Some Claustro-cortical Connections in the Cat and Baboon as Studied by Retrograde Horseradish Peroxidase Transport D. RICHE AND J. LANOIR Departement d e Neurophysiologie appliquee, CNRS-LPN, 91 190 Gif-sur-Yvette and Departement d e NeurophysiologiegeneraleCNRS, 31 ch. Joseph Aiguier, 13274, Marseille Ce'dex 2 France
ABSTRACT 1. The mammalian Claustrum (C1) is a convergent multisensory structure of unknown function, and disputed ontogenetic origin. Its cortical projections, hitherto unknown, have been studied in cat and baboon by means of the horseradish peroxidase (HRP) technique. HRP was injected into the gyrus proreus (frontal eye field) of cats, and separately into the frontal eye fields, visual areas, and motor-premotor areas of the baboon cortex. 2. Differential retrograde transport to the C1 was demonstrated, such that in the cat the ipsilateral dorsal C1 was shown to be the principal origin of claustroproreate projections. In the baboon, the whole C1 projects onto area 8, while only the posteroventral part of the nucleus sends efferents to the visual cortex. The projection to the motor and premotor areas is present, but does not seem to be "essential." 3. Discussion of the physiological literature, together with anatomical evidence of reciprocal cortico-claustral projections to closely similar regions of the C1 lead to the suggestion that the C1 is concerned with the integration of messages subserving visually-directed movements. Some other functional implications are also discussed. The mammalian Claustrum (C1) is a nucleus with well-defined limits and not inconsiderable volume, especially in the cat. I t is situated laterally in the cerebral hemisphere, just below the neocortical mantle, limited medially by the external capsule and ventrolaterally by the extreme capsule. The cat C1 is inserted between the putamen and the ventral wall of the anterior sylvian gyrus and is in the form of a drop narrowed and elongated in the caudal direction. Two parts (Narkiewicz, '64; Druga, '66a; Chadzypanagiotis and Narkiewicz, '71) may therefore be recognized: the dorsal or insular C1, and the ventral or prepyriform C1. In the monkey and baboon, the C1 takes the form of a narrow sheet separating the putamen from the insular cortex; the anteroventral part, less well-defined than the rest, is considered as being simply adjacent to the amygdala (Rae, '54; Berke, '60). The ontogenetic origin of the C1 has been a matter of dispute for a century. Various theories, none of which J. COMP. NEUR. (1978)177: 435-444
can be regarded as proven, have been advanced: individualization from the deep layers of the insular cortex, a purely pallial origin, a mixed pallial and striatal origin (see historical review in Filimonoff, '66). From the hodological point of view, there is a striking dichotomy between the well-studied afferent connections and the almost totally unknown efferent projections. Cortical afferents in particular have been extensively investigated, in the rabbit (Carman et al., '64), the cat (Narkiewicz, '64, '72; Druga, '66b, '68, '71; Otellin, '69; Chadzypanagiotis and Narkiewicz, '71), and the monkey (Mettler, '47; Showers, '58; Berke, '60; Black and Myers, '62; Nauta, '62, '64; De Vito and Smith, '64; Astruc, '71; Kunzle, '75; Leichnetz and Astruc, '75; Kunzle and Akert, '77). As far as efferents are concerned, there is known to be a massive projection to the putamen in the cat, with sparing of the pallidum (Druga, '72). Data on neocortical projections in the cat
435
436
D. RICHE AND J. LANOIR
are limited to the retrograde degenerative changes observed by Narkiewicz ('64) after large neocortical ablations. A single observation also in the macaque by Kievit and Kuypers ('75) who noted that scattered C1 cells are sometimes labelled following injection of HRP into various areas of the rostral pole of the ipsilateral hemisphere. The present work offers new findings on t h e neocortical projection of the CI in the cat and baboon.
TABLE 1
Site ofHRPtnjection and survival time Case
c1 c2
c3
MATERIALS A N D METHODS
P1
Experiments were carried out on three cats and eight adolescent baboons (Papio papiol weighing between 3 and 7 kg. Unilateral intracortical injections of 30% horseradish peroxidase (HRP), (Sigma type VI) were made under Nembutal anaesthesia. In the cats (C1, C2, C3), a single injection was made stereotaxically (A 29, L 1) into t h e gyrus proreus with a glass micropipette of tip diameter approximately 30 pm. The micropipette was connected with the needle of a Hamilton syringe by a polyethylene tubing of the same diameter. The whole system was then filled up with coloured paraffin oil. A micrometer gage was used to push HRP into the cortex. I n t h e baboons (P1 to P8), a Hamilton syringe with attached needle was used to make freehand injections a t several points within a given cortical area. This technique made i t possible to fill almost the whole of the cortical tissue within a functional territory, and a t t h e same time to avoid subjacent structures. The injected cortical territories, in different animals, were: the frontal eye field, corresponding to the area 8 of Brodmann ('051, (P1, P2, P3), the visual areas (P4, P5) and the motor and premotor cortex (P6, P7, P8). Other details of t h e injections are given in table 1. At the end of t h e survival time, the animals were sacrificed under deep anaesthesia by intracardiac perfusion of 5-7 1 of a fixative consisting of 1%glutaraldehyde, 1%paraformaldehyde buffered to pH 7.2 with 0.12 M phosphate. The brains were removed and kept for 16 hours a t 4°C in a phosphate buffer solution containing 30%sucrose. They were then cut i n serial frozen section a t 40 p m , transversely in the case of all the cat brains and some of t h e baboon brains, while others of the latter were cut in parasagittal section. All sections from cat brains and one in five from baboon brains were treated histochemically by t h e method of Graham and Karnovsky ('66). Alternate sec-
P2 P3 P4
P5 P6 P7 P8
Areas
Quantity
injected
4
Survival time hours
Frontal eye field (+4J Head of Caudate n . Frontal eye field Frontal eye field ( + white m a t t e r ) 8 ( = frontal eye field) 8 8 + 9 ( + white matter) 17 ( + 18,191 17 ( + 18,19) 4 6+4 6+4
1
24
0.5 0.2
17 24
1
24
4 8
24 24
2.5
24
3.5
72
2.4 3 2.7
24 24 71
tions were lightly counterstained with cresyl violet. The labelled cells were mapped by using a n X-Y plotter coupled to the stage of a microscope equipped with a darkfield condenser. For the cat brains, frontal planes were defined a s in t h e atlas of Snider and Niemer ('61). The extent of the injection sites in cases C2 and C3 a r e illustrated in figure 1A on outlines of frontal sections. The injection site in C1 (not shown) was superimposable on that of C2, but in addition involved the dorsomedial edge of the head of t h e caudate nucleus. All 8 cortical injection sites in t h e baboons have been shown on a n outline of the left hemisphere in lateral view (fig. 1B). Frontal and parasagittal planes were defined with reference to the atlas of Riche e t al. ('68). RESULTS
1. Cat Injection of HRP into the gyrus proreus causes t h e labelling of a large number of cells in the dorsal part of the ipsilateral C1. These cells are found from the rostral pole of the nucleus back as far as plane A 11.5, but their distribution is not uniform. Almost all the peroxidase-positive cells are concentrated in the zone of constriction between the dorsal and ventral parts of t h e nucleus (fig. 2). This region of high density corresponds exactly with t h e "medial marginal zone" as defined by Druga ('66a). In case C3, with a small injection volume, there are only ten labelled cells,
CLAUSTRO-CORTICAL CONNECTIONS IN CAT AND BABOON
437
A bbreuratcons
aaa, area amydaloidea anterior ap, area prepyriformis as, arcuate sulcus CI, claustrum cp, cortex pyriformis cs, central sulcus FEF, frontal eye field gea, gyrus ectosylvianus anterior gem, gyrus ectosylvianus medialis gsa, gyrus sylvianus anterior
gsp, gyrus sylvianus posterior
GP, globus pallidus nam, nucleus amygdalae anterior nC, nucleus caudatus Pu, putarnen sp, sulcus pseudosylvius sr, sulcus rhinalis ss, sulcus Sylvius TO, tractus opticus VPM, ventroposterornedial nucleus of the thalamus
Fig. 1A Sites of HRP injection into the rostral pole of the left hemisphere in cats C2 andC3. The centre of the injection site is shown in black, while its periphery, colouredyellow in the sections, is dotted. In case C2, the track of the micropipette (A 29, L 1) is shown in section n"16. B Cortical injection sites in the baboon shown on a lateral projection of the left hemisphere.
but they are all in this restricted zone between frontal planes A 13 and A 12. In every case, the prepyriform part of C1 shows, on darkfield examination, a few labelled cells further posteriorly, between planes A 12.5 and A 11. The entire caudal pole of C1, posterior to plane A 10.5, was entirely devoid of labelled cells. The anterior sylvian gyrus (insular cortex), which is adjacent to the C1, shows a high density of labelled cells (fig. 2: A 15.5-A 13). In the ventral wall of this gyrus, and in the depths of the pseudosylvian sulcus, the labelled somata are located in the deep cell layers. In the contralateral hemisphere, two or three peroxidase-positive cells per section were observed in the rostral half of the C1, anterior to planes A 15-A 14.5, but with the same
preferential localization in the transitional zone between dorsal and ventral parts as on the ipsilateral side. Similarly, and in numbers roughly equal to those in the C1, labelled neurones could also be observed in the ventral wall of the anterior sylvian gyrus and the depths of the pseudosylvian sulcus. From the cytological point of view, at least two types of neurone accumulate HRP in the soma and dendritic tree. The majority are stellate cells of 20-25 p m diameter (figs. 3a,b). The other category, amounting to about 20% of the total, consists of large pyramidal neurones with a diameter of 20 p m x 35-40 pm (fig. 3c). There is possibly a third type of labelled cell, constituting about one tenth of the population and made up of small fusiform cells.
438
D. RICHE AND J. LANOIR
Fig. 2 Cartography of labelled cells in the claustrum and adjacent structures of cat C2. Borders and points have been traced on the X-Y plotter, each point representing the site of a labelled cell seen by darkfield examination. Each outline shows the total number of labelled cells in six successive sections. Frontal planes defined according to the atlas of Snider and Niemer ('61)
2. Baboon
Injections of HRP into frontal area 8 (fig. 1B: P1, P2, P3) result in the appearance of labelled cells in the whole of the ipsilateral C1. Case P3, which received t h e largest quantity of HRP (8 pl, the injected area including the underlying white matter), exhibited t h e highest density of labelled cells. They are most numerous in the anterodorsal region of t h e structure, which is the part nearest the site of injection. This preferential labelling can be clearly seen on the outlines of parasagittal sections from animal P3 (fig. 4A). In cases P 2 and P3, a very small number of labelled cells were also observed in the contralateral C1. The visual areas were injected in two baboons (fig. 1B: P4, P5). Although principally filling area 17, in both cases t h e HRP also
penetrated into areas 18 and 19. The distribution of labelled cells in t h e ipsilateral C1 differs from that following frontal injection, brightfield examination showing t h a t labelled neurones are concentrated in the posteroventral corner of t h e nucleus in parasagittal sections (fig. 4B). I n neither of these animals were any peroxidase-positive cells seen in the C1 contralateral to the site of injection. In three further animals, one part of the motor area (P6) and t h e motor plus a small part of the premotor area (P7, P8) were injected respectively. In the case of injection restricted to a part of area 4, no labelled somata were observed in the ipsilateral C1. When a greater part of area 4 and a part of area 6 were involved, peroxidase-positive neurones were present in t h e ipsilateral C1. Case P7 has shown only very few labelled cells. In
CLAUSTRO-CORTICAL CONNECTIONS IN CAT AND BABOON
439
Fig. 3 Darkfield photomicrographs showing retrograde labelling of ipsilateral claustro-cortical cells in the marginal zone of the CI. Cat C1, frontal plane A 14. a, b Coexistence of multipolar and pyramidal neurones. c A large pyramidal neurone with its apical dendrite.
Fig. 4 Cartography of labelled cells in parasagittal sections of the injected hemisphere in the baboon. A Following HRP injection into area 8 (baboon P3), labelled cells can be seen throughout the claustrum. B Following HRP injection into the visual areas (baboon P5), almost all the labelled cells are concentrated in the ventral part of the claustrum a t posterior levels. Each outline shows the total number of labelled cells in four successive sections. Parasagittal planes defined according to the atlas of Riche e t al. (‘68).
case P8, labelling was evident but seemed, however, to spare t h e ventral region of the structure. Following both frontal and occipital HRP injection, different neurone types within t h e C1 showed peroxidase labelling. Fusiform,
ovoid, or polygonal cells were filled, and were classified according to their diameter as “small” (10-20 p m ; fig. 5 above) and “large” (20-30 p m ; fig. 5 below). The distribution of these cell types within t h e C1 appears to be diffuse, with no obvious organization. However,
440
D. RICHE AND J. LANOIR
Fig. 5 Brightfield photomicrographs showing various types of ipsilateral claustro-cortical neurones in the baboon. a Two specimens of small (above) and large (below) neurones. b Different cellular types may be observed close to one another, a phenomenon which occurred throughout the entire structure and for each site of injection.
peroxidase-positive large ovoid cells were frequently observed on edge of the nucleus. DISCUSSION
(1) The results of our cat experiments demonstrate the existence of a direct projection from the C1 to gyrus proreus. Even when t h e volume of tissue injected was very small (C31, a t least two types of relatively large C1 neurones - mainly multipolar, and to a lesser extent pyramidal - were labelled. In Nissl sections of the C1, these two cell types have been observed mingled with smaller round or fusiform cells (Chadzypanagiotis and Narkiewicz, ’71). The zone of the C1 in which labelled cells were found in our experiments is, according to Druga (’66a), t h e region with t h e highest cell density. Furthermore, of t h e five cell types identified by this author, t h e large multipolar and pyramidal cells are precisely those which
predominate in this “medial marginal zone.” The rare small fusiform cells showing peroxidase activity have also been described in normal material by the above-named authors. While the projection of the C1 on t h e gyrus proreus may be a new finding, i t was 11 years ago t h a t Druga (’66b) demonstrated connections i n the opposite direction as a result of making localized lesions in the frontal cortex of the cat. His series included two cases (K31, K51) in which the lesion corresponded to the site of HRP injection i n two of our animals ((22, C3). The region of the C1 in which he found terminal axon degeneration is the same as t h a t in which labelled neurones were demonstrated in our material. However, while terminal degeneration following a lesion of gyrus proreus was confined to t h e rostra1 quarter of t h e insular C1, in our material labelled cells were absent only from t h e caudal pole of the
CLAUSTRO-CORTICAL CONNECTIONS IN CAT AND BABOON
structure. I t may be concluded that the claustral territory projecting to gyrus proreus includes the regions which receive reciprocal afferents from gyrus proreus, but it also extends further caudally. In the same publication, Druga (’66b) maintained that degenerated terminals surround claustral cells of different types. I t is of interest to speculate a s to whether the various neuronal categories contacted by the terminals of cortical efferents may be the same as those showing peroxidase accumulation. In such a case, the cortico-claustral contacts (chiefly axodendritic) might be in the neuropile of those claustral neurones which project reciprocally on the cortical area of origin. Such a hypothesis is also compatible with the electron microscopic observations of Narkiewicz (’721, who showed that the presynaptic endings in the C1 which degenerate following ablation of the proreate and sigmoid gyri are axodendritic and make asymmetric synapses. They were principally seen on fine dendritic arborisations and spines, more rarely on large dendrites. The author concludes that “These findings seem to prove that frontoclaustral connections do exist and that they are mostly if not exclusively axodendritic of a n asymmetrical type which according to Gray’s hypothesis may suggest that they are excitatory (Gray, ’591.”As in the specific sensory nuclei of the thalamus, there are many topically-organized closed circuits connecting certain zones of the C1 with their respective cortical projection area. The last important point is that our HRPinjections into the gyrus proreus cover in whole, or in part, the territory recognized as being the frontal oculomotor area (Schlag and Schlag-Rey, ’70; Brodal and Brodal, ’71). The existence of a massive monosynaptic projection of the insular cortex onto the gyrus proreus doubtless explains why some authors (Horcholle e t al., ’68; Imbert and Buisseret, ’74) situate the feline frontal eye field in the insulo-orbital cortex.’ (2) In the baboon, retrograde axonal transport of HRP has enabled us to demonstrate that an efferent claustral system projects to different neocortical areas. Following an injection which filled the whole of area 8, labelled cells were found throughout the C1, so that it may be concluded t h a t the entire C1 projects to area 8. The existence of a projection in the opposite direction, from frontal cortex to the C1, was suggested by Berke (’60).
44 1
However, his interpretation of results following various frontal and temporal ablations failed to make clear whether the degenerating axons were terminal, or represented fibres of passage. Scollo-Lavizzari and Akert (’63) did not include the C1 in their analysis of the subcortical projections from area 8, but later Nauta (’64) thought that there was “little doubt that the prefrontal cortex entertains direct efferent connections with the Claustrum” while Astruc (’71) was affirmative: “there is a substantial projection from area 8 to the Claustrum.” Using the autoradiographic tracing technique, Kunzle and Akert (’77) confirmed recently the existence of a direct projection from area 8 in Macaca fascicularis: two foci of moderate grain density were seen in the ipsilateral C1, one in the central moiety, the other in the most ventral and caudal part. In case P3, the high density of peroxidase-positive neurones in the anterodorsal region of the C1 must be related to the fact that the injected HRP also penetrated into area 9. This interpretation is confirmed by recent experiments in which this area was selectively injected. Conversely, Nauta (’64) has described cortico-claustral fibres originating in area 9. Results obtained after injections into motor and premotor areas lead us to think that - in comparison with little animals where the injected quantity of HRP must be very small (less than 0.5 pl) to accurately localize the origin of the projection (Sotelo and Riche, ’74) in the baboon it was necessary to inject a larger amount of HRP (several pl) to find substantial labelling. In the motor field of Papio, when a small part, restricted to the face area, was injected, no labelled cells could be observed (P6). For an injection site situated more dorsally and involving partially area 6 (P8), a significant labelling appeared. Concerning the reciprocal descending component, Nauta (’64) failed to observe fibre degeneration in C1 after lesions in area 6 and area 4 of the precentral gyrus. Using the autoradiographic technique with injections in the face area of the precentral motor cortex in the macaque, Kunzle (’75) found grain patches in the most rostrodorsal part of the C1. In good agreement with the localization of our HRP-positive neurones, he found also that the projec-
’
The simultaneous presence of labelled neurones in both the CI and the insular cortex following frontal HRP injection certainly dws not prove that the former derives from the latter, a s some authors believe (Filimonoff, ‘66).The finding nevertheless speaks in favor of a common relationship.
442
D. RICHE AND J. LANOIR
tions of the arm-trunk area seemed to cover approximately three-fifths of t h e dorsal C1. I n sum, reciprocal connections exist very likely in monkey and baboon between the C1 and t h e cortex of both the frontal pole (areas 8 + 9) and the precentral gyrus. The ascending component described here seems therefore to form part of a closed circuit. Occipital injections demonstrate a clear posteroventral localization of positive somata within t h e C1. There is thus a preferential relationship linking t h e posterior poles of both the C1 and the neocortex. Comparison of t h e distribution of labelled cells following frontal precentral and occipital injections respectively reveals a rostrocaudal organization of corticipetal efferents. Narkiewicz ('64) found topographically-organized retrograde degenerative changes in t h e C1 of the cat after extensive removals of some neocortical regions. Ablations of t h e sensorimotor cortex produced degenerations in the dorsorostral portion, removals of the visual region caused severe alterations in t h e dorsocaudal portion of the C1. For the author, "the simplest and most likely i n t e r p r e t a t i o n is t h a t t h e degenerations reveal t h e existence of direct claustro-cortical, or, possibly, cortico-claustral connections." The same antero-posterior organization is also described for projections in t h e opposite, cortico-claustral direction (Carman e t al., '64; Druga, '66b, '68; Narkiewicz, '72; Kunzle, '75). According to some authors, t h e rostro-caudal organization may also be associated with a medio-lateral one. So far as we know, there are no publications on connections between primary visual areas and the C1 in primates. I t is therefore necessary to discuss our results in relation to data on the cat. Chadzypanagiotis and Narkiewicz ('71), confirming t h e results of Druga ('68),described a projection of t h e visual cortex onto t h e dorsocaudal part of the C1. Furthermore, in explanation of precocious cellular (therefore retrograde) degeneration following cortical ablation, these authors suggested a reciprocal claustro-cortical component linking t h e same regions. Our results in the baboon confirm this hypothesis. I t may be postulated that t h e posteroventral part of the C1 is connected by a closed circuit with t h e ipsilateral visual areas. (3)The diversity of neuronal types in t h e primate C1 is not in itself a new finding (Rae, '54). We have yet been able to show t h a t several neurone types project to the frontal cortex, and t h a t apparently these same types send
also axons to the occipital cortex. Despite the rostro-caudal organization, neurones projecting both frontally and occipitally co-exist in t h e posteroventral part of the nucleus. For any given cell type, it remains to be seen whether independent somata project to each of the two cortical areas, or whether a single cell may project to both, in opposite directions, by means of a bifurcated axon. (4) When examined by evoked potential and single-unit methods, t h e C1 appears to be a multisensory nucleus, showing little spontaneous activity, on which converge somatic, auditory, visual, olfactory, gustatory, vestibular and viscero-vagal afferents (Bonvallet et al., '52; Segundo and Machne, '56; Spector, '65; Spector and Hassmannova, '66; Hassmannova and Spector, '67). The somaesthetic modality is predominant, activating almost all excitable neurones with short latency. In agreement with anatomical findings, cells of unimodal (lemniscal) type a r e preferentially localized in t h e rostra1 half of the C1, between planes A 17.5 and A 15 in the cat (Spector et al., '70, '74, '75).Units exhibiting heteromodal somatosensory or polysensory convergence are more numerous posteriorly, where visual evoked potentials have the greatest amplitude and shortest latency (Rapisarda e t al., '69). I t appears from the above discussion t h a t the C1 is connected through short loops to two cortical areas (FEF, primary visual cortex) involved in visual mechanisms. Because of the low frequency of spontaneous discharge, the transmission would appear to occur phasically, under special conditions, perhaps in conjunction with body, head or eye movements. This new feature has to be kept in mind when searching for the functional r61e of this nonhomogenous multisensory structure. As pointed out by Spector e t a1 ('751, t h e C1 is also connected with different thalamic nuclei and many other subcortical structures. I t seems concerned w i t h "long loops" (Buser a n d Imbert, '75) of different modalities; each loop relays through its specific thalamic nucleus and corresponding cortical area before relaying back to t h e C1. Even longer "visual" circuits might also involve t h e superior colliculus, the thalamic visual associative nuclei, and t h e posterior group (PO). Since a projection of both t h e proreate and orbital gyri onto t h e principal sensory (and interpolaris) trigeminal nuclei has been demonstrated (Rinvik, '68; Wold and Brodal, '741, i t may be t h a t t h e C1 is also involved in a complex somatomo-
CLAUSTRO-CORTICAL CONNECTIONS IN CAT AND BABOON
tor circuit consisting of: C1- FEF -, sensory trigeminal nuclei -, VPM C1. As most authors agree t h a t t h e n. interpolaris does not project to t h e VPM (Darian-Smith et al., '63; Woda et al., '771, t h e existence of such a circuit remains to be proved for this part of the trigeminal complex. This suggestion is put forward simply as a working hypothesis.
-.
ACKNOWLEDGEMENTS
We wish to thank G. Ghilini, A. Christolomme, R. Guillon, B. Guibert and M. Gavioli for invaluable technical assistance. This study has been partly supported by grant of the D.G.R.S.T. 74.7.71.376. LITERATURE CITED Astruc, J. 1971 Corticofugal connections of area 8 (frontal eye field) in Macaca rnulatta. Brain Res., 33: 241-256. Berke, J. 1960 The claustrum. the external capsule and the extreme capsule of Macaca mulatta. J. Comp. Neur., 125; 297-331. Black, P., and R. E. Myers 1962 Connection of the occipital lobe. Anat. Rec., 242: 216. Bonvallet, M., P. Dell and A. Hugelin 1952 Projections olfactives, gustatives, viscerales, vagales, visuelles et auditives au niveau des formations grises anterieures du chat. J. Physiol. (Paris), 44: 222-224. Brodmann, K. 1905 Beitrage zur histologischen Lokalisation der Grosshirnde. J. Psychol. Neurol. (Leipzig), 4: 177-226. Buser, P., and M. Imbert 1975 Neurophysiologie fonctionnelle. Hermann, Paris, 465 pp. Carman, J . B., W. M. Cowan and T. P. S. Powell 1964 The cortical projection upon the claustrum. J. Neurol. Neurosurg. Psychiat., 27: 46-51. Chadzypanagiotis, D., and 0. Narkiewicz 1971 Connections of the visual cortex with the claustrum. Acta Neurobiol. Exp. (Warsaw),32: 291-311. Darian-Smith, I., G. Phillips and R. D. Ryan 1963 Functional organization in the trigeminal main sensory and rostral spinal nuclei of the cat. J. Physiol. (London), 268: 129-146. De Vito, J. L., and 0. A. Smith 1964 Subcortical projections of the prefrontal lobe of the monkey. J . Comp. Neur., 223: 413-424. Druga, R. 1966a The claustrum of the cat (Fells d o rnestical. Folia Morphol. (Praha), 14: 7-16. 1966b Cortico-claustral connections. I. Frontoclaustral connections. Folia Morphol. ( P r a h a ) , 24: 391-399. 1968 Cortico-claustral connections. 11. Connections from the parietal, temporal and occipital cortex to the claustrum. Folia Morphol. (Praha), 26: 142-149. 1971 Projection of prepyriform cortex into claustrum. Folia Morphol. (Praha), 29: 405-410. 1972 Efferent projections from the claustrum (an experimental study using Nauta's method). Folia Morphol. (Praha), 20: 163-165. Filimonoff, I. N. 1966 The claustrum, i t s origin and development. J . Hirnforsch., 8: 503-528. Graham, R. C., and M. J. Karnovsky 1966 The early stages of absorption of injected horseradish peroxidase in the proximal tubule of mouse kidney: ultrastructural cytochemistry by a new technique. J. Histochem. Cytochem., 14: 291-302.
443
Gray, E. G. 1959 Axo-somatic and axo-dendritic synapses of the cerebral cortex: an electron microscope study. J. Anat. (London),93: 420-433. Hassmannova, J., and I . Spector 1967 Relations fonctionnelles entre relais acoustique (GM) e t le Claustrum. J . Physiol. (Paris).59: 427-428. Horcholle, G., M. Imbert and S. Tyc-Dumont 1968 Decharges phasiques et nystagmiques du nerf oculo-moteur externe evoquees par la stimulation du cortex orbitaire chez le chat. J . Physiol. (Paris),60: 462-463. Imbert, M., and P. Buisseret 1974 Visually triggered oculomotor discharges with and without eye movements. Brain Res., 72: 239-243. Kievit, J., and H. G. Kuypers 1975 Subcortical afferents to the frontal lobe in the Rhesus monkey studied by means of retrograde horseradish peroxidase transport. Brain Res., 85: 261-266. Kunzle, H. 1975 Bilateral projections from precentral motor cortex to the putamen and other parts of the basal ganglia. An autoradiographic study in Macaca fascicularis. Brain Res., 88: 195-209. Kunzle, H., and K. Akert 1977 Efferent connections of cortical area 8 (frontal eye field) in Macaca fascicularis. A reinvestigation using the autoradiographic technique. J. Comp. Neur., 2 73: 147-164. Leichnetz, G. R., and J. Astruc 1975 Efferent connections of the orbitofrontal cortex in the marmoset (Saguinus oedipus). Brain Res., 84: 169-180. Mettler, F. A. 1947 Extra-cortical connectionsof the primate frontal cerebral cortex. 11. Corticofugal connections. J. Comp. Neur.. 86: 119-166. Narkiewicz, 0. 1964 Degenerations in the claustrum after regional neocortical ablations in the cat. J. Comp. Neur., 123: 335-356. 1972 Fronto-claustral interrelations in cats and dogs. Acta Neurobiol. Exp. (Warsaw),32: 141-150. Nauta, W. J . H. 1962 Neural associations of the amygdaloid complex in the monkey. Brain, 85: 505-520. 1964 Some efferent connections of the prefrontal cortex in the monkey. In: The Frontal Granular Cortex and Behavior. J . M. Warren and K. Akert, eds. McGraw-Hill, New York, pp. 397-409. Otellin, V. A. 1969 Projections of the acoustic cortex (area Ep) to the claustrum. Proc. Acad. Sci. (USSR), 287: 1198-1200. Rae, A. S. 1954 The form and structure of the human claustrum. J. Comp. Neur., 200: 15-40. Rapisarda, C., A. Azzaroni and F. Infantellina 1969 An electrophysiological analysis of the visual projections to the claustrum in unanaesthetized cats. Arch. Sci. Biol. (Bologna), 53: 130-148. Riche, D., A. Christolomme, J. Bert and R. Naquet 1968 Atlas Stereotaxique du cerveau de babouin lPapio papto). C.N.R.S., Paris, 207 pp. Rinvik, E. 1968 The corticothalamic projection from the gyrus proreus and the medial wall of the rostral hemisphere in the cat. An experimental study with silver impregnation methods. Exp. Brain Res., 5: 129-152. Schlag, J., and M. Schlag-Rey 1970 Induction of oculomotor responses by electrical stimulation of the prefrontal cortex in the cat. Brain Res., 22: 1-13. Scollo-Lavizzari, G., and K. Akert 1963 Cortical area 8 and its thalamic projections in Macaca mulatta. J. Comp. Neur., 222: 259-267. Segundo, J . P., and X. Machne 1956 Unitary responses to afferent volleys in lenticular nucleus and claustrum. J . Neurophysiol., 19: 325-339. Showers, M. J. 1958 Correlation of medial thalamic nuclear activity with cortical and subcortical neuronal arcs. J. Comp. Neur., 209: 261-315.
444
D. RICHE AND J. LANOIR
Snider, R. S., and W. T. Niemer 1961 A stereotaxic atlas of the cat brain. Univ. Chicago Press, Chicago, Illinois. Sotelo, C., and D. Riche 1974 The smooth endoplasmic reticulum and the retrograde and fast orthograde transport of horseradish peroxidase in the nigro-striato-nigral loop. Anat. Embryol., 146: 209-218. Spector, I. 1965 Projections sensorielles au niveau du claustrum. J. Physiol. (Paris), 57: 702-703. Spector, I., and J. Hassmannova 1966 Afferences sensorielles activant les cellules du claustrum. J. Physiol. (Paris), 58: 619-620. Spector, I., J . Hassmannova and D. Albe-Fessard 1970 A macrophysiological study of functional organization of the claustrum. Exp. Neurol., 29: 31-51.
1974 Sensory properties of single neurons of cat’s claustrum. Brain Res., 66: 39-65. 1975 Somatosensory properties of neurones in the cat’s claustrum. I n : The Somato-Sensory System. H. H. Kornhuber, ed. Thieme, Stuttgart, pp. 135-144. Woda, A,, J. Azerad and D. Albe-Fessard 1977 Mapping of the trigeminal peripheral field, dental pulp afferents and sensory complex of the cat. Characterization of its neurons by stimulations of thalamic projections. J. Physiol. (Paris), 73: 367-378. Wold, J. E., and A. Brodal 1974 The cortical projection of the orbital and proreate gyri to the sensory trigeminal nuclei in the cat. An experimental anatomical study. Brain Res., 65: 381-395.
ABSTRACT 1. The mammalian Claustrum (C1) is a convergent multisensory structure of unknown function, and disputed ontogenetic origin. Its cortical projections, hitherto unknown, have been studied in cat and baboon by means of the horseradish peroxidase (HRP) technique. HRP was injected into the gyrus proreus (frontal eye field) of cats, and separately into the frontal eye fields, visual areas, and motor-premotor areas of the baboon cortex. 2. Differential retrograde transport to the C1 was demonstrated, such that in the cat the ipsilateral dorsal C1 was shown to be the principal origin of claustroproreate projections. In the baboon, the whole C1 projects onto area 8, while only the posteroventral part of the nucleus sends efferents to the visual cortex. The projection to the motor and premotor areas is present, but does not seem to be "essential." 3. Discussion of the physiological literature, together with anatomical evidence of reciprocal cortico-claustral projections to closely similar regions of the C1 lead to the suggestion that the C1 is concerned with the integration of messages subserving visually-directed movements. Some other functional implications are also discussed. The mammalian Claustrum (C1) is a nucleus with well-defined limits and not inconsiderable volume, especially in the cat. I t is situated laterally in the cerebral hemisphere, just below the neocortical mantle, limited medially by the external capsule and ventrolaterally by the extreme capsule. The cat C1 is inserted between the putamen and the ventral wall of the anterior sylvian gyrus and is in the form of a drop narrowed and elongated in the caudal direction. Two parts (Narkiewicz, '64; Druga, '66a; Chadzypanagiotis and Narkiewicz, '71) may therefore be recognized: the dorsal or insular C1, and the ventral or prepyriform C1. In the monkey and baboon, the C1 takes the form of a narrow sheet separating the putamen from the insular cortex; the anteroventral part, less well-defined than the rest, is considered as being simply adjacent to the amygdala (Rae, '54; Berke, '60). The ontogenetic origin of the C1 has been a matter of dispute for a century. Various theories, none of which J. COMP. NEUR. (1978)177: 435-444
can be regarded as proven, have been advanced: individualization from the deep layers of the insular cortex, a purely pallial origin, a mixed pallial and striatal origin (see historical review in Filimonoff, '66). From the hodological point of view, there is a striking dichotomy between the well-studied afferent connections and the almost totally unknown efferent projections. Cortical afferents in particular have been extensively investigated, in the rabbit (Carman et al., '64), the cat (Narkiewicz, '64, '72; Druga, '66b, '68, '71; Otellin, '69; Chadzypanagiotis and Narkiewicz, '71), and the monkey (Mettler, '47; Showers, '58; Berke, '60; Black and Myers, '62; Nauta, '62, '64; De Vito and Smith, '64; Astruc, '71; Kunzle, '75; Leichnetz and Astruc, '75; Kunzle and Akert, '77). As far as efferents are concerned, there is known to be a massive projection to the putamen in the cat, with sparing of the pallidum (Druga, '72). Data on neocortical projections in the cat
435
436
D. RICHE AND J. LANOIR
are limited to the retrograde degenerative changes observed by Narkiewicz ('64) after large neocortical ablations. A single observation also in the macaque by Kievit and Kuypers ('75) who noted that scattered C1 cells are sometimes labelled following injection of HRP into various areas of the rostral pole of the ipsilateral hemisphere. The present work offers new findings on t h e neocortical projection of the CI in the cat and baboon.
TABLE 1
Site ofHRPtnjection and survival time Case
c1 c2
c3
MATERIALS A N D METHODS
P1
Experiments were carried out on three cats and eight adolescent baboons (Papio papiol weighing between 3 and 7 kg. Unilateral intracortical injections of 30% horseradish peroxidase (HRP), (Sigma type VI) were made under Nembutal anaesthesia. In the cats (C1, C2, C3), a single injection was made stereotaxically (A 29, L 1) into t h e gyrus proreus with a glass micropipette of tip diameter approximately 30 pm. The micropipette was connected with the needle of a Hamilton syringe by a polyethylene tubing of the same diameter. The whole system was then filled up with coloured paraffin oil. A micrometer gage was used to push HRP into the cortex. I n t h e baboons (P1 to P8), a Hamilton syringe with attached needle was used to make freehand injections a t several points within a given cortical area. This technique made i t possible to fill almost the whole of the cortical tissue within a functional territory, and a t t h e same time to avoid subjacent structures. The injected cortical territories, in different animals, were: the frontal eye field, corresponding to the area 8 of Brodmann ('051, (P1, P2, P3), the visual areas (P4, P5) and the motor and premotor cortex (P6, P7, P8). Other details of t h e injections are given in table 1. At the end of t h e survival time, the animals were sacrificed under deep anaesthesia by intracardiac perfusion of 5-7 1 of a fixative consisting of 1%glutaraldehyde, 1%paraformaldehyde buffered to pH 7.2 with 0.12 M phosphate. The brains were removed and kept for 16 hours a t 4°C in a phosphate buffer solution containing 30%sucrose. They were then cut i n serial frozen section a t 40 p m , transversely in the case of all the cat brains and some of t h e baboon brains, while others of the latter were cut in parasagittal section. All sections from cat brains and one in five from baboon brains were treated histochemically by t h e method of Graham and Karnovsky ('66). Alternate sec-
P2 P3 P4
P5 P6 P7 P8
Areas
Quantity
injected
4
Survival time hours
Frontal eye field (+4J Head of Caudate n . Frontal eye field Frontal eye field ( + white m a t t e r ) 8 ( = frontal eye field) 8 8 + 9 ( + white matter) 17 ( + 18,191 17 ( + 18,19) 4 6+4 6+4
1
24
0.5 0.2
17 24
1
24
4 8
24 24
2.5
24
3.5
72
2.4 3 2.7
24 24 71
tions were lightly counterstained with cresyl violet. The labelled cells were mapped by using a n X-Y plotter coupled to the stage of a microscope equipped with a darkfield condenser. For the cat brains, frontal planes were defined a s in t h e atlas of Snider and Niemer ('61). The extent of the injection sites in cases C2 and C3 a r e illustrated in figure 1A on outlines of frontal sections. The injection site in C1 (not shown) was superimposable on that of C2, but in addition involved the dorsomedial edge of the head of t h e caudate nucleus. All 8 cortical injection sites in t h e baboons have been shown on a n outline of the left hemisphere in lateral view (fig. 1B). Frontal and parasagittal planes were defined with reference to the atlas of Riche e t al. ('68). RESULTS
1. Cat Injection of HRP into the gyrus proreus causes t h e labelling of a large number of cells in the dorsal part of the ipsilateral C1. These cells are found from the rostral pole of the nucleus back as far as plane A 11.5, but their distribution is not uniform. Almost all the peroxidase-positive cells are concentrated in the zone of constriction between the dorsal and ventral parts of t h e nucleus (fig. 2). This region of high density corresponds exactly with t h e "medial marginal zone" as defined by Druga ('66a). In case C3, with a small injection volume, there are only ten labelled cells,
CLAUSTRO-CORTICAL CONNECTIONS IN CAT AND BABOON
437
A bbreuratcons
aaa, area amydaloidea anterior ap, area prepyriformis as, arcuate sulcus CI, claustrum cp, cortex pyriformis cs, central sulcus FEF, frontal eye field gea, gyrus ectosylvianus anterior gem, gyrus ectosylvianus medialis gsa, gyrus sylvianus anterior
gsp, gyrus sylvianus posterior
GP, globus pallidus nam, nucleus amygdalae anterior nC, nucleus caudatus Pu, putarnen sp, sulcus pseudosylvius sr, sulcus rhinalis ss, sulcus Sylvius TO, tractus opticus VPM, ventroposterornedial nucleus of the thalamus
Fig. 1A Sites of HRP injection into the rostral pole of the left hemisphere in cats C2 andC3. The centre of the injection site is shown in black, while its periphery, colouredyellow in the sections, is dotted. In case C2, the track of the micropipette (A 29, L 1) is shown in section n"16. B Cortical injection sites in the baboon shown on a lateral projection of the left hemisphere.
but they are all in this restricted zone between frontal planes A 13 and A 12. In every case, the prepyriform part of C1 shows, on darkfield examination, a few labelled cells further posteriorly, between planes A 12.5 and A 11. The entire caudal pole of C1, posterior to plane A 10.5, was entirely devoid of labelled cells. The anterior sylvian gyrus (insular cortex), which is adjacent to the C1, shows a high density of labelled cells (fig. 2: A 15.5-A 13). In the ventral wall of this gyrus, and in the depths of the pseudosylvian sulcus, the labelled somata are located in the deep cell layers. In the contralateral hemisphere, two or three peroxidase-positive cells per section were observed in the rostral half of the C1, anterior to planes A 15-A 14.5, but with the same
preferential localization in the transitional zone between dorsal and ventral parts as on the ipsilateral side. Similarly, and in numbers roughly equal to those in the C1, labelled neurones could also be observed in the ventral wall of the anterior sylvian gyrus and the depths of the pseudosylvian sulcus. From the cytological point of view, at least two types of neurone accumulate HRP in the soma and dendritic tree. The majority are stellate cells of 20-25 p m diameter (figs. 3a,b). The other category, amounting to about 20% of the total, consists of large pyramidal neurones with a diameter of 20 p m x 35-40 pm (fig. 3c). There is possibly a third type of labelled cell, constituting about one tenth of the population and made up of small fusiform cells.
438
D. RICHE AND J. LANOIR
Fig. 2 Cartography of labelled cells in the claustrum and adjacent structures of cat C2. Borders and points have been traced on the X-Y plotter, each point representing the site of a labelled cell seen by darkfield examination. Each outline shows the total number of labelled cells in six successive sections. Frontal planes defined according to the atlas of Snider and Niemer ('61)
2. Baboon
Injections of HRP into frontal area 8 (fig. 1B: P1, P2, P3) result in the appearance of labelled cells in the whole of the ipsilateral C1. Case P3, which received t h e largest quantity of HRP (8 pl, the injected area including the underlying white matter), exhibited t h e highest density of labelled cells. They are most numerous in the anterodorsal region of t h e structure, which is the part nearest the site of injection. This preferential labelling can be clearly seen on the outlines of parasagittal sections from animal P3 (fig. 4A). In cases P 2 and P3, a very small number of labelled cells were also observed in the contralateral C1. The visual areas were injected in two baboons (fig. 1B: P4, P5). Although principally filling area 17, in both cases t h e HRP also
penetrated into areas 18 and 19. The distribution of labelled cells in t h e ipsilateral C1 differs from that following frontal injection, brightfield examination showing t h a t labelled neurones are concentrated in the posteroventral corner of t h e nucleus in parasagittal sections (fig. 4B). I n neither of these animals were any peroxidase-positive cells seen in the C1 contralateral to the site of injection. In three further animals, one part of the motor area (P6) and t h e motor plus a small part of the premotor area (P7, P8) were injected respectively. In the case of injection restricted to a part of area 4, no labelled somata were observed in the ipsilateral C1. When a greater part of area 4 and a part of area 6 were involved, peroxidase-positive neurones were present in t h e ipsilateral C1. Case P7 has shown only very few labelled cells. In
CLAUSTRO-CORTICAL CONNECTIONS IN CAT AND BABOON
439
Fig. 3 Darkfield photomicrographs showing retrograde labelling of ipsilateral claustro-cortical cells in the marginal zone of the CI. Cat C1, frontal plane A 14. a, b Coexistence of multipolar and pyramidal neurones. c A large pyramidal neurone with its apical dendrite.
Fig. 4 Cartography of labelled cells in parasagittal sections of the injected hemisphere in the baboon. A Following HRP injection into area 8 (baboon P3), labelled cells can be seen throughout the claustrum. B Following HRP injection into the visual areas (baboon P5), almost all the labelled cells are concentrated in the ventral part of the claustrum a t posterior levels. Each outline shows the total number of labelled cells in four successive sections. Parasagittal planes defined according to the atlas of Riche e t al. (‘68).
case P8, labelling was evident but seemed, however, to spare t h e ventral region of the structure. Following both frontal and occipital HRP injection, different neurone types within t h e C1 showed peroxidase labelling. Fusiform,
ovoid, or polygonal cells were filled, and were classified according to their diameter as “small” (10-20 p m ; fig. 5 above) and “large” (20-30 p m ; fig. 5 below). The distribution of these cell types within t h e C1 appears to be diffuse, with no obvious organization. However,
440
D. RICHE AND J. LANOIR
Fig. 5 Brightfield photomicrographs showing various types of ipsilateral claustro-cortical neurones in the baboon. a Two specimens of small (above) and large (below) neurones. b Different cellular types may be observed close to one another, a phenomenon which occurred throughout the entire structure and for each site of injection.
peroxidase-positive large ovoid cells were frequently observed on edge of the nucleus. DISCUSSION
(1) The results of our cat experiments demonstrate the existence of a direct projection from the C1 to gyrus proreus. Even when t h e volume of tissue injected was very small (C31, a t least two types of relatively large C1 neurones - mainly multipolar, and to a lesser extent pyramidal - were labelled. In Nissl sections of the C1, these two cell types have been observed mingled with smaller round or fusiform cells (Chadzypanagiotis and Narkiewicz, ’71). The zone of the C1 in which labelled cells were found in our experiments is, according to Druga (’66a), t h e region with t h e highest cell density. Furthermore, of t h e five cell types identified by this author, t h e large multipolar and pyramidal cells are precisely those which
predominate in this “medial marginal zone.” The rare small fusiform cells showing peroxidase activity have also been described in normal material by the above-named authors. While the projection of the C1 on t h e gyrus proreus may be a new finding, i t was 11 years ago t h a t Druga (’66b) demonstrated connections i n the opposite direction as a result of making localized lesions in the frontal cortex of the cat. His series included two cases (K31, K51) in which the lesion corresponded to the site of HRP injection i n two of our animals ((22, C3). The region of the C1 in which he found terminal axon degeneration is the same as t h a t in which labelled neurones were demonstrated in our material. However, while terminal degeneration following a lesion of gyrus proreus was confined to t h e rostra1 quarter of t h e insular C1, in our material labelled cells were absent only from t h e caudal pole of the
CLAUSTRO-CORTICAL CONNECTIONS IN CAT AND BABOON
structure. I t may be concluded that the claustral territory projecting to gyrus proreus includes the regions which receive reciprocal afferents from gyrus proreus, but it also extends further caudally. In the same publication, Druga (’66b) maintained that degenerated terminals surround claustral cells of different types. I t is of interest to speculate a s to whether the various neuronal categories contacted by the terminals of cortical efferents may be the same as those showing peroxidase accumulation. In such a case, the cortico-claustral contacts (chiefly axodendritic) might be in the neuropile of those claustral neurones which project reciprocally on the cortical area of origin. Such a hypothesis is also compatible with the electron microscopic observations of Narkiewicz (’721, who showed that the presynaptic endings in the C1 which degenerate following ablation of the proreate and sigmoid gyri are axodendritic and make asymmetric synapses. They were principally seen on fine dendritic arborisations and spines, more rarely on large dendrites. The author concludes that “These findings seem to prove that frontoclaustral connections do exist and that they are mostly if not exclusively axodendritic of a n asymmetrical type which according to Gray’s hypothesis may suggest that they are excitatory (Gray, ’591.”As in the specific sensory nuclei of the thalamus, there are many topically-organized closed circuits connecting certain zones of the C1 with their respective cortical projection area. The last important point is that our HRPinjections into the gyrus proreus cover in whole, or in part, the territory recognized as being the frontal oculomotor area (Schlag and Schlag-Rey, ’70; Brodal and Brodal, ’71). The existence of a massive monosynaptic projection of the insular cortex onto the gyrus proreus doubtless explains why some authors (Horcholle e t al., ’68; Imbert and Buisseret, ’74) situate the feline frontal eye field in the insulo-orbital cortex.’ (2) In the baboon, retrograde axonal transport of HRP has enabled us to demonstrate that an efferent claustral system projects to different neocortical areas. Following an injection which filled the whole of area 8, labelled cells were found throughout the C1, so that it may be concluded t h a t the entire C1 projects to area 8. The existence of a projection in the opposite direction, from frontal cortex to the C1, was suggested by Berke (’60).
44 1
However, his interpretation of results following various frontal and temporal ablations failed to make clear whether the degenerating axons were terminal, or represented fibres of passage. Scollo-Lavizzari and Akert (’63) did not include the C1 in their analysis of the subcortical projections from area 8, but later Nauta (’64) thought that there was “little doubt that the prefrontal cortex entertains direct efferent connections with the Claustrum” while Astruc (’71) was affirmative: “there is a substantial projection from area 8 to the Claustrum.” Using the autoradiographic tracing technique, Kunzle and Akert (’77) confirmed recently the existence of a direct projection from area 8 in Macaca fascicularis: two foci of moderate grain density were seen in the ipsilateral C1, one in the central moiety, the other in the most ventral and caudal part. In case P3, the high density of peroxidase-positive neurones in the anterodorsal region of the C1 must be related to the fact that the injected HRP also penetrated into area 9. This interpretation is confirmed by recent experiments in which this area was selectively injected. Conversely, Nauta (’64) has described cortico-claustral fibres originating in area 9. Results obtained after injections into motor and premotor areas lead us to think that - in comparison with little animals where the injected quantity of HRP must be very small (less than 0.5 pl) to accurately localize the origin of the projection (Sotelo and Riche, ’74) in the baboon it was necessary to inject a larger amount of HRP (several pl) to find substantial labelling. In the motor field of Papio, when a small part, restricted to the face area, was injected, no labelled cells could be observed (P6). For an injection site situated more dorsally and involving partially area 6 (P8), a significant labelling appeared. Concerning the reciprocal descending component, Nauta (’64) failed to observe fibre degeneration in C1 after lesions in area 6 and area 4 of the precentral gyrus. Using the autoradiographic technique with injections in the face area of the precentral motor cortex in the macaque, Kunzle (’75) found grain patches in the most rostrodorsal part of the C1. In good agreement with the localization of our HRP-positive neurones, he found also that the projec-
’
The simultaneous presence of labelled neurones in both the CI and the insular cortex following frontal HRP injection certainly dws not prove that the former derives from the latter, a s some authors believe (Filimonoff, ‘66).The finding nevertheless speaks in favor of a common relationship.
442
D. RICHE AND J. LANOIR
tions of the arm-trunk area seemed to cover approximately three-fifths of t h e dorsal C1. I n sum, reciprocal connections exist very likely in monkey and baboon between the C1 and t h e cortex of both the frontal pole (areas 8 + 9) and the precentral gyrus. The ascending component described here seems therefore to form part of a closed circuit. Occipital injections demonstrate a clear posteroventral localization of positive somata within t h e C1. There is thus a preferential relationship linking t h e posterior poles of both the C1 and the neocortex. Comparison of t h e distribution of labelled cells following frontal precentral and occipital injections respectively reveals a rostrocaudal organization of corticipetal efferents. Narkiewicz ('64) found topographically-organized retrograde degenerative changes in t h e C1 of the cat after extensive removals of some neocortical regions. Ablations of t h e sensorimotor cortex produced degenerations in the dorsorostral portion, removals of the visual region caused severe alterations in t h e dorsocaudal portion of the C1. For the author, "the simplest and most likely i n t e r p r e t a t i o n is t h a t t h e degenerations reveal t h e existence of direct claustro-cortical, or, possibly, cortico-claustral connections." The same antero-posterior organization is also described for projections in t h e opposite, cortico-claustral direction (Carman e t al., '64; Druga, '66b, '68; Narkiewicz, '72; Kunzle, '75). According to some authors, t h e rostro-caudal organization may also be associated with a medio-lateral one. So far as we know, there are no publications on connections between primary visual areas and the C1 in primates. I t is therefore necessary to discuss our results in relation to data on the cat. Chadzypanagiotis and Narkiewicz ('71), confirming t h e results of Druga ('68),described a projection of t h e visual cortex onto t h e dorsocaudal part of the C1. Furthermore, in explanation of precocious cellular (therefore retrograde) degeneration following cortical ablation, these authors suggested a reciprocal claustro-cortical component linking t h e same regions. Our results in the baboon confirm this hypothesis. I t may be postulated that t h e posteroventral part of the C1 is connected by a closed circuit with t h e ipsilateral visual areas. (3)The diversity of neuronal types in t h e primate C1 is not in itself a new finding (Rae, '54). We have yet been able to show t h a t several neurone types project to the frontal cortex, and t h a t apparently these same types send
also axons to the occipital cortex. Despite the rostro-caudal organization, neurones projecting both frontally and occipitally co-exist in t h e posteroventral part of the nucleus. For any given cell type, it remains to be seen whether independent somata project to each of the two cortical areas, or whether a single cell may project to both, in opposite directions, by means of a bifurcated axon. (4) When examined by evoked potential and single-unit methods, t h e C1 appears to be a multisensory nucleus, showing little spontaneous activity, on which converge somatic, auditory, visual, olfactory, gustatory, vestibular and viscero-vagal afferents (Bonvallet et al., '52; Segundo and Machne, '56; Spector, '65; Spector and Hassmannova, '66; Hassmannova and Spector, '67). The somaesthetic modality is predominant, activating almost all excitable neurones with short latency. In agreement with anatomical findings, cells of unimodal (lemniscal) type a r e preferentially localized in t h e rostra1 half of the C1, between planes A 17.5 and A 15 in the cat (Spector et al., '70, '74, '75).Units exhibiting heteromodal somatosensory or polysensory convergence are more numerous posteriorly, where visual evoked potentials have the greatest amplitude and shortest latency (Rapisarda e t al., '69). I t appears from the above discussion t h a t the C1 is connected through short loops to two cortical areas (FEF, primary visual cortex) involved in visual mechanisms. Because of the low frequency of spontaneous discharge, the transmission would appear to occur phasically, under special conditions, perhaps in conjunction with body, head or eye movements. This new feature has to be kept in mind when searching for the functional r61e of this nonhomogenous multisensory structure. As pointed out by Spector e t a1 ('751, t h e C1 is also connected with different thalamic nuclei and many other subcortical structures. I t seems concerned w i t h "long loops" (Buser a n d Imbert, '75) of different modalities; each loop relays through its specific thalamic nucleus and corresponding cortical area before relaying back to t h e C1. Even longer "visual" circuits might also involve t h e superior colliculus, the thalamic visual associative nuclei, and t h e posterior group (PO). Since a projection of both t h e proreate and orbital gyri onto t h e principal sensory (and interpolaris) trigeminal nuclei has been demonstrated (Rinvik, '68; Wold and Brodal, '741, i t may be t h a t t h e C1 is also involved in a complex somatomo-
CLAUSTRO-CORTICAL CONNECTIONS IN CAT AND BABOON
tor circuit consisting of: C1- FEF -, sensory trigeminal nuclei -, VPM C1. As most authors agree t h a t t h e n. interpolaris does not project to t h e VPM (Darian-Smith et al., '63; Woda et al., '771, t h e existence of such a circuit remains to be proved for this part of the trigeminal complex. This suggestion is put forward simply as a working hypothesis.
-.
ACKNOWLEDGEMENTS
We wish to thank G. Ghilini, A. Christolomme, R. Guillon, B. Guibert and M. Gavioli for invaluable technical assistance. This study has been partly supported by grant of the D.G.R.S.T. 74.7.71.376. LITERATURE CITED Astruc, J. 1971 Corticofugal connections of area 8 (frontal eye field) in Macaca rnulatta. Brain Res., 33: 241-256. Berke, J. 1960 The claustrum. the external capsule and the extreme capsule of Macaca mulatta. J. Comp. Neur., 125; 297-331. Black, P., and R. E. Myers 1962 Connection of the occipital lobe. Anat. Rec., 242: 216. Bonvallet, M., P. Dell and A. Hugelin 1952 Projections olfactives, gustatives, viscerales, vagales, visuelles et auditives au niveau des formations grises anterieures du chat. J. Physiol. (Paris), 44: 222-224. Brodmann, K. 1905 Beitrage zur histologischen Lokalisation der Grosshirnde. J. Psychol. Neurol. (Leipzig), 4: 177-226. Buser, P., and M. Imbert 1975 Neurophysiologie fonctionnelle. Hermann, Paris, 465 pp. Carman, J . B., W. M. Cowan and T. P. S. Powell 1964 The cortical projection upon the claustrum. J. Neurol. Neurosurg. Psychiat., 27: 46-51. Chadzypanagiotis, D., and 0. Narkiewicz 1971 Connections of the visual cortex with the claustrum. Acta Neurobiol. Exp. (Warsaw),32: 291-311. Darian-Smith, I., G. Phillips and R. D. Ryan 1963 Functional organization in the trigeminal main sensory and rostral spinal nuclei of the cat. J. Physiol. (London), 268: 129-146. De Vito, J. L., and 0. A. Smith 1964 Subcortical projections of the prefrontal lobe of the monkey. J . Comp. Neur., 223: 413-424. Druga, R. 1966a The claustrum of the cat (Fells d o rnestical. Folia Morphol. (Praha), 14: 7-16. 1966b Cortico-claustral connections. I. Frontoclaustral connections. Folia Morphol. ( P r a h a ) , 24: 391-399. 1968 Cortico-claustral connections. 11. Connections from the parietal, temporal and occipital cortex to the claustrum. Folia Morphol. (Praha), 26: 142-149. 1971 Projection of prepyriform cortex into claustrum. Folia Morphol. (Praha), 29: 405-410. 1972 Efferent projections from the claustrum (an experimental study using Nauta's method). Folia Morphol. (Praha), 20: 163-165. Filimonoff, I. N. 1966 The claustrum, i t s origin and development. J . Hirnforsch., 8: 503-528. Graham, R. C., and M. J. Karnovsky 1966 The early stages of absorption of injected horseradish peroxidase in the proximal tubule of mouse kidney: ultrastructural cytochemistry by a new technique. J. Histochem. Cytochem., 14: 291-302.
443
Gray, E. G. 1959 Axo-somatic and axo-dendritic synapses of the cerebral cortex: an electron microscope study. J. Anat. (London),93: 420-433. Hassmannova, J., and I . Spector 1967 Relations fonctionnelles entre relais acoustique (GM) e t le Claustrum. J . Physiol. (Paris).59: 427-428. Horcholle, G., M. Imbert and S. Tyc-Dumont 1968 Decharges phasiques et nystagmiques du nerf oculo-moteur externe evoquees par la stimulation du cortex orbitaire chez le chat. J . Physiol. (Paris),60: 462-463. Imbert, M., and P. Buisseret 1974 Visually triggered oculomotor discharges with and without eye movements. Brain Res., 72: 239-243. Kievit, J., and H. G. Kuypers 1975 Subcortical afferents to the frontal lobe in the Rhesus monkey studied by means of retrograde horseradish peroxidase transport. Brain Res., 85: 261-266. Kunzle, H. 1975 Bilateral projections from precentral motor cortex to the putamen and other parts of the basal ganglia. An autoradiographic study in Macaca fascicularis. Brain Res., 88: 195-209. Kunzle, H., and K. Akert 1977 Efferent connections of cortical area 8 (frontal eye field) in Macaca fascicularis. A reinvestigation using the autoradiographic technique. J. Comp. Neur., 2 73: 147-164. Leichnetz, G. R., and J. Astruc 1975 Efferent connections of the orbitofrontal cortex in the marmoset (Saguinus oedipus). Brain Res., 84: 169-180. Mettler, F. A. 1947 Extra-cortical connectionsof the primate frontal cerebral cortex. 11. Corticofugal connections. J. Comp. Neur.. 86: 119-166. Narkiewicz, 0. 1964 Degenerations in the claustrum after regional neocortical ablations in the cat. J. Comp. Neur., 123: 335-356. 1972 Fronto-claustral interrelations in cats and dogs. Acta Neurobiol. Exp. (Warsaw),32: 141-150. Nauta, W. J . H. 1962 Neural associations of the amygdaloid complex in the monkey. Brain, 85: 505-520. 1964 Some efferent connections of the prefrontal cortex in the monkey. In: The Frontal Granular Cortex and Behavior. J . M. Warren and K. Akert, eds. McGraw-Hill, New York, pp. 397-409. Otellin, V. A. 1969 Projections of the acoustic cortex (area Ep) to the claustrum. Proc. Acad. Sci. (USSR), 287: 1198-1200. Rae, A. S. 1954 The form and structure of the human claustrum. J. Comp. Neur., 200: 15-40. Rapisarda, C., A. Azzaroni and F. Infantellina 1969 An electrophysiological analysis of the visual projections to the claustrum in unanaesthetized cats. Arch. Sci. Biol. (Bologna), 53: 130-148. Riche, D., A. Christolomme, J. Bert and R. Naquet 1968 Atlas Stereotaxique du cerveau de babouin lPapio papto). C.N.R.S., Paris, 207 pp. Rinvik, E. 1968 The corticothalamic projection from the gyrus proreus and the medial wall of the rostral hemisphere in the cat. An experimental study with silver impregnation methods. Exp. Brain Res., 5: 129-152. Schlag, J., and M. Schlag-Rey 1970 Induction of oculomotor responses by electrical stimulation of the prefrontal cortex in the cat. Brain Res., 22: 1-13. Scollo-Lavizzari, G., and K. Akert 1963 Cortical area 8 and its thalamic projections in Macaca mulatta. J. Comp. Neur., 222: 259-267. Segundo, J . P., and X. Machne 1956 Unitary responses to afferent volleys in lenticular nucleus and claustrum. J . Neurophysiol., 19: 325-339. Showers, M. J. 1958 Correlation of medial thalamic nuclear activity with cortical and subcortical neuronal arcs. J. Comp. Neur., 209: 261-315.
444
D. RICHE AND J. LANOIR
Snider, R. S., and W. T. Niemer 1961 A stereotaxic atlas of the cat brain. Univ. Chicago Press, Chicago, Illinois. Sotelo, C., and D. Riche 1974 The smooth endoplasmic reticulum and the retrograde and fast orthograde transport of horseradish peroxidase in the nigro-striato-nigral loop. Anat. Embryol., 146: 209-218. Spector, I. 1965 Projections sensorielles au niveau du claustrum. J. Physiol. (Paris), 57: 702-703. Spector, I., and J. Hassmannova 1966 Afferences sensorielles activant les cellules du claustrum. J. Physiol. (Paris), 58: 619-620. Spector, I., J . Hassmannova and D. Albe-Fessard 1970 A macrophysiological study of functional organization of the claustrum. Exp. Neurol., 29: 31-51.
1974 Sensory properties of single neurons of cat’s claustrum. Brain Res., 66: 39-65. 1975 Somatosensory properties of neurones in the cat’s claustrum. I n : The Somato-Sensory System. H. H. Kornhuber, ed. Thieme, Stuttgart, pp. 135-144. Woda, A,, J. Azerad and D. Albe-Fessard 1977 Mapping of the trigeminal peripheral field, dental pulp afferents and sensory complex of the cat. Characterization of its neurons by stimulations of thalamic projections. J. Physiol. (Paris), 73: 367-378. Wold, J. E., and A. Brodal 1974 The cortical projection of the orbital and proreate gyri to the sensory trigeminal nuclei in the cat. An experimental anatomical study. Brain Res., 65: 381-395.
