containing hilus cells were evenly dis- lack of HRP containing cells in the regio

692

S. LAURBERG

inferior, after injection into the area dentata as mentioned above, most likely was not due merely to a defective retrograde axon transport of HRP from the hilus fasciae dentatae t o regio inferior. These experiments included injections of similar doses of HRP (to those for area dentata) into the regio inferior (figs. 7a,b; cases 464, 465). This resulted in the labelling of a large number of regio inferior cells in t h e pyramidal cell layer on both sides (described more fully on p. 690).Apparently axons from regio inferior to area dentata either do not exist or are rather few in number. Summarizing, the cells in the hilus fasciae dentatae are the major and probably the only source of the commissural and intrinsic projections to the area dentata. The ipsilateral and the commissural projections to the septal portion of the area dentata originate from

Fig. 6 Photomicrograph of the hilus fasciae dentatae and the intrahilar part of regio inferior 3.5 mm dorsal to the temporal tip of the hippocampus. Several HRP-containing hilus cells are diffusely scattered throughout the hilus. No regio inferior cells contain HRP. Case 461. X 90.

hilus cells located a t identical septo-temporal levels. Projections from the septalpart of the regio inferior Fink-Heimer material The septal part of the hippocampus was electrolytically lesioned in two animals (cases 300, 301). The lesion was made from a dorsal approach, in the extraventricular part of the regio inferior, with a small needle track through the most dorsal part of the regio superior. Ipsilateral degeneration The regio superior. The degeneration in stratum oriens was dense and uniform throughout the layer a t the level of the lesion, but i t could be followed for only a relatively short distance in the temporal direction. The degeneration in stratum radiatum displayed the same degree of density and uniformity a t the level of the lesion as the degeneration in stratum oriens. In the temporal direction the degeneration showed a gradual decrease and eventually ceased about 3 mm from the temporal tip. The decrease was most pronounced a t the border with regio inferior and in the temporal direction progressed from this site both toward the subiculum and toward the surface of the stratum radiatum. This resulted in a broadening zone of normal tissue between the degeneration in the regio inferior and in the superficial part of the regio superior (fig. 8a). Regio inferior. Degeneration was present in stratum oriens along the full subiculo-dentate extent of the layer a t the level of the lesion. More temporally, degeneration was absent from the layer, except in an area a t the same subiculo-dentate position as the lesion. Degeneration was found in all parts of stratum radiatum a t the level of the lesion and was evenly distributed throughout the depth of the layer. In the temporal direction the degeneration soon began decreasing in the deep parts of the layer and was reduced to scattered impregnated particles only, while dense degeneration in the outer part of the layer persisted (fig. 8a). The superficial degeneration persisted along the full subiculo-dentate distance through nearly the entire septo-temporal extent of radiatum. Only the last temporal 1.5 mm of the layer did not contain degeneration.


Commissural projections The degeneration in the contralateral Ammon’s horn was weaker than that on the ipsilateral side but showed the same gradients along the septo-temporal axis. The only major difference was a denser commissural projection to stratum oriens of both regio superior and regio inferior. The commissural degeneration could be traced for a slightly shorter distance in the temporal direction. Injection of HRP into the septal part of the regio inferior The two cases in which HRP was injected into the septal part of the regio inferior (partly described p. 690: cases 464, 465) contained axons labelled by anterograde transport of HRP in addition to cells labelled by retrograde transport. These axons were distributed in a

693

pattern that was qualitatively similar to the pattern of degeneration seen in Fink-Heimer preparations after lesions of the same part of the regio inferior (fig. 8b). It is most likely due to a different sensitivity of the two methods that the projection to regio superior appeared much weaker in the HRP-preparations than in the Fink-Heimer preparations and could not be traced as far temporally. Projections from the middle septo-temporal parts of the regio inferior (1) Lesions of the intrahilar part of regio inferior In three cases lesions were placed in the intrahilar part of regio inferior. In two of these (376, 439) regio inferior was approached from a postero-medial angle to avoid regio superior. Ipsilateral projections. In regio superior

7a Fig. 7 a. Camera lucida drawing of a horizontal section through the dorsal curvature of the hippocampus. Case 464 in which HRP was injected into the septal part of the contralateral regio inferior two days before sacrifice. The dots represent the approximate number and location of HRP-containing neurons. The small arrow points to the border between the regio superior and the regio inferior. The only hippocampal neurons that contain HRP are located in the pyramidal cell layer of regio inferior. b. Photomicrograph of an area outlined by rectangle in figure 7a. Several cells in the pyramidal cell layer of the intrahilar part of regio inferior contains HRP, in contrast to the lack of labelled neurons in the hilus fasciae dentatae. 7a X 6.5; 7b X 44.

694

S. LAURBERG

(studied in cases 376 and 439)no degeneration was observed in stratum oriens. Degeneration occurred in stratum radiatum and was of equal density along the whole subiculo-dentate axis a t all septo-temporal levels, except for a reduction close to the subieulum. However, it varied much in density from proximal t o distal along the pyramidal cell dendrites. There, the gradient differed according t o the septo-temporal level of the lesion. At the level of the lesion, the degeneration was sparse in the inner one-fourth of the layer, and dense in the outer three-fourths. With increasing distance from the lesion in the temporal direction, the zone of weak degeneration gradually widened and finally dense degeneration was found only in a narrow zone immediately beneath the molecular layer; only a few

impregnated particles were present in the rest of the layer. Septa1 t o the lesion, the degeneration decreased gradually in the superficial part of the layer. In the most septal part of the projection field, most of the degeneration was found close to stratum pyramidale with only scattered degenerating particles above. Regio inferior, exclusive of the intrahilar part. In stratum oriens degeneration was found only in the extraventricular part a t the level of the lesion. The degeneration in stratum radiatum was considerably weaker than that in the same layer of regio superior, and was found along the full subiculo-dentate extent of regio inferior. It was most intense near the lesion and dwindled towards regio superior. Along the septo-temporal axis the degeneration varied in the same way as in regio su-

Fig. 8 a. Photomicrograph of a Fink-Heimer impregnated section from case 300, in which the septal part of the extraventricular part of regio inferior was lesioned on the same side. In the superficial part of stratum radiatum of regio inferior, degeneration is found along the entire subiculo-dentate extent. Degeneration in regio superior at this level is confined t o the part of stratum radiatum that borders the subiculum. b. Photomicrograph showing the location of anterogradely transported HRP at a level temporal t o an injection of HRP into the septal part of the extraventricular regio inferior on the same side. At this level anterogradely transported HRP is located only in the superficial part of stratum radiatum of regio inferior, resembling the longitudinal association path described by Lorente de NO (‘34).8a X 31; 8b X 20.

