ACTA NEUROBIOL. EXP. 1989, 49: 141-151

CONTRALATERAL CONNECTIONS OF THE DOG'S FRONTAL ASSOCIATIOK CORTEX

Graiyna RAJKOWSKA and Anna KOSMAL

Department of Neurophysiology, Nencki Institute of Experimental Biology 3. Pasteur St., 02-093 Warsav~, Poland

Keu words: frontal association cortex, contralateral connections, dog, HRP method

Abstract. We studied the topography of contralateral connections of both prefrontal and premotor regions of the dog's frontal association cor- tex (FAC) by charting distributions of retrogradely labeled cells following unilateral MRP injections to various areas of this cortex. Generally, in the contralateral hemisphere the labeled cells were most numerous in the FAC areas localized homotopically to the injection sites, less numerous in FAC areas heterotopic to injections, and the least numerous in cortical areas situated outside the I'rontal lobe. The nonfrontal areas which project to the dorsal azd ventral FAC dlffer from one another. Dorso-caudal parts of the cingular and insular areas, as well as the auditory, somatosensory and visual association cortices project to the dorsal FAC, while the ventro-rostra1 parts of the cingular and insular areas, together with the prepiriform and periamygdaloid areas of the olfactory cortex as well as the subcallosal area send their axons to the ventral FAC. Thus, the dorsal and ventral FAC areas are supplied by contralatera! afferents originating from different cortical areas. Similar organization of ipsilateral FAC connections was described previously.

INTRODUCTION

There are few papers on the organization of contralateral connections of the frontal lobe cortex in such species as monkey (5, 7, 16, 18), cat (3, 4, 11, 15, 19) and rat (1, 2, 6, 13), while in the dog this projection

has never been investigated. A general opinion is that the most numer- ous transcallosal connections run between prefrontal areas situated ho- motopically. It has been reported that a weaker contralateral projection to this cortex arises from heterotopic areas of the frontal lobe as well as from the cingular, insular and sensory association cortices (3, 5, 7, 15). A heterotopical transcallosal projection from the retrosplenial area to the prefrontal cortex has been reported in the cat (3) and rat (1). Ad- ditionally in the rat, callosal afferents to the prefrontal cortex have been found in the perirhinal, entorhinal and presubicular areas (1, 2, 13). Thus, the contralateral projection originates in the cortical areas which are the source of strong ipsilateral connections as well (3, 7, 13). All those cal- losal axons reach the opposite hemisphere via respective parts of the corpus callosum (16, 19). The homotopical afferents pass through the rostral part of the corpus callosum, while the heterotopical ones run through both its rostral and caudal parts.

In the present paper the topography of contralateral connections of the frontal association cortex in the dog is analyzed. A comparison is also made with the previously described organization of ipsilateral con- nections in the dog, as well as with the pattern of contralateral frontal connections in other mammalian species.

MATERIAL AND METHOD

In 23 young dogs unilateral injections of horseradish peroxidase (HRP) were made to various areas of the frontal association cortex under Nembutal anesthesia (35 mg/kg of body weight). In each subject the cortex was injected by means of a Hamilton syringe with 30-50°/o HRP solution 9n 8-12 neighbouring points at the depth of 2 mm from the cortical surface. The total volume of injected HRP in a single animal was about 1.5 p1. After 48 h the animals were deeply anesthetized and perfused transcardially with NaCl (0.9°/o), followed by an aldehyde mixture in 0.1 phosphate buffer at ph = 7.4. The brains were removed and stored for 48 h at 4OC in the phosphate buffered sucrose (30°/o, ph = 7.4) and then frozen they were cut coronally in sections 40 pm thick. Every 10th section was collected and processed with tetramethyl- benzidine as the chromagen (14) to reveal the HRP-positive neurons. Another series of sections was counterstained with thionine.

RESULTS

Contralateral cortico-cortical connections of the frontal association cortex (FAC, see the extent of this cortex in Fig. 1) were examined on the basis of 23 cases of frontal injections. All cases of FAC injections

Fig. 1. Extent of the frontal association cortex (FAC) in the dog's brain. A, lateral surface of the hemisphere; B, medial surface; C, coronal sections from rostra1 (1) to caudal (4) direction. FAC involves two cortical regions (according to 17) pre- motor (squared area) and prefrontal (vertical lines area) and is divided into many small areas (according to 8, 9) delineated by broken lines. For the used names,

see the list of abbreviations.

were arranged in two groups, dorsal and ventral, according to localiz- ation of injections within the frontal lobe and a pattern of distribution of retrogradely labeled neurons. The dorsal group involves 17 cases of injections into different areas of the premotor and dorsal prefrontal regions (Fig. 2). The ventral group involves 6 cases of injections into areas of the ventral prefrontal region (Fig. 3). All injections caused the retrograde labeling of neurons in the cortex of the opposite hemisphere in areas homotopic and heterotopic to the insection sites.

The results of the dorsal group are illustrated with a chosen repre- sentative, case D6 (Fig. 4). In this subject the injection was large and involved the cortex of the dorsal prefrontal region of both aspects of the hemisphere, dorso-lateral (PR, PRL, PORd areas) and dorso-medial (part of the PR area). A zone of diffusion of HRP covered some further parts of the above mentioned areas, as well as XM prefrontal area and CA premotor area (Fig. 4A and B). Following the D6 injection the greatest number of labeled neurons was observed in the contralateral

Fig. 2. Localization of FAC injections included in the dorsal group; black areas indicate sites of the highest concentration of injected enzyme; broken lines show the zone of enzyme diffusion. Note tha.t injections Dl-D4 are situated in the premotor region, while D5-Dl7 - in the dorsal prefrontal region. Among the latter ones, injections D5-D9 cover prefrontal areas situated most dorsally on one or both (lateral, a) and (medial, b) aspects of the hemisphere, