THE JOURNAL OF COMPAHATIVE NEUROLOGY 197:Sl-98 (1981)

Connections of the Corticomedial Amygdala in the Golden Hamster. I. Efferents of the “Vomeronasal Amygdala”

GOLDA A. KEVETTER AND SARAH S. WIN.ANS Neurusciencr Program and Department of Anatomy, The University of Michigan, Ann Arbor, Michigan 48109

ABSTRACT The medial (M) arid posteromedial cortical (C3) amygdaloid nu- clei and the of the accessory (NAOT) are designated the ”vomeronasal amygdala” because they are the only components of the amygdala to receive a direct projection from the accessory (AOB). The efferents of M and C3 were traced after injections of “H-proline into the amygdala in male golden hamsters. Frozen sections of the were processed for autoradiography. The efferents of the “vomeronasal amygdala” are largely to areas which are primary and secondary terminal areas along the vomeronasal pathway, although the efferents from C3 and M terminate in different layers in these areas than do the projections from the vomeronasal nerve or the AOR. Specifically, C3 projects ipsilaterally to the internal granule cell layer of the AOB, the cellular layer of NAOT, and layer Ib of M. Additional fibers from C3 terminate in a retrocommis- sural component of the bed nucleus of the (BNST) bilaterally, and in the cellular layers of the contralateral C3. The medial nucleus projects to the cellular layer of the ipsilateral NAOT, layer Ib of C3, and bilaterally to the medial component of RNST. Projections from M to non-vomeronasal areas terminate in the medial -anterior hypothalamic junction, ventromedial nucleus of the , ventral premammillary nucleus and possibly in the ventral These results demonstrate reciprocal connections between primary and sec- ondary vomeronasal areas and between the secondary areas themselves. They suggest that M, but not C3,projects to areas outside this vomeronasal network. The medial amygdaloid nucleus is therefore an important link between the vo- meronasal organ and areas of the not receiving direct vomeronasal input.

Two separate chemosensory organs, the ol- together they provide information necessary factory and vomeronasal, are found in the nasal for the maintenance of sexual behavior iWi- cavities of most mammalian species (McCot- nans and Powers, ’77). ter, ’12; Wysocki, ’80). They have been im- Of the many areas in the ventral forebrain plicated in various species-specific behaviors that receive input from the olfactory and vo- and endocrine functions (Johnset al., ’78;Flem- meronasal pathways, the amygdala appears to ing et al., ’79; Reynolds and Keverne, ’79).Ear- be particularly important in mediating the lier work involving removal of the olfactory chemosensory control of copulation in male bulbs produced deficits in male and female sex- ual behavior and aggressive and maternal be- haviors (for review see Alberts, ’74; Cain, ’74). This work was part of a dissertation submitted in partial fulfillmerit The emphasis in this laboratory has been to of the requirements for the degree of Doctor of Philosophy in Neuro- elucidate the role of chemosensory structures science, The Ilniversity nf Michigan in the sexual behavior of male hamsters. The results of behavioral studies on male hamsters Current address: Marinc Biomedical Institute, 200 Univcrsity with peripheral deafferentation of the ol- tloulevard, Galveston, Texas 77550 factory, vomeronasal, or both olfactory and Please send reprint requests to Ur. Sarah S Winans, Department of vomeronasal organs suggest that these two sys- Anatomy, 5705 Medical Science 11, The University of Michigan, Ann tems are complementary in function and t,hat Arbor. MichrFan 48109.

0021-9967/81/1971-0081$05.00 0 1981 ALAN R. LISS, INC. 82 G.A. KEVE’ITER AND S.S. WINANS hamsters. Devor 1’73) studied mating behavior and olfactory areas in the amygdala is main- in these animals after lesions of the lateral tained in non-converging projections from olfactory tract (LOT), which conveys fibers those two areas in the rat and cat. The only from both the main olfactory bulb and acces- point of interaction identified in these species sory (vomeronasal) olfactory bulb to the piri- was the projection of the endopiriform nucleus form lobe. When Devor cut the LOT im- (which receives input from the olfactory cortex) mediately caudal to the olfactory bulbs, the to the vomeronasal amygdala (Krettek and males could not locate buried food pellets. They Price, ’77b). showed little sniffing investigation and no The present paper describes the projections of mounts with a receptive female. When the LOT the medial and posteromedial cortical nuclei, was cut at the level of the caudal olfactory rostral and caudal components of the “vomero- tubercle, the males found the buried food, in- nasal amygdala.” The companion paper vestigated the receptive female and showed (Kevetter and Winans, ’81) examines the con- some aspects of sexual , however these nections of C1 and C2 in the “olfactory amyg- males also failed to engage in mounts, intro- dala.” missions or . Although the caudal LOT cuts in these animals spared main ol- MATERIALS AND METHODS factory bulb input to the anterior olfactory nu- cleus, rostral and olfactory Electrolytic lesions were placed in 21 adult tubercle, they interrupted vomeronasal system male hamsters, Mesocricetus aumtus, anesthe- efferents from the accessory olfactory bulb to tized with 50 mg sodium pentobarbitallkg body the amygdala and bed nucleus of the stria ter- weight. Anodal direct current of 0.5-1.5 mA minalis, as well as main olfactory bulb fibers to was passed through epoxy-coated insect pins the amygdala, posterior piriform cortex and (371 pm diameter) for 5-15 seconds to produce . Lehman et a1 (’78) hy- lesions in selected parts of the amygdala. pothesized that of these caudal areas, the The animals were anesthetized 2-8 days amygdala may be of major importance in after surgery and perfused through the heart mediating olfactory and vomeronasal influ- with isotonic saline followed by 10% phos- ences on sexual behavior. Their studies showed phate-buffered formalin. Brains were re- that male hamsters with bilateral lesions of the moved, embedded in an egg-yolk-gelatin mix- rostral corticomedial amygdala, particularly ture, and sectioned in the coronal plane on a the medial nucleus, failed to mate, whereas freezing microtome at 25 pm. Sections 125 pm animals with sham operations or lesions of only or 250pm apart were stained by the Fink- the caudal corticomedial amygdala mated Heimer I1 technique (Fink and Heimer, ’67). normally. Adjacent sections were stained with cresyl vio- Although the corticomedial amygdala re- let to determine the extent and accuracy of the ceives both vomeronasal and olfactory inputs, lesion and for cytoarchitectonic localization of the fibers from the two systems terminate in degenerating fibers and terminals. adjacent, non-overlapping areas. The main ol- Autoradiographic studies were done on 50 factory bulb projects to the lateral part of the male hamsters. One to five microcuries of corticomedial area-the anterior cortical (C1) tritiated proline [L-[3,4-’H(N)]-Proline;25- and the posterolateral cortical (C2) nuclei- 50 Cihmole; New England Nuclear) in 0.01- while the accessory olfactory bulb projects to 0.05 pl physiological saline were stereo- the medial area-the medial nucleus (M), and taxically injected over 45 minutes into one of the posteromedial cortical nucleus (C3) (Scalia the four nuclei of the corticomedial amygdala and Winans, ’75; Broadwell, ’75; Oevor, ’76). via a 1 pl or 5 p1 Hamilton syringe. Survival Thus, although behavioral evidence after times varied from six hours to six days. Most deafferentation suggests that these two sys- animals were perfused at 12, 24, or 48 hours. tems are both involved in chemosensory control Perfusion with formalin and embedding of the of reproductive behavior, no convergence of the brains were performed as described above. two pathways has been demonstrated to the Twenty micrometer frozen sections, 100 or level of the amygdala. The present study was 200 pm apart, were mounted on gelatin-coated initiated to determine the projections of the slides, coated with Kodak NTB2 emulsion, and vomeronasal and olfactory nuclei of the exposed for %6 weeks at 4°C. The slides were corticomedial amygdala in the hamster. The then developed with Kodak D-19 developer, studies of Krettek and Price (’77a,b ’78a,b) and couterstained with cresyl violet, and observed de Olmos (’72)have suggested that, in general, microscopically under brightfield and darkfield the medialflatera1 segregation of vomeronasal illumination. EFFERENTS OF THE “VOMEKONASAL AMYGDALA 83 In 12 additional hamsters, electrolytic le- M. Additional brains with injections or lesions sions were stereotaxically placed in the stria involving areas adjacent to M or C3, but not terminalis. Eight of these animals received in- involving these target nuclei, were examined jections of 2pCi tritiated proline in 20nl as controls. physiological saline in the amygdala two weeks after the lesion was made. The remaining four Projections of the medial nucleus animals received identical injections three Twelve animals had lesions or injections that days after the stria terminalis was damaged. primarily involved M. The maximum extent of Twenty-four hours after the proline injections, the injection in the coronal plane in two brains all 12 animals were perfused, and the brains is shown in Figure 1. The electrode or needle were processed for autoradiography as de- tracts commonly traversed the cortex, corpus scribed above. Sections of four of these brains callosum, , optic tract, tha- were a1 so processed for Fink-Heimer impreg- lamic reticular nucleus, and the hippo- nation. campus. Project.ions from these areas could be traced directly from the electrode tract in RESULTS Fink-Heimer material. All cases in which the The projections reported here were seen after stria terminalis was damaged were excluded a lesion or tritiated proline was placed in either from this part of the study. Projections of the the rostral (M) or caudal (C3) vomeronasal medial nucleus are shown diagrammatically on amygdala. Conclusions from the autoradio- the right side of Figures 2 and 3. graphic and Fink-Heimer impregnation tech- The pattern of terminal fields differed de- niques were nearly identical. The description pending on the amount of M involved. The data and the illustrations in this account are suggested that neurons in the rostral and cau- primarily from autoradiographic material. The dal parts of M project to different areas of the Fink-Heimer material was prepared for sup- brain, but this could not be determined with plemental analysis, and proved particularly certainty from the available material. useful in distinguishing efferent fibers from Efferent fibers of the medial nucleus travel terminal fields. In the following description the medial to the central nucleus in the most me- terms “terminals” and “terminal fields” are dial component of the stria terminalis to a posi- used interchangeably for the deposit of silver tion ventral to the fornix, and continue grains seen with the light microscope over rostrally in the medial third of the stria termi- cytoarchitectonically defined areas that re- nalis. At the level of the anterior thalamus, ceived projections from M and C3. these fibers exit and course ventromedially in The injection locus in either M or C3 was the middle portion of the exiting bundle to the defined as the region containingcell bodies that anterior hypothalamic-medial preoptic area appeared black due to heavy incorporation of junction (AH-MPOA), at the level of the optic tritiated proline (Cowan et al., ’72). Sur- chiasm (Figs. 2D, E; 4). A dense accumulation rounding this locus was a uniform distribution of silver grains in the AH-MPOA indicates an of silver grains, which diminished gradually area of termination at this level. Caudal to with distance from the tip of the needle. Nuclei AH-MPOA the number of silver grains di- adjacent to M or C3 were variably included in minishes dramatically, al though sparse and this diffusion zone; however, the specific nuclei evenly spread grains are present throughout involved differed from brain to brain. Conclu- the mediohasal hypothalamus. In the core of sions about the projections of a given nucleus the ventromedial nucleus of the hypothalamus were based on comparison of results from injec- Pigs. 3G, H, I; 5) and more posteriorly in the tions that involved that nucleus but had no ventral premammillary nucleus (Fig. 2J), a other areas in common. This resulted in the high density of silver grains was also observed. elimination of projections arising from over- In Fink-Heimer materi a1 degenerating fibers lapping areas. were found throughout the mediobasal hypo- With two exceptions the projection sites de- thalamic area. However, fine, dust-like de- scribed below appeared in at least ten of the 12 posits representing degenerating boutons were brains in which M was the focus of the injection seen only at the junction of the AH-MPOA, in and in eight of ten brains in which the injection the ventromedial nucleus of the hypothalamus, was centered in C3. The two exceptions were and in ventral premammillary nucleus. terminal fields over the nucleus of the acces- Some fibers in the most medial part of the sory olfactory tract, and contralateral bed nu- stria terminalis terminate in the medial part of cleus ofthe stria terminalis, which were seen in the bed nucleus of the stria terminalis (BNST) only eight of the 12 brains having injections in (Figs. 2C, D; 6A, B). The label is seen lateral to 84 G.A. KEVElTER AND S.S. WINANS the stria medullaris and extends ventro- After interruption of the stria terminalis medially under this fiber bundle. Other fibers (ST) Fink-Heimer impregnation demonstrated continue rostrally, decussate in the dorsal part degenerating argyrophilic particles in the of the , and terminate lat- AH-MPOA, ventromedial nucleus of the hypo- eral to the stria medullaris in the contralateral thalamus, premammillary nucleus, and BNST. BNST (Fig. 2D). This labeling is consistently In the hamsters in which the ST was inter- lighter and more caudal than that observed in rupted prior to the injection of tritiated proline the contralateral BNST after injections of C3. into the amygdala, silver grains were seen in After injections of M. silver grains were always the ST up to the point of the lesion but not seen in the anterior commissure, but the con- beyond. This was in striking contrast to the tinuation of this labeling into the contralateral heavy labeling of fibers in the ST, which dis- BNST was not always apparent. tributed to the mediobasal hypothalamus and

Abbreuiations AC, anterior fornix commissure lateral ACC , nucleus nucleus of accumbens the amygdala AH, anterior LOT, olfactory AHA, amygdalo-hippocamp~il tract area medial AOB, accessory nucleus of olfactory bulh the amygdala AOBgr, internal MPOA, medial granule cell preoptic area layer, MPOA-AH, medial Accessory preoptic olfactory bulh area-anterior BLa, anterior hyothalamic division junction hasolatern1 NAOT. nucleus of nucleus of the accessoly the amygdala olfactory BLP, posterior tract division NLOT, nucleus of basolateral the lateral nucleus of olfactory the amygdala tract BNST, bed nucleus optic chiasm of the stria optic tract terminalis olfactory c1, anterior tu txrcle cortical ventral nucleus of premammillary the amygdala nucleus c2, posterolateral subiculum cortical stria medul- nucleus of laris the amygdala stria C3, posteromedial terminalis cortical tenia tecta nucleus of ventromedial the amygdala nucleus of CA1, hippocampal the field hypothalamus Ce, central nucleus of the amygdala CPF, piriform cortex EC . entorhinal cortex EFFERENTS OF THE “VOMERONASAI, AMYGDALA” 85

Fig. 1. Brightfield photomicrographs and matching tracing of injection sites centered in M and C3. Area of stippling corresponds to locus of the injection as defined by heavily labeled neurons. B and D show two of the largest injection sites used for this analysis. A. animal 972 pCi H-proline in 20 nl physiological saline. R. animal 57;5 p .‘H-prolinein 50 nl physiological saline. C. animal 101; 2 pCR ’H-proline in 20 nl physiological saline. D. animal 95; 2 pCi ’H-proline in 20 nl physiological saline. Animals 97, 57, and 95 survived 48 hours; animal 101 survived 12 hours. 86 G A KEVETTER AND S S WINANS

A

B

C

D

E

F

Fig. 2 and 3. Schematic drawings of coronal sections of the hamster brain showing the neuronal projections from the medial nucleus on the right (closed circles) and those from the postemmedial cortical nucleus (open circles) on the left. EFFEKENTS OF THE “VOMERONASAL AMYCDALA” 88 G.A. KEVYTTER AND S.S. WTNANS RNST in experiments with injections of W- proline in M. These results suggested that most, if not all, of the medial hypothalamic projections from the amygdala travel through the stria terminalis. The medial nucleus also sends efferents to C3 and may project further caudally to the sub- iculum. These fibers do not travel in the stria terminalis; most course caudally at the medial corner of the amygdalo-hippocampal area (AHA) and eventually terminate in layer Ib of C3 (Fig. 7A, U). Analysis of these brains re- vealed a lightly labeled field of silver grains in the molecular layer of the ventral suhiculum. In brains in which the injection was almost entirely restricted to M, this field was faint; therefore it is likely that other amygdaloid nu- clei included in the large injections contributed substantially to this projection. Fibers from M also terminate rostrally in the cellular layer of the nucleus of the accessory olfactory tract INAOT). This nucleus is located ventral and rostra1 to M and therefore was often labeled by diffusion from the injection site. However, the projection from M to NAOT was apparent in both Fink-Heimer sections and in brains with limited injections of M that were processed autoradiographically. Projections of the posteromedial cortical nucleus Ten animals had ,"H-proline injections into C3. The injection sites in two of these brains are shown in Figure 1, and the projections of C3 are diagrammed in Figures 2 and 3 on the left. The needle tract passed through the cortex, corpus callosum, and internal capsule in all cases. The and entorhinal cortex were also traversed in several cases. Projec- tions of these areas were distinguished from those ofC3 using the criteria described earlier. Efferents from C3 travel medially and rostrally along the ventral surface of the brain to terminate in layer Ib on the ventral and medial surfaces of M. Some fibers from this group continue rostrally to terminate in the cellular layer of the NAOT (Figs. 2E 8). Other fibers from C3 enter the stria terminalis travel- ing dorsally and rostrally through the central nucleus of the amygdala to enter the sub-

Fig. 4. Junction of the anterior hypothalamus and me- dial preoptic area (MPOA-AH) after medial nucleus injection in hamster 70.A. eresyl violet. A. darkfield photomicrograph of coronal section in (A). Compare to Figure 2E. EFFERENTS OF THE 'VOMERONASAL .WYGDALA" 89

i 90 G.A. KEVETTER AND S.S. WINANS EFFERENTS OF THE "VOMERONASAL AMYGDALA" 91 92 G.A. KEVETTER AND S.S. WINANS

Fig. 8. Nucleus of the accessory olfactory tract (NAOT). Label which resulted from an injection of C3 in hamster 95. A. cresyl violet. B. darkfield photomicrograph of coronal section in (A).Compare to Figure 2E. EFFERENTS OF THE "VOMERONASAL AMYGDALA" 93 ventricular component of this bundle (de the cellular layers. As illustrated in this exam- Olmos, '72). Some ST fibers end ipsilaterally in ple, the terminals from M and C3 also do not a terminal field in a lateral part of medial overlap. This is true in all cases except in the BNST at the level of the decussation of the projections of M and C3 to NAOT, which appear anterior cornmissure, while others decussate in to be nearly congruent. In the case of the BNST, the dorsal part of the commissure and termi- the small subarea of BNST that receives direct nate contralaterally in the corresponding por- input from the AOB is at the caudal edge of the tion of the retrocornmissural BNST (Figs. 2C, area that receives from M, and this area, in 6C, D). Additional fibers from the anterior turn, appears to extend only slightly into the commissure enter the central subventricular more rostral BNST, in which fibers from C3 component of the contralateral stria terminalis terminate. and travel caudally to the cellular layer of the contralateral C3 (Figs. 31, J, K; 7C, D). Ipsilat- Accessory olfactory bulh era1 to the injection, some fibers were observed In the hamster only C3 of the corticomedial to follow the intermediate olfactory tract, an amygdala was found ta project to the AOB (Fig. anterior extension of the parolfactory stria 9). terminalis (de Olmos, 72; de Olmos and In- After horseradish peroxidase injections into gram, '72). These fibers continue rostrally the AOB of the rat, de Olmos et al. ('78) iden- within the olfactory peduncle ventral to the tified a projection from M and NAOT, as well as ependymal cells, and their terminals are found from C3, back to the AOB. We did not observe in the internal granule cell layer of the ac- this projection in the hamster, but de Olmos et cessory olfactory bulh (AOB) (Fig. 9). al. mentioned that while HRP was located in many cells of C3 and NAOT, only a few cells DISCUSS I0N were labeled in M. Thus it is possible that if this projection exists in the hamster, it was too The AOB, which receives peripheral input lightly labeled in our material to be distin- exclusively from the VNO, projects to the guished from background labeling. NAOT, M, C3, and BNST. The projections de- The projection of C3 back to the AOB pro- scribed here from M and C3 provide anatomical vides for feedback control of the output of the evidence for a vomeronasal network that inter- bulb. The vomeronasal nerves terminate in the connects all of these AOB projection areas !Fig. glomerular layer of the AOB on dendrites of the 10). Both M and C3 project to NAOT, the ipsi- rnitral cells (Barber and Raisman, '74; Barber lateral and contralateral BNST, and to each and Field, '751, which in turn project to the other. In addition, C3 projects to the contra- BNST, M, and C3 (Scalia and Winans, '75; lateral C3, and back to the AOB. With the ex- Rroadwell, '75; Devor, '76). Neurons of C3 pro- ception of the medial arnygdaloid nucleus, ject back to the internal granule cells of the which projects to the preoptic area and medial AOR (Raisman, '72; Barber and Field, '75). In basal hypothalamus, no area that receives from the MOB, the granule cells have been reported the AOB projects to areas outside of this to have an inhibitory effect on the mitral cells vomeronasal network. The observation that M, (Rall and Shepherd, '68; Shepherd, '72). Thus if but not C3, projects to the MPOA and medial the functional connections within the AOB are hypothalamus may explain the findings of similar to those of the main bulb, C3 may have Lehman et al. ('781, that the rostral cortico- an inhibitory effect on the output of the AOB. medial amygdala plays a more important role in the chemosensory control of copulatory be- Nucleus of the accessory olfactory tract havior in male hamsters than does the caudal The nucleus of the accessory olfactory tract corticomedial amygdala. (NAOT)is a small cluster of cells in the anterior Although M, C3, and the AOB project to one amygdala. Rostrally, it is ventral and lateral to another as well as to NAOT and BNST, the the nucleus of the lateral olfactory tract, and efferents from the amygdala and the AOB do more caudally it occupies a position ventral to not terminate in the same laminae or parts of M. A projection from AOB to the molecular these areas. In general, the AOR innervates the layer Ia of NAOT (Scalia and Winans, '75; distal dendrites of the neurons in its terminal Kevetter, '78) is complemented by previously areas, while the fibers from the amygdala ter- unidentified projections from both M and C3 to minate closer to, or on, the cell bodies. For the deeper cellular area of this nucleus. Of example, within C3, the AOB projects to layer these last two projections, the one from C3 to Ia; M, to layer Ib; and the contralateral C3, to NAOT is heavier and more easily identified 94 G.A. KEVEmER AND S.S. WINANS EFFERENTS OF THE “VOMERONASAL AMYGDALIY 95

Fig. 10. Schematic diagram of the ventral surface of the hamster brain showing the efferents of‘the accessory olfactory hulh fthin solid line), medial nucleus of the amygdala, M (heavy didlines), and posterornediai cortical nucleus, C3 (dashed lines).

than that from M. The NAOT was the only area outer molecular layer (Iai of both M and C3, M in the hamster in which the projections from M projects to layer Ib of C3, and fibers originating and C3 were found to share the same terminal in C3 terminate in the Ib-IIjunction ofM. Simi- field. lar intrinsic connections between the two vomeronasal amygdaloid nuclei were described A niygdnln for the rat by de Olmos 1’72). Krettek and Price The nuclei M and C.3 of the “vomeronasal (’78bi reported a projection from M to the mo- amygdala” are reciprocally connected by lecular layer iI) of C3, which presumably cor- intra-amygdaloid fibers, which terminate deep responds to the results observed here in ham- to the terminals of fibers from the AOB. While ster, in which this projection is restricted to projections from the AOB terminate in the layer Tb. Although they could not dissociate the 96 G.A. KEVE'ITER AND S.S.WINANS intra-amygdaloid projections of C3 and AHA, commissure or a component of the BNST lo- Krettek and Price ('78b! did describe a projec- cated within the commissure. Terminals on tion to the cellular layers (11 and 111) of M from such cells are also a possibility in the hamster, one or both of these nuclei. This dd'fers slightly since the labeling seen in the anterior commis- from the projection from C3 to layer Ib-11junc- sure, especially after injections involving M, tion of M in hamster. Fibers from C3 that enter was heavier than that seen leaving the com- the contralateral stria terminalis continue missure and terminating in the contralateral caudally to the amygdaloid region, where they BNST. end in the cellular layers of the contralateral C3, and this commissural projection is similar Mediobasal hypothalamus to that previously reported for the rat (de Olmos and Ingram, '72) and the rabbit ilammers, '72). After injections or lesions of M, fibers from the ventral stria terminalis course medio- Bed nucleus of the stria termimlis ventrally at the level of the anterior thalamus Medial and lateral subdivisions of BNST and terminate in the AH-MPOA junction area. were recognized by Krettek and Price ('7th) in This projection was seen in the hamster as pre- the rat and the cat on the basis of both cytoar- viously reported in the rat by Leonard and Scott chitectural differences and differences in the ('711, de Olmos ('721, and de Olmos and Ingram projections they received from the amygdala. ('72). The density of labeling decreases caudal The medial subdivision of RNST in those spe- to the AH-MPOA area and increases again in cies receives fibers from M, C3, and the AOB the region of the ventromedial nucleus of the (Broadwell,'75; Krettek and Price, '78a). In the hypothalamus, where both the cellular core of hamster M projects to the caudal BNST, lateral the ventromedial nucleus of the hypothalamus and ventral to the stria meduilaris at the level and a laterally adjacent area, Diepen's nucle- of the anterior thalamus, while C3 projects to a us tuberis lateralis, were heavily labeled after more rostral part of medial RNST, which is injections of M. Krettek and Price ('78a) con- ventral to the anterior commissure. The topog- sider this area a ventromedial extension of raphy of these projections differs from that ob- BNST; however, the terminal field includes the served in the rat by Krettek and Price ('78a), medial extent of the AH-MPOA region, adja- who described a medial to lateral (,ratherthan a cent to the ventricle, while the ventromedial rostral to caudal) organization for the termi- BNST, which receives a projection from C3, is nals from C3 and M, respectively. It is impor- located rostral to this AH-MPOA junction. tant to note that fibers that contribute the Projections from M to Diepen's nucleus have terminals in the AH-MPOA junction after in- also been seen in the rat (de Olmos, '72; de jections involving M exit more caudally and Olmos and Ingram, '72; Leonard and Scott, '71; laterally from the stria terminalis than do fi- Krettek and Price, '78a) and cat iKrettek and bers from C3 that project to this ventromedial Price, '784. portion of the rostral BNST. Although input from the posterior part of the Projections to the contralateral BNST decus- amygdala to the ventromedial hypothalamus sate in the dorsal component of the anterior was seen in the hamster, it appeared to arise in commissure and terminate in the medial divi- AHA, rather than from C3, as suggested by sion of the contralateral RNST with a rostro- Leonard and Scott ('71) and de Olmos ('72). In caudal organization, which overlaps the input the present study, injections of tritiated proline from the ipsilateral amygdala. In other words, involving AHA resulted in heavy projections to both ipsilateral and contralateral M terminate the shell of the ventromedial nucleus of the in a more caudal area than ipsilateral and hypothalamus, AH-MPOA, and the premam- contralateral C3. Amygdaloid projections to millary nucleus, whether C3 was involved or the contralateral BNST have not been de- not. These findings are similar to those of Kret- scribed in autoradiographic studies. After tek and Price ('78a) in rat and cat; however, small lesions in the corticomedial amygdala, those authors suggested that the projection to Leonard and Scott ('71) saw evidence of termi- the premammillary nucleus from the caudal nal degeneration in the anterior commissure of amygdala was from C3. In the hamster mate- the rat but could not trace degeneration to the rial, this latter projection also appears to origi- contralateral side. They concluded, as did Val- nate in AHA. verde ('65), that many of the fibers of this com- The projections of M to the AH-MPOA, missural component of the stria terminalis end mediobasal hypothalamus, and possibly to the on dendrites of the bed nucleus of the anterior subiculum were the only projections of the EFFEKENTS OF THE "VOMERONASAL AMYGDALA" 97 corticomedial amygdala to areas not receiving and the MPOA would be of major functional a primary or secondary projection from the significance, since bilateral lesions of the stria . Since lesions of the rostral, terminalis, in contrast to bilateral lesions of M, but not the caudal, corticomedial amygdala re- do not eliminate the male's copulatory behav- sult in marked deficits in tlhesexual behavior of ior. male hamsters iLehman et al., '781, it is rea- sonable to postulate that in the hamster a nec- essary pathway for the expression of mating ACKNOWLEDGMENTS behavior includes the AH-MPOA and projec- The authors would like to thank William tions from M to the AH-MPOA and mediobasal Rrudon for his assistance with the illustrations hypothalamus. This hypothesis is supported by and Bonita Johnson for typing the manuscript. several other lines of evidence. Lesions of This research was supported in part by MPOA and medial hypothalamus rostral to the NRSA grant TS2-MHI4279 to G.A.K., Rack- mammillary bodies severely disrupt copulatory ham dissertation grant to G.A.K. from Horace behavior (Heimer and Larsson, '67; Paxinos H. Rackham School of Graduate Studies, Uni- and Bindra, '73; Caggiula et al., '73),and stimu- versity of Michigan, and NIH grant RO1- lation of the MPOA facilitates sexual perform- NS14071 to S.S.W. ance of male rats (Malsbury, '71). In addition, estrogen and androgen concentrating neurons are found in the MPOA and in the medial nu- LITERATURE CITED cleus of the amygdala, and M contains the Alberta, J.R. i1974) Producing and interpreting experi- heaviest concentration of androgen target mental olfactory deficits. Phys. and Behav., 1Zt657-670. areas in the amygdala of the rat (Pfaff and Barber, P.C., and P.M. Field (1975) Autoradiographic dem- onstration of afferent connections of the accessory olfac- Keiner, '72; Grant and Stumpf, '75; Sar and tory bulb in the mouse. Brain Res., 85:201-203. Stumpf, '75; Krieger et al., '76). Barber, P.C., and G. Raisman (19741 An autoradiographic The pathway by which those neurons of M investigation of the projection of the vomeronasal organ to that mediate male sexual behavior reach the the accessory olfactory bulb in the mouse. Brain Kes., MPOA and hypothalamus is still to be deter- x1:21-30. Broadwell. R.D. I 1976) Olfactory relationships of the mined. Our experiments on the projections of telencephalon and diencephalon in the rabbit. I. An auto- the amygdala after severing the stria termi- radiographic study of the efferent connections of the main nalis suggest that the majority of the cells that and accessory olfactory bulb. J. Comp. Neurol., 163.32% 346. project to AH-MPOR and VMH send their Caggiula, A.R., S.M. Antelman, and M.d. Zigmond (1973) axons through the stria. This conclusion is en- Disruption of copulation in male rat.s afkr hypothalamic tirely in agreement with the findings of previ- lesions: A behavioral, anatomical and neurochemical ous investigators who used degeneration tech- analysis. Brain Res., 59.275287. niques. In the experiments in which the ST was Cain, D.P. (1974) The role of the olfactoly bulb in limbic mechanisms. Psych. Bull., 81:654-671. cut prior to injections of "H-proline, the longest Cowan, W.M.,D.I. Gottlieb, A.E. Hendrickson, J.L. Price, exposure time for the sections was two weeks, and T.A. Woolsey 119721 The autoradiographic demon- and the amount of ,'H-proline injected in these stration of axonal connections in the central nervous sys- animals was only 2~Ci-i.e., less than the tem. Brain Iles.. 37.21-51. amount injected in many of the animals used de Olmos, J.S. (19721The amygdaloid projection field in the rat iis studied with the cupric-silver method. In: The for identifying terminal fields. Thus, very Neurobiologyofthe Amygdala. R. Eleftheriou, ed. Plenum sparse projections to the medial hypothalamus Press, New York, pp. 145204. via a non-stria1 pathway may not have been de Olmos, J., H. Hardy, and L. Heimer (1978) The afferent observed in this material. connections of the accessory olfactory bulb formations in the rat: An experimental HRP-study. J. Comp. Neurol., In addition, recent studies using horseradish 181,213-243. peroxidase have provided positive evidence for de Olmos, J.S., and W.R. Ingram (1972)The projection field the existence of a small number of neurons of the stria terminalis in the rat brain. An experimental whose axons apparently take a non-stria1 study. J. Comp. Neurol., 146.30%-331. ('77) Devor, M. (19731 Components of mating dissassociated by route. McBride and Sutin found HRP label lateral olfactory tract transection in male hamsters. Brain in cells of both the basomedial and the medial Res., 64.437-441. amygaloid nuclei after injecting HRP into the Devor, M. 11976)Fiber trajectories of olfactory bulb efferents ventromedial hypothalamus of cats in which in the hamster. J. Comp. Neurol., l66:31~48. the stria terminalis had been cut. Preliminary Fink, K.P., and L. Heimer (1967) Two methods for selective silver impregnation of degenerating axons and their studies by Lehman in our laboratory suggest synaptic endings in the . Brain that this is also the case in the hamster. A Kes., 4.3fisL374. non-strial connection between the amygdala 98 G.A. KEVETTER AND S.S. WINANS

Fleming, A,,F. VaeCarino. L. Tamhosso, and P. Chee I 1979i Malsbury, C.W (19713 Facilitation of male rat copulatory Vomeronasal and olfactory sysbm modulation of mater- behavior by electrical stimulation of the medial preoptic nal behavior in the rat. Science, 203:37%374. area. Phys. Behav., 7.797 805. Grant, L.U., and W.E. Stumpf(1975) Hormone uplake sites McBrrde, R.T,., and J. Sutin (1977) Amygdalold and pontine in relation to CNS biogenic amine systcms. In: Anat.omica1 projections to the ventromedial nucleus of the hypo- Neuroendocrinology. W.E. Stumpf and L.D. Grant, eds. S. thalamus. J. Comp. Neurol., f 74:377-396. Karger, New York, pp. 445463. McCotter, K.E. (19121The connection of the vomeronasal Heimer, L., and K. Larsson (1967)Mating behavior of male nerves with the accessory olfactory bulb in the opossum rats after olfactory bulb lesions. Phys. and Behav.,Z.207- and other mammal?. Anal.. Rec., 6:29%:318. 209. Paxinos, G., and D. Rindra 11973) Hypothalamic and mid- Johns, M.A., H.H. Feder, B.R. Komisaruk, and A.D. Mayer brain neural pathways involved in eating, drinking, ir- 11978) [Jrine-induced reflex ovulNtion in anovulatnry rats ritability, , and copulation in rats. J. Comp. may he a vomeronasal effect. Nature, 272:44&448. Phys. Psych., 82r1-14. Kevetter, G.A. (1978) Efferents of the corticomdidial amyg- Pfa, D.W., and M. Keiner I 1972) Estradiol-concentrating dala in the golden hamster. Doctoral dissertation, Univer- cells in the rat amygdala as part of a limbic hypothalamic sity of Michigan. hormone-sensitive system. In: The Neurobiology of the Kevetter, G.A.. and S.S. Winans (1981)Connectir IS of the Amygdala. B. Eleftheriou, ed. Plenum Press, New York. corticomedial amygdala in the golden hamster. 11. Effe- pp. 775785. rents of the "olfactory amygdala." .J. Cornp. Neurol., Raisman, G. (1972) An experimental study of the projection 197:99-111. of the amygdala to the accessory olfactory bulb and its Krettek, J.E., and J.L. Price 11977a) Projections from the relationship to the concept of a dual . Exp. amygdaloid complex to the and thalamus Brain Res., 14.395408. in the rat and cat. J. Comp. Neurol., 172587-722. Rall, W., and G.M. Shepherd (1968) Theoretical reconstruc- Krettek, J .K., and J.L. Price il977b) Projections fmm the tion of field potentials and dendrodendritlc synilptic inter- amygdaloid complex and adjacent olfactoly structures to actions in olfactory bulb. J. Neurophysiol., 32.884915. the entorhinal cortex and to the subiculum in the rat and Reynolds, J., and E.B. Keverne (1979) The accessory olfac- cat. d. Comp. Neurol., 172:7%752. tory system and its role in the pheromonally mediated Krettek, d.E., and J.1,. Pricr (1978a) Amygdaloid projections suppression ofoestrus in grouped mice. J. Reprod. Fertil., to subcortical structures within the basal forcbrain and 57:31-35. brainstem in the rat and cat. J. Comp. Neurol.. Sar, M., and W.E. Stumpf (1975)Distribution of androgen- 178:225-254. concentrating neurons in rat brain. In: Anatomical Neuro- Krettek, J.R., and J.L. Price llY78bl A description of the endocrinolog-4-. W.E. Stumpf and L.U. Grant, eds. S. amygddoid complex in the rat and cat with observations Karger. Basel. DD. 12&13:j. , II on intra-amygdaloid connections. J. Comp. Neurol., Scalia:' P., and S.S. Winana (1Y75) The differential projec- 178:25&280. tions of the olfactory hulb and accessory olfactory bulb in Krieger, M.S., J.1. Morrell, and D.W. Pfaff(l976) Autoradio- mammals. J. Comp. Neurol., 1611:31-56. graphic localization of est,radinl-concentratingcells in the Shepherd, G.M. (1972) Synaptic organization of the mam- female hamster brain. Neuroendocrinology, 22 .19P205. malian olfactory bulb. Physiol. Rev., 52:864-917. Lammers, H.J. (1972) The neural connections of the amyg- Valverde, F. (1965) Studies on the piriform lobe Harvard daloid complex in mammals. In: The Neurobiology of the University Press, Camhridge, Mass. Amygdala. B. Elefteriou, ed. Plenum Press, Kew York, pp. Winans, S.S., and J.B. Powers (1977) Olfactory and vom- 12S144. eronasal deafferentalion of male hamsters. Histological Lehman, M.N., G.A. Kevetter. and J.B. Powers (1978)Sex- and behavioral analyses. Rrain Res., 126:32%344. ual and aggressive hehavior inmale hamsters aftcr lesions Wysocki, C.J. (19801 A review of the neurobehavioral evi- ofthe corticomedial amygdala. SOC.Neurosci. Prog. Ahst., denre for the involvement of the vomeronasal system in 251. mammalian reproduction. Neurosci. Hiobehav. Kev. (in Leonard, C.M., andJ.W. Scott (1971)Originanddistrihution press). of the amygdalofugal pathways in the rat: An experi- mental neuronnatomical qtudy. J. Comp. Neurol., 141:31% 330.