695


perior. The degeneration was heaviest near the zone of mossy fiber septa1 to the lesion, but densest in the superficial part of the layer temporal to the lesion. Contralateral projection. In regio superior only a weak degeneration was found in stratum oriens a t the level of the lesion. The degeneration instratum radiatum was weaker than that on the ipsilateral side, but showed the same gradients along the subiculo-dentate and septo-temporal axes and in the proximodistal direction within the layer. In regio inferior the terminal degeneration subsequent to the relatively small lesion was sparse. In stratum oriens only a few particles were found close to the hilus fasciae dentatae. In stratum radiatum degeneration occurred along the full extent of the subiculo-dentate axis. It was most dense near the hilus and decreased gradually toward regio superior. The degeneration showed the same proximo-distal variations depending on septo-temporal level as it did on the ipsilateral side. The absolute

_____

9

289

. .. . . . .

376

V

-

value of the septo-temporal spread, of the ipsilateral and commissural projections t o the Ammon’s horn, was difficult to ascertain due to the smallness of the lesions. (2) Combined lesions of the intrahilar and extraventricular parts of the regio inferior A lesion was placed in the extraventricular part of regio inferior in six animals (fig. lo), in three animals (272, 306, and 307) via a lateral approach through the ipsilateral regio superior and in the remaining cases (378, 424, 436) from a postero-medial direction, without involvement of the other areas of the hippocampus. These lesions included interruption of the ipsilateral and commissural fibers from the intrahilar part of the regio inferior. The terminal degeneration described below was therefore the combined projections from the intrahilar and extraventricular part of the regio inferior.

439

Fig. 9 Camera lucida drawing indicating the location of three lesions involving the intrahilar part of regio inferior. The numbers refer to specific animals discussed in the text. X 23.

272

... . . . .

378

-

306 4 24

-

307

_ _ _ _ -436

10 Fig. 10 Camera lucida drawing indicating the location of lesions in six animals in which the extraventricular part of regio inferior was lesioned. The numbers refer to specific animals discussed in the text. x 23.

696

S. LAURBERG

Ipsilateral projections. In regio superior, analysed in cases 378, 424, 436, degeneration was present in stratum oriens and was of the same density throughout the layer at all septo-temporal levels. The degeneration had a septo-temporal divergence of the same magnitude as that in stratum radiatum. Degeneration in stratum radiatum was much denser than in stratum oriens (fig. l l c ) and more pronounced than after lesions restricted t o the intrahilar part of the regio inferior; however, it showed similar spatial variations at levels septal and temporal to the lesion (figs. lla,b). In the ventricular part of the regio inferior no degeneration was found in stratum oriens a t levels temporal t o the lesion. A t the level of the lesion and more septally weak degeneration was found in all parts of regio inferior (fig. Ilc). It was most pronounced near the lesion and gradually decreased towards regio superior. Degeneration was observed in all parts of stratum radiatum. It was much denser than the degeneration in stratum oriens of regio superior, but weaker than in radiatum of regio superior (fig. llc). At levels temporal t o the lesion the degeneration prevailed near the border to stratum moleculare (fig. l l b ) . At septal levels, the degeneration was most dense close to the layer of mossy fibers. Commissural projections. In regio superior, the degeneration was denser than on the ipsilateral side in stratum oriens, but much weaker in stratum radiatum (figs. llc,d). The degeneration in stratum radiatum and stratum oriens were of approximately the same intensity and showed the same septotemporal and subiculo-dentate variations on the two sides. Also, the spread of the ipsilatera1 and commissural fibers to regio superior was of the same considerable magnitude. In both layers the degeneration could be traced for 1.2 mm temporal to the lesion, and i t extended all the way to the septal tip of the region. Regio inferior. The degeneration was much weaker in regio inferior than in regio superior (fig. l l d ) . In stratum oriens it changed along the septo-temporal axis: temporal to the lesion, i t was seen only in the intrahilar and extraventricular parts of the layer. Towards the septum it expanded toward regio superior. In the septal end of the projection field, the degeneration was of about the same density along the subiculo-dentate axis of the layer.

The degeneration in stratum radiatum (fig. l l d ) was weaker than in stratum oriens, and was less dense than in radiatum on the side of the lesion. Degeneration was found in all parts of regio inferior, but it was maximal corresponding to the part lesioned on the opposite side. Temporal to the lesion, the degeneration was concentrated in the superficial parts of the layer with only moderate degeneration deeper, whereas septal to the lesion the degeneration was maximal immediately above the mossy fiber layer. The septo-temporal spread of the commissural and ipsilateral projections to regio inferior were similar. The temporal spread of the degeneration both in stratum radiatum and oriens amounted to only a few hundred microns while the septal spread was extensive. In the cases with relatively large lesions, the degeneration could be traced to the septal tip. (3) Lesion of the ventricular part of regio inferior Lesions were made in the ventricular part of regio inferior via a lateral approach through the temporal bone in three rats (284, 286, 287). In all cases the lesion involved regio superior, but since it is well documented that regio superior does not give rise to either commissural projections to the Ammon's horn (Gottlieb and Cowan, '73; Hjorth-Simonsen and Laurberg, '77) or ipsilateral projections to regio inferior (Lorente de NO, '34; HjorthSimonsen, '73), the involvement of this area Fig. 11 a. Photomicrograph from the hippocampus a t the level of a lesion of the extraventricular part of regio inferior. In regio superior and the ventricular part of regio inferior, degeneration appears equally dense in the superficial and deep parts of stratum radiatum. b. Photomicrograph of a section from the hippocampus taken temporal to a lesion of the extraventricular part of regio inferior. In stratum radiatum of both regio superior and regio inferior, degeneration is most dense in the superficial part toward stratum moleculare and diminishes in the deeper part of the layer. c. Photomicrograph of parts of regio superior and regio inferior outlined by the rectangle in figure lla. The degeneration is heavier in regio superior than in regio inferior, and the degeneration is more pronounced in stratum radiatum than in stratum oriens. Arrow points to the transition zone. d. Photomicrograph from the same area and dorso-ventral level as seen in figure llc, but from the opposite hippocampus of the same animal. (The mirror image is shown in order to simplify the comparison with figure llc.) The distribution of silver particles is generally the same as in figure llc. Compared to the ipsilateral side, degeneration is less dense in stratum radiatum of the Ammon's horn, but heavier in stratum oriens. Arrow points to the transition zone between the regio superior and the regio inferior. lla,b x 26; llc,d x 134.


Figure 11

697

698

S. LA1JRBERG

was not expected to influence the observations. Ipsilateral projections. The degeneration in regio superior will not be described in detail because the lesion transects fibers en passage from the contralateral hippocampus from the ipsilateral regio inferior and autochthonous fibers of regio superior itself. Degeneration was found in both stratum radiatum and oriens, but it was absent from stratum moleculare. Regio inferior, excluding the part lesioned. In stratum oriens the degeneration was heavy throughout the layer a t the level of the lesion. The number of silver particles reduced towards the dentate area. In the temporal direction, the amount of degeneration decreased in two ways: from the dentate end and up from the proximal parts of the pyramidal cell dendrites (fig. 13b). The degeneration disappeared completely a few hundred microns temporal to the lesion. Septal to the lesion the degeneration was concentrated at the proximal parts of the basal dendrites of the pyram_ _ _ _ - 284

12

-

286

.. . . . . .

287

u

Fig 12 Camera lucida drawing of the approximate extent of three lesions involving the ventricular part of regio inferior at a middle dorso-ventral level of the hippocampus. The numbers refer to animals discussed in the text. x 23.

idal cells (fig. 13a) and could be followed for more than 2 mm. The degeneration in stratum radiatum was generally heavier than that in stratum oriens. It varied along both the subiculo-dentate and the septo-temporal axes. At the level of the lesion, degeneration was found in all parts of the layer. It was most intense close to the lesion and diminished gradually towards the hilus fasciae dentatae. It occupied the entire depth of the layer, and was maximal close to the zone of mossy fibers. Septal to the lesion, the degeneration gradually disappeared from both the dentate end and the superficial part of the layer (fig. 13c). At levels temporal to the lesion the degeneration disappeared from the dentate end and from the deep part of the layer (fig. 13d). The spread was about 1.0 mm in the temporal direction and more than 2 mm in the septal direction. Commissural projection. Regio superior. At the level of the lesion the degeneration in stratum oriens was of uniform density throughout the layer. More temporally it gradually disappeared from the dentate end of the layer. Septal to the lesion the degeneration receded progressively from the subicular end. The degeneration in stratum radiatum was much weaker than in stratum oriens. At the level of the lesion, i t was relatively uniform both along the subiculo-dentate axis and across the layer. It only became attenuated close to the border to the subiculum. At levels temporal to the lesion, the degeneration was concentrated superficially in the layer with only little degeneration beneath. Septal to the lesion the degeneration dominated deep in the layer. Regio inferior. Dense degeneration was found in stratum oriens a t the level of the lesion. The degeneration was present in all parts of regio inferior, but was most dense in the ventricular part, diminishing towards the area dentata. It was equally pronounced throughout the depth of the layer. Septal to Fig. 13 Photomicrographs of Fink-Heimer preparations showing the extraventricular part of regio inferior in an animal with a lesion of the ventricular part of regio inferior on the same side. At levels septal to the lesion, the degeneration is heaviest along the proximal parts of the pyramidal cell dendrites in both stratum radiatum (fig. 13a) and stratum oriens (fig. 13c). At levels temporal to the lesion the degeneration is maximal along the distal part of the pyramidal cell dendrites in both stratum radiatum (fig. 13b) and stratum oriens (fig. 13d). 13a,b X 530; 13c,d X 470.


Figure 13

699

700

S. LAURBERG

the lesion the degeneration was most dense superficially in the layer, generally vanishing a few hundred microns from the lesion. Only in the ventricular part of the region could it be followed for approximately 2 mm in the septal direction. Temporal to the lesion the degeneration withdrew from the dentate end and from the deep parts of the layer. It disappeared after a few hundred microns. The degeneration in stratum radiatum was sparse. At the level of the lesion, it was found in all subiculo-dentate parts of the layer. It was most dense in the ventricular part and decreased towards the dentate area. Through the depth of the layer no gradients were found. Septa1 to the lesion the degeneration was found only in the ventricular part immediately above the mossy fiber zone. Temporal to the lesion it was present along the entire subiculodentate extent of the layer. It was most dense superficially. The septo-temporal spread was of the same size as in stratum oriens. DISCUSSION

Thecommissuraldegenerationin theAmmon ’s horn after complete transection of the psalterium ventrale The commissural projection to the Ammon’s horn in the rat has been described in detail by Blackstad (’56) with Nauta’s (’50) method for the tracing of degenerating fibers and terminals and by Gottlieb and Cowan (‘73) with injections of tritiated leucine. In both of these works, commissural fibers were found to project t o all parts of stratum oriens and radiatum, of both regio superior and inferior. Blackstad reported that the “main middle” part of the regio inferior was nearly normal, with only a small number of scattered degeneration particles. This finding agrees with the observations reported here on the distribution of commissural fibers in the temporal parts of the regio inferior. However, in the septal end, a dense degeneration was found in stratum radiatum in all parts of regio inferior. In the earlier works cited above there was no description of variations in the projection along the septo-temporal axis of the Ammon’s horn. The present study shows that variations in distribution pattern exist in all parts of the projection field along the septo-temporal axis of the Ammon’s horn. The density of the projections to both stratum radiatum and oriens was greatest in the septal parts of the Ammon’s

horn, and diminished gradually towards the temporal tip. A similar septo-temporal gradient has been found by Raisman et al. (’651, Mosko et a,. (’73) and Hjorth-Simonsen and Laurberg (‘77) in the commissural projection to the inner part of the molecular layer of the fascia den tata. Hjorth-Simonsen and Laurberg (’77) found, however, that the commissural projec tion to the hilus fasciae dentatae (CA4) had an inverse septo-temporal gradient, i.e., it was heaviest in the temporal part. The septo-ternporal gradients in the commissural projection to the fascia dentata and the Ammon’s horn may be part of a general septo-temporal gradient in the development of the connections of the Ammon’s horn and the fascia dentatii. Zimmer and Haug (’781, using Timm s sulphide silver method to identify the terminal field of intrinsic and extrinsic fiber syt3terns, have found that the temporal parts are the first to differentiate in the commissural/ ipsilateral zone of the molecular layer of the fascia dentata and in stratum radiatum and oriens of the Ammon’s horn. Moreover, Fricke and Cowan (‘77) have shown, by injection of tritium labelled amino acids into the hilar region that the ipsilateral afferents to the inner part of the molecular layer of the fascia dentata are present a t an earlier stage of development than the commissural afferents. By talring into account the works of Fricke artd Cowan (‘77) and Zimmer and Haug (‘78) artd by assuming that the commissural fibers arrive after the association fibers to stratum oriens and radiatum of Ammon’s horn, the septo-temporal gradients in the commissural projections to the hippocampus can be explained by temporal factors. If the cornmissural fibers to the Ammon’s horn arrive later than the intrinsic fibers, and since the temporal parts of Ammon’s horn and the fascia dentata develop before the septal part, the commissural projections will in general have to project more septally. In the temporal part of the regio inferior, gradients in two directions were found in the projections to both stratum radiatum and stratum oriens: In both, the degeneration was weakest in the ventricular part, except near the border of regio superior, and most dense about the proximal parts of the pyramidal cell dendrites. One would predict from the temporal theory that the ventricular part of regio inferior should be the first to differentiate,


and that terminals would be formed first on the distal parts of the pyramidal cell dendrites within both stratum radiatum and stratum oriens. Several studies invite this inference. Angevine (‘65) (mouse) and Hine and Das (’74) (rat), using tritium marked thymidine have shown that the cellular proliferation terminates earlier in subfield CA3a, which corresponds roughly to the ventricular part of regio inferior, than in any other part of the pyramidal layer of regio inferior. Furthermore, Zimmer and Haug (‘78) found that the earliest part of the rat regio inferior to show Timm staining was CA3a and that the staining appeared first a t the distal parts of the pyramidal cell dendrites within stratum radiatum as well as stratum oriens. These three papers therefore suggest that the gradients in the commissural projections to the intermediate and temporal parts of stratum radiatum and oriens of regio inferior might be due to temporal factors. Little is known about the organization of the commissural projections in animals other than rats. Hjorth-Simonsen (‘77) showed that the commissural fibers passing through the psalterium ventrale of the rabbit are also distributed to both stratum oriens and radiatum of Ammon’s horn. The distribution is quantitatively different, however, from that in the rat, particularly in stratum radiatum. In the rabbit the projection to this layer was sparse in regio superior compared to regio inferior, and a distinct septo-temporal gradient was observed only in stratum radiatum. This gradient was qualitatively similar to the one observed in the rat. Additional studies, involving a comparison of the time of development of the pyramidal cells of Ammon’s horn and the time of arrival of the commissural projections in the two species, might clarify whether or not the differences in the organization of the commissural connections in rats and rabbits are related to different sequences of the development of cells and afferent systems.

The commissural and intrinsic connections of the area dentata Origin. According to Raisman et al. (‘651, Gottlieb and Cowan (‘73), and Hjorth-Simonsen and Laurberg (’771, both the hilus fasciae dentatae and the molecular layer of the fascia dentata receive a commissural projection through the psalterium ventrale. Gottlieb and Cowan confirmed Blackstad’s (‘56) original

701

observation of a commissural projection to the inner one-third of the molecular layer of the fascia dentata and found labelling in the contralateral fascia dentata only after injections into the hilar region. In that study the hilar region included the CA4 (hilus fasciae dentatae) and CA3c. Hjorth-Simonsen and Laurberg (‘77) found that neither regio superior nor the ventricular part of regio inferior was the origin of the projections to the area dentata. Moreover, they observed degeneration both in the molecular layer of the fascia dentata and in the hilus fasciae dentatae in cases with lesions of the CA4 which did not involve regio inferior. However, they were unable to exclude the possibility that such fibers also arise in the intrahilar part of regio inferior, since a lesion there would have transected fibers from the CA4 to the contralateral area dentata. An intrinsic projection restricted to the same inner zone of the molecular layer of the fascia dentata as the commissural projection was described by Zimmer (‘71) and found to originate in either or both of CA3c and CA4. The existence and origin of this system was confirmed by Gottlieb and Cowan (’72) with labelled amino acids and by Lynch et al. (‘76) with HRP. It is thus well established that the hilar region (CA4ICA3c) projects to the ipsilateral as well as the contralateral fascia dentata, but the exact origin of these fibers was unknown. In this study injections of HRP into the area dentata resulted in HRP-labelled cells in both the ipsilateral and contralateral hilus fasciae dentatae. Thus the hilus has both ipsilateral and commissural projections, to the hilus, the fascia dentata, or both. Hjorth-Simonsen and Laurberg (’77) found, after complete transections of the psalterium ventrale, that a t septal levels impregnated particles occurred in close to negligible numbers in the hilus fasciae dentatae, but were densely crowded in the fascia dentata. Therefore, the HRP-containing cells found in the present study in the contralateral hippocampus after injections into the septal part of area dentata largely represent cells that project to the molecular layer of the fascia dentata. An ipsilateral fiber system interconnecting different parts of the hilus fasciae dentatae along the septo-temporal axis has not been described, but since there are decussating interhilar fibers the existence of a corresponding

702

S. LAURBERG

ipsilateral projection would seem likely. Indeed, the occurrence of HRP-containing cells in the same hemisphere as the area dentata HRP injections do indicate the existence of cells projecting t o the hilus, the molecular layer of the fascia dentata, or both. The experimental material described here offered no positive evidence that regio inferior is an additional origin of ipsilateral and commissural fibers to area dentata. In case 461, in which the HRP injection was restricted to area dentata, no regio inferior cells were labelled with HRP. In the other cases, in which there was also a small injection into regio inferior, few HRP-containing regio inferior cells were found a t the level of injection, and they were diffusely distributed along the subiculo-dentate axis of the region. Since both the CA3a and CA3b do not project to the ipsilateral or contralateral area dentata (Zimmer, '71; Gottlieb and Cowan, '72, '73; Lynch et al., '76; Hjorth-Simonsen and Laurberg, '77) the regio inferior cells containing HRP in these cases can be explained by the small injection into the regio inferior. In conclusion, this study indicates that the hilus fasciae dentatae is the sole source of the crossed and uncrossed hippocampal afferents to area dentata. Septo-temporal divergence. Gottlieb and Cowan ('73) after injection of tritiated amino acids into the hilar region a t middle rostrocaudal levels found labelling in the contralateral molecular layer of the fascia dentata, distributed over most of its rostro-caudal extent; only in the caudalmost part of the molecular layer was the labelling negligible. HjorthSimonsen and Laurberg ('77) made localized lesions of the hilus fasciae dentatae equally a t middle dorso-ventral levels. The degeneration observed in the contralateral molecular layer of the fascia dentata extended only as far as about 1 mm temporal to the lesion, but in the other direction all the way to the septal tip of the fascia dentata. The large septal spread of the commissural projection to the area dentata, from the middle rostro-caudal parts of the hilus has been confirmed in this study. In harmony with this, injections of HRP into the septal part of the area dentata resulted in HRP-labelled cells in the hilus along the major part of the septo-temporal axis. Only the most temporal 20-40%of the hilus lacked HRP-containing cells. Since the commissural projections from middle dorso-ventral levels of the hilus were also found to terminate in

the septal parts of the contralateral area dentata, it has been uncertain whether or not the septal parts of the hilus fasciae dentatae ccntribute to the commissural projection to the area dentata. This study has shown that the septal part of the hilus fasciae dentatae does contribute fibers to the contralateral area dentata, because in all of the cases with injcctions of HRP into the area dentata, HRP-ccntaining cells were found in the septal parts of the hilus. The septo-temporal spread of the ipsilateral association system terminating in the molecular layer of fascia dentata was described by Zimmer ('71). The projection from the temporal levels of the hilar region was described as spreading only moderately. In contrast, lesions of the hilar region a t middle dorso-ventral levels gave degeneration extending all the way to the septal tip of the layer. Similar clbservations were made in the course of the present investigation. Moreover, injections of HRP into the septal part of the area dentata resulted in HRP-labelled cells a t the same septo-temporal levels of the hilus, in both heniispheres in all cases (occupying the septal three- to four-fifths of the hilus). This indicates that the septal spread of the intrinsic and commissural projections to the dentate area may be of the same order of magnitude in general. Origin of the commissural and intrinsic connections to stratum radiatum and oriens of Ammon's horn Commissural projections Several studies indicate that regio superior does not give off fibers to the contralateral hippocampus. Gottlieb and Cowan ('73) found no evidence of such fibers after injection of labelled amino acids into the anterior part of regio superior. Segal and Landis ('74) after injection of HRP into the hippocampus found no HRP-containing cells in the contralateral regio superior. Hjorth-Simonsen and Laurberg ('77) saw no degeneration in the contralateral hippocampus after lesion of regio superior a t a middle septo-temporal level. Finally, Andersen et al. ('73) were unable to record antidromic population spikes in regio superior after stimulation of the fimbria. Since, moreover, the dentate granule ce 11s likewise emit no crossing axons (Cajal, 1893; Blackstad et al., '70) only the hilus fasciae dentatae and regio inferior remain as possikde


sources of the commissural projection to stratum radiatum and oriens of Ammon’s horn. Hjorth-Simonsen and Laurberg (‘77) found no degeneration in the contralateral Ammon’s horn in one animal with a lesion in the hilus. This observation was confirmed in the present study where six animals with lesions in the hilus showed no evidence of degeneration in the contralateral Ammon’s horn. In addition, in two animals, in which HRP was injected in the septa1 part of regio inferior with only a minor diffusion into the area dentata, very few cells in the opposite hilus contained HRP. This part of the present study thus adds further support to the conclusion that regio inferior must be the source of the commissural fibers to stratum radiatum and oriens of Ammon’s horn. Ipsilateral projections From the Golgi studies of Schaffer (1892), Cajal (1893, ’68) and Lorente de NO (‘34) and the degeneration study by Hjorth-Simonsen (‘73) it is well established that regio inferior has ipsilateral projections to both regio superior and regio inferior. According to Lorente de NO (’34), Hjorth-Simonsen (’73) and Andersen et al. (‘73) there are no reciprocal connections from regio superior to regio inferior. However, both Lorente de NO and Hjorth-Simonsen believed t h a t the hilus fasciae dentatae too had ipsilateral projections to regio superior and inferior. The present study permitted a correction of this standpoint. Hjorth-Simonsen’s conclusion was based on a single case (3821, which had a large lesion in the hilus fasciae dentatae, which, however, encroached on the intrahilar part of regio inferior as stated by the author (and seen in his fig. 4). The degeneration described in the Ammon’s horn by Hjorth-Simonsen is similar to that found in this study in animals with lesions involving the intrahilar part of regio inferior. Lorente de NO (‘34) shows drawings (fig. 9, cells 21, 22, and fig. 11, cell l) of three modified CA4 pyramidal cells with axonal arborizations. Zn t h e s e , axonal branches can be followed to the molecular layer of fascia dentata, consistent with the conclusions drawn in the present study that cells of the hilus fasciae dentatae are the source of the intrinsic connections to area dentata. Two of Lorente de N6’s cells (21 and 22 of fig. 9)have a branch, which can be traced to

703

the deep part of regio inferior near the fimbria. In the present study labelled fibers were found in a similar place after the injection of HRP into the contralateral area dentata. It is possible that the branches shown by Lorente de NO could be axons destined for the contralateral area dentata. From none of the hilar cells shown by Lorente de NO can axonal branches be followed to regio superior and except for those mentioned above, axonal branches are only drawn in that part of regio inferior adjoining the CA4. In cases with localized lesions of the hilus, no degeneration was found in the present study in regio superior, and only a few coarse particles in regio inferior near the lesion. On the other hand, in two animals in which HRP was injected largely into regio inferior, only a few hilus cells contained HRP, and most likely due to a minor diffusion from the injection into the area dentata. Altogether the findings of this study indicate that if the hilus fasciae dentatae takes part in the ipsilateral projection to Ammon’s horn, the content of fibers must be rather limited. Regio inferior seems to be the only important source of ipsilateral afferents to stratum radiatum and oriens of the Ammon’s horn. These afferents will be discussed in further detail in the following paragraphs. Commissural and intrinsic projections from the middle dorso-ventralpart of the regio inferior Commissural projections To regio superior. Gottlieb and Cowan (‘73) found that regio inferior projected t o regio superior, but they were unable to determine whether all parts of regio inferior gave rise to the projection. Moreover, with different injection sites within the region, they demonstrated substantial differences in the relative density of projections to stratum oriens and stratum radiatum, but their material did not allow to systematically explain these differences. The work presented here provides additional information about these projections. Several factors causing differential distribution interact to give the final pattern. For instance, a t the mid septo-temporal level all parts of regio inferior project to regio superior, but not to the same extent. The projection from the extraventricular and the intrahilar parts are heavier than that from the ventricular part. Moreover, the projections

704

S. LAURBERG

from the various subiculo-dentate parts of Septo-temporal spread. As pointed out by regio inferior terminate with a differential Gottlieb and Cowan (’73) the septo-temporal strength in stratum radiatum and oriens: de- spread of these connections is extensive. They generation was found largely in stratum described two cases (R47, R48) with injections radiatum after lesions in the intrahilar part of in the CA3 and with maximal labelling in the regio inferior, but in stratum oriens after CA3b a t the middle of the rostro-caudal exterd lesions of ventricular regio inferior. Finally, of the hippocampus. No grains were found in differences in amount of fibers depending on the temporal part, while grains were found the septo-temporal level of the origin exist from the level of the injection all the way into (vide infra). the septal tip. These observations are compatiTo regzo inferior. There is a n extensive com- ble with those made in this study. After missural projection to stratum radiatum and lesions of the extraventricular part a t middle oriens of regio inferior, but little is known dorso-ventral levels, all cases in the present about the detailed organization of this projec- study had a temporal divergence of only a few tion. Regarding its origin, Gottlieb and Cowan hundred microns but an extensive sept 31 (‘73) found that probably all parts of regio spread. In the cases with the largest lesions 3f inferior give rise to the projection, but since the extraventricular part, the degeneration they did not have sufficiently small injections could be followed all the way into the septal in individual parts of regio inferior their con- tip. clusion was presented as tentative. They always found the largest number of grains in Ipsilateral projection the part homotopic to the center of the injecThe organization of the intrinsic systems tion, but because of the spread of the injected were originally described in Golgi material. isotope, they could not decide whether each Schaffer (1892) and Cajal (1893) found that part was connected to the homotopic spot, ex- the regio inferior pyramidal cells had thick reclusively, or to all parts of regio inferior. They current collaterals t h a t traverse the superficoncluded that the commissural connections cial part of regio inferior and regio superior. appeared to be organized homotopically, in the From these, branches a r e distributed to sense that each of its subfields appeared to be stratum radiatum and stratum oriens. Lorelated principally (if not exclusively) to the rente de NO (‘34) found additional pyramidal corresponding subfield on the opposite side. cells with one or two collaterals ascending to However, the commissural projection seemed stratum radiatum, forming a dense fiber plcxto be widely spread along the septo-temporal us there mainly in CA2 and CA3a. Thus, two axis of the hippocampal formation. types of regio inferior pyramidal cells were The present study has provided some addi- seen-one with Schaffer collaterals intercontional information about these connections. It necting parts of the Ammon’s horn along the appears that all parts of regio inferior take subiculo-dentate axis and the other interconpart in the projections to both stratum necting different parts of regio inferior along radiatum and oriens. Confirmation has been the septo-temporal axis (the longitudinal asobtained of Gottlieb and Cowan’s conclusion sociation path). The pyramidal cells with that the maximal projection of a given part is Schaffer collaterals were located in CA2 and to the corresponding part, but t h a t there is a CA3a-b, those giving rise to the longitudinal divergence in the sense t h a t a lesion in one association path in CA3b-c. part leads to degeneration in all parts. After An inherent well recognized danger of delesions of the intrahilar part or combined generation studies is the potential interruplesions of the intrahilar and extraventricular tion of fibers passing through the lesion. In part, a divergence of degeneration is found degeneration studies of the ipsilateral hipespecially at levels septal to the lesion, where pocampal pathways a concomitant interrupit occupies the entire subiculo-dentate length tion of commissural fibers may readily enof both stratum radiatum and stratum oriens. cumber the distinction of ipsilateral from From the ventricular part, the greatest spread commissural fibers. However, in the present of fibers along the subiculo-dentate axis was context some knowledge available in earlier found at the level of the lesion. Summarizing, papers permits partial circumvention of this the projection to regio inferior is organized difficulty. Lorente de NO (‘34, fig. 38) using homotopically only in that the projection is Golgi preparation showed that fibers arriving densest in the part of regio inferior t h a t corre- through the fimbria take a different course to sponds to the lesion. the various parts of the hippocampus. Fibers


destined for regio superior and the ventricular part of regio inferior do not traverse the extraventricular and intrahilar parts. Fibers to the hilus near part of the regio inferior do not pass through the ventricular part of regio inferior. Gottlieb and Cowan (‘73) have confirmed that the commissural fibers to the Ammon’s horn have this organization. It therefore seems reasonable to believe that the majority if not all of the degenerating terminals found in this study in regio superior and the ventricular part of regio inferior, after lesions close to the hilus, are ipsilateral, and that the degeneration found in regio inferior close to the hilus, after lesion of the ventricular part, is largely due to ipsilateral fibers. In his study of intrinsic hippocampal connections in the adult rat, Hjorth-Simonsen (‘73) made use of chronic deafferentation to prevent degeneration due to interruption of commissural fibers. Experimental animals were used in which the hippocampal commissures had been transected a t the eighth postnatal day. In the adult rats lesions were then placed in different parts of regio inferior a t middle dorso-ventral levels. It is of interest to compare some of Hjorth-Simonsen’s findings with those made in the present study. Hjorth-Simonsen in his material saw no evidence of the longitudinal association path of Lorente de NO since the degeneration in regio inferior, irrespective of the subiculo-dentate position of the lesion, showed septo-temporal divergence slightly less than the degeneration in regio superior. This is in agreement with the results presented here that no evidence was found for a longitudinal association path a t middle dorso-ventral levels. Hjorth-Simonsen (‘73) further found that the projection from regio inferior and superior was restricted to stratum radiatum and oriens leaving lacunosum-moleculare (stratum moleculare) virtually normal. This fits with the observation in this study of no unequivocal degeneration in stratum moleculare of Ammon’s horn when the projections from the entorhinal area were spared. The ipsilateral projections from the different parts of regio inferior appear to be organized parallel to the commissural projections. The results of this study indicate that the ipsilateral projections arising from the pyramidal cells in regio inferior closest to the hilus terminate in stratum radiatum of regio superior, and that the terminal zones of the pyramidal cells lying farther from the hilus shift from stratum radiatum to stratum oriens with in-

705

creasing distance of the cells from the hilus. This shift corresponds to the pattern described for the commissural projections t o regio superior. Commissural and intrinsic projections from the septal part of regio inferior

Little has been known about the organization of the commissural and intrinsic projections from the septal part of the regio inferior. However, Hjorth-Simonsen (’77) placed lesions in the rostral (septal) extreme of Ammon’s horn in the rabbit, and observed degeneration in the rostral two-thirds of stratum radiatum, most dense in the outer part of the layer. The degeneration was largely present in regio inferior but extended for a short distance into regio superior. This degeneration pattern in the stratum radiatum resembles the longitudinal association path described by Lorente de NO (‘34). In the present study on the rat both lesions and HRP injections in the septal part of regio inferior equally revealed a projection in the superficial part of stratum radiatum that could be followed farther temporally in regio inferior than in regio superior. It is therefore possible that the longitudinal association path of Lorente de NO has its origin exclusively in the septal part of regio inferior. After injection of tritium labelled amino acids into regio inferior Gottlieb and Cowan (‘73) saw that the projection to stratum radiatum of the contralateral regio superior varied in density along the subiculo-dentate axis. The portion near the subiculum in some cases contained the greatest number of grains, in other cases the lowest number, for unexplained reasons. In the present study, varying densities along the subiculo-dentate axis were found in the projection to the contralateral stratum radiatum of regio superior after localized lesions in regio inferior. The material indicates that these gradients are related to the septo-temporal position of the lesion. At levels increasingly temporal to a lesion in the extraventricular part of the septal regio inferior, the degeneration diminished and disappeared from that part of the stratum radiatum abutting on regio inferior. The most temporal part of the projection was confined to stratum radiatum adjoining subiculum. In cases with lesions of the extraventricular part of regio inferior a t middle dorso-ventral levels, t h e degeneration in stratum radiatum of regio superior was of equal density along most of the subiculo-

706

S. LAURBERG

dentate axis, and only lessened near the subiculum.

missural fibers arrive later to their target than the ipsilateral fibers.

Comparison between all commissural and ipsilateral projections from regio inferior

The general organization of the projection:: to the hippocampus The general hodological relation between the parts of the hippocampal region has now become more readily discernible than before. In particular this concerns the relations of the area dentata. It seems appropriate to end W Ith a brief synoptic treatment of this subject. A massive input to the hippocampal region originates in the entorhinal cortex. (Other less abundant afferents from the brain stem must be omitted from consideration a t this moment.) It is well established that the entorhinal area has excitatory ipsilateral and partly contralateral projections to the fascia dentata (as well as to the Ammon’s horn and the molecular layer of the subiculum) (Andersen et al., ’66;Lomo, ’71a,b; Hjorth-Simcinsen, ’72; Hjorth-Simonsen and Jeune, ’72; Steward, ’76; Steward and Scoville, ’76). The fascia dentata and the hilus are intimately .nterconnected by two-way connections, which need not be detailed here. The granule cells project to the ipsilateral hilus. According to Blackstad et al. (‘70) and Gaarskjaer (‘78) this projection shows only a moderate septo-temporal spread. In contrast, the hilus cells have both ipsilateral and commissural projections to the area dentata, both showing a large septo-temporal divergence. Swanson and Cowan (‘77) after injection of tritiated amino acids into CA4 observed no labelled extrahippocampal projections. The present study additionally indicates t h a t there is no substantial projection from the hilus fasciae dentatae to the Ammon’s horn and that the only major intrahippocampal connections of the hilus are the bilateral projections to the dentate area. This implies that the hilus neurons take part only in the local circuitries of the areae dentatae themselves.2 Since the hilus fasciae dentatae has no major connection with the Ammon’s horn, and the Ammon’s horn does not project to the area dentata, the mossy fibers are the only major connection between the area dentata and the Ammon’s horn. Hjorth-Simonsen (‘73) and Swanson and Cowan (‘77) describe a unidirwtional organization of the ipsilateral projec-

The commissural and ipsilateral projections from regio inferior appeared very similar irrespective of the placement of the lesion along the subiculo-dentate and septo-temporal axes. In all cases where comparison was possible the two types of projection displayed equivalent variations along the subiculo-dentate axis of the Ammon’s horn, and their septo-temporal divergence was of the same order of magnitude. Moreover, the variations in laminar distribution were similar. At levels temporal to the lesion, both projections to the stratum radiatum of the Ammon’s horn appeared concentrated on the distal part of the pyramidal cell dendrites, while septal to the lesion both were concentrated on the proximal parts of the pyramidal cell dendrites. A similar shift along the proximo-distal parts of the pyramidal cell dendrites was observed in both projections to the stratum oriens of regio inferior when the ventricular part was lesioned. It is of interest that both of the ipsilateral and commissural projections to stratum radiatum, a t levels temporal to the cell of origin, concentrate on the distal parts of the apical dendrites. In analogous manner, the temporal part of the Ammon’s horn and the distal part of the pyramidal cell dendrites apparently receive intrinsic/commissural projections before the septal parts do (Zimmer and Haug, ’78).This septo-temporal gradient, common t o the shorter ipsilateral projection and the longer (later arriving) commissural projection, therefore can hardly be fully explained by temporal factors. The only major difference between the ipsilateral and commissural projections to Ammon’s horn was that the commissural projections from the ipsilateral ones shifted from stratum radiatum to stratum oriens in both regio superior and regio inferior as described above. It has been reported that the apical dendrites of the Ammon’s horn pyramidal cells develop earlier than the basal dendrites (Purpura and Pappas, ’68; Stensaas, ’68; Minkwitz, ’76). The general shift toward stratum oriens of the commissural projections from all parts of regio inferior could be explained by temporal factors if the com-

‘The connections of the rat hilus fasciae dentatae described in this study therefore fit well with the conception of the hilua as being a part of an “area dentata” (Blackstad,‘56) rather than as a part of the Ammon’s horn (Lorente de NO, ’34, and others).


tions, regio inferior supplying regio superior and the Ammon’s horn the subiculum. This investigation demonstrates a one-way connection even between area dentata and regio inferior. It is striking and hardly not without important functional implications that the area dentata, the Ammon’s horn, and the subiculum all receive an input directly from the entorhinal area and that each of these areas has a unidirectional connection with the next, in the direction from the area dentata to the subiculum. ACKNOWLEDGMENTS

This study was partly supported by USPHS Research Grant NS07998 to Professor Th. W. Blackstad, to whom I am gratefully indebted. I thank Doctor A. Hjorth-Simonsen, Doctor M. West for valuable suggestions while the manuscript was prepared, and Mrs. E. Kjar Hansen, Mrs. L. Munkk, Miss M. Sorensen, Miss K. Wiedemann, Mr. B. Krunderup and Mr. A. Meier for technical assistance. LITERATURE CITED Andersen, P., B. H. Bland and J. D. Dudar 1973 Organization of th e hippocampal output. Exp. Brain Res., 17: 152-168. Andersen, P., B. Holmqvist and P. E. Voorhoeve 1966 Entorhinal activation of dentate granule cells. Acta Physiol. Scand., 66: 448-460. Angevine, J. B., Jr. 1965 Time of neuron origin i n t h e hippocampal region. An autoradiographic study in t he mouse. Exp. Neurol., Suppl. 2. Blackstad, T. W. 1956 Commissural connections of t he hippocampal region in the rat with special reference to their mode of termination. J. Comp. Neur., 105: 417-538. Blackstad, T. W., K. Brink, J. Hem and B. Jeune 1970 Distribution of hippocampal mossy fibers in t he rat. An experimental study with silver impregnation methods. J. Comp. Neur., 138: 433-450. Cajal, S. Ramon y 1893 Estructura del asta de Ammon. Anal. Soc. esp. Hist. Nat. Madr., 22: 53-114. English translation in Cajal, ’68. 1968 The Structure of Ammon’s Horn. Charles C Thomas, Springfield, Illinois, 78 pp. Fricke, R., and W. M. Cowan 1977 An autoradiographic study of t h e development of t h e entorhinal and commissural afferents to the dentate gyrus of the rat. J. Comp. Neur., 173: 231-250. Gaarskjaer, F. H. 1978 Organization of t he mossy fiber system of the rat studied in extended hippocampi. 11. Experimental analysis of fiber distribution with silver impregnation methods. J. Comp. Neur., 178: 73-88. Gottlieb, D. I., and W. M. Cowan 1972 Evidence for a temporal factor in t h e occupation of available synaptic sites during th e development of t he dentate gyrus. Brain Res., 41: 452-456. 1973 Autoradiographic studies of t h e commissural and ipsilateral association connections of t he hippocampus and dentate gyrus of t he rat. I. The commissural connections. J. Comp. Neur., 149: 393-422. Haug, F.-M. 9. 1974 Light microscopical mapping of the

707

hippocampal region, the pyriform cortex and the corticomedial amygdaloid nuclei of the rat with Timm’s sulphide silver method. I. Area dentata, hippocampus and subiculum. Z. Anat. Entwick1.-Gesch., 145: 1-27. Hine, R. J., and G. D. Das 1974 Neuroembryogenesis in the hippocampal formation of the rat. Z. Anat. Entwick1.Gesch., 144: 173-186. Hjorth-Simonsen, A. 1970 Fink-Heimer silver impregnation of degenerating axons and terminals on mounted cryostat sections of fresh and fixed brains. Stain Tech., 45: 199-204. 1972 Projections of the lateral part of the entorhinal area to the hippocampus and fascia dentata. J. Comp. Neur., 146: 219-232. 1973 Some intrinsic connections of the hippocampus in the rat: An experimental analysis. J. Comp. Neur., 147: 145-162. 1977 Distribution of commissural afferents to the hippocampus of the rabbit. J. Comp. Neur., 176: 495-514. Hjorth-Simonsen, A., and B. Jeune 1972 Origin and termination of the hippocampal perforant path in the ra t studied by silver impregnation. J. Comp. Neur., 144: 219-232. Hjorth-Simonsen, A., and S. Laurberg 1977 Commissural connections of the dentate area in the rat. J. Comp. Neur., 174: 591-606. Laatsch, R. H., and W. M. Cowan 1967 Electron microscopic studies of the dentate gyrus of the rat. 11. Degeneration of commissural afferents. J. Comp. Neur., 130: 241-262. Lorente de NO, R. 1934 Studies on the structure of the cerebral cortex. 11. Continuation of the study of the ammonic system. J. Psychol. Neurol. (Leipzig), 46: 113-177. Lynch, G., C. Gall, R. Rose and C. Cotman 1976 Changes in the distribution of the dentate gyrus associational system following unilateral or bilateral entorhinal lesions in the adult rat. Brain Res., 110: 57-71. h m o , T. 1971a Patterns of activation in a monosynaptic cortical pathway: the perforant path input to the dentate area of t h e hippocampsl formation. Exp. Brain Res., 12: 18-45. 1971b Potentiation of monosynaptic EPSPs in the perforant path-dentate granule cell synapse. Exp. Brain Res., 12; 46-63. Mesulam, M. M. 1976 The blue reaction product in horseradish peroxidase neurohistochemistry: incubation parameters and visibility. J. Histochem. Cytochem., 24 (12): 1273-1280. Minkwitz, H.-G. 1976 Zur Entwicklung der Neuronenstruktur des Hippocampus wahrend der pra- und postnatalen Ontogenese der Albinoratte. I. Mitteilung: Neurohistologische Darstellung der Entwicklung langaxoniger Neurone aus den Regionen CA3 und CA4. J. Hirnforsch., 17: 213-231. Mosko, S., G. Lynch and C. W. Cotman 1973 The distribution of the septa1 projections to the hippocampus of the rat. J. Comp. Neur., 152: 163-174. Nauta, W. J. H. 1950 Uber die sogenannte terminale Degeneration im Zentralnervensystem und ihre Darstellung durch Silberimpragnation. Schweiz. Arch. Neur. Psychiat., 66: 353-376. Purpura, D. P., and G. D. Pappas 1968 Structural characteristics of neurons in the feline hippocampus during postnatal ontogenesis. Expt. Neurol., 22: 379-393. Raisman, G., W. M. Cowan and T. P. S. Powell 1965 The extrinsic afferent, commissural and association fibres of the hippocampus. Brain, 88: 963-996. Schaffer, K. 1892 Beitrag zur Histologie der Ammonshorn-Formation. Arch. mikr. Anat., 39: 611-632. &gal, M., and S. Landis 1974 Afferent8 to the hippocampus

708

S. LAURBERG

of the r a t studied with t h e method of retrograde transport of horseradish peroxidase. Brain Res., 78: 1-15, Stensaas, L. J. 1968 The development of hippocampal and dorsolateral pallial regions of t h e cerebral hemisphere in fetal rabbits. VI. Ninetymillimeter stage, cortical differentiation. J. Comp. Neur., 132: 93-108. Steward, 0. 1976 Topographic organization of t h e projections from the entorhinal area to the hippocampal formation of the rat. J. Comp. Neur., 167: 285-314. Steward, 0..and S. A.Scoville 1976 Cells of origin of entorhinal cortical afferents to the hippocampus and fascia dentata of the rat. J.Comp. Neur., 169: 347-370.

Swanson, L. W., and W. M. Cowan 1977 An autoredi ographic study of the organization of the efferent connec tions of the hippocampal formation in the rat. J. Comp. Neur., 172: 49-84. Zimmer, J. 1971 Ipsilateral afferents to t h e commissural zone of t h e fascia dentata, demonstrated in decommissurated rats by silver impregnation. J. Comp. Neur., 142: 393-416. Zimmer, J.,and F:M. Haug 1978 Laminar differentation of the hippocampus, fascia dentata and subiculum in developing rats, observed with the Timm sulphide silver method. J. Comp. Neur., 179: 581-618.

s.

Intrinsic and commissural connections of the rat inferior colliculus.

Commissural connections between the auditory cortices of the rat.

Commissural connections of the dentate area in the rat.

The intrinsic, association and commissural connections of area 17 on the visual cortex.

Redirected perforant and commissural of connections of eutopic and ectopic neurons in the hippocampus of methylazoxymethanol-acetate treated rats.

Developmental studies of thalamocortical and commissural connections in the rat somatic sensory cortex.

The commissural fibre connections of the primary somatic sensory cortex.

Ultrastructure of neurons containing somatostatin in the dentate hilus of the rat hippocampus after cerebral ischaemia, and a note on their commissural connections.

Synaptic connections of neuropeptide Y (NPY) immunoreactive neurons in the hilar area of the rat hippocampus.

Commissural connections mediate inhibition for the computation of interaural level difference in the barn owl.

Target cell specificity of synaptic connections in the hippocampus.

Commissural responses of rat retrohippocampal neurons.

Correction: Corticospinal and Reticulospinal Contacts on Cervical Commissural and Long Descending Propriospinal Neurons in the Adult Rat Spinal Cord; Evidence for Powerful Reticulospinal Connections.

Direct efferent and afferent connections of the hippocampus with the neocortex in the marmoset monkey.

The reduction of volume and fiber bundle connections in the hippocampus of EGR3 transgenic schizophrenia rats.

Physiological studies of the reciprocal connections between the hippocampus and entorhinal cortex.

The connections of presubiculum and parasubiculum in the rat.

Association and commissural fiber systems of the olfactory cortex of the rat.

An autoradiographic study of the commissural and ipsilateral hippocampo-dentate projections in the adult rat.

Differential efferent connections of the brain stem to the hippocampus in the cat.

Lack of Intrinsic GABAergic Connections in the Thalamic Reticular Nucleus of the Mouse.

Intrinsic connections within the pedunculopontine tegmental nucleus are critical to the elaboration of post-ictal antinociception.

Effects of alcoholization on the rat hippocampus.

Synaptic connections formed by grafts of different types of cholinergic neurons in the host hippocampus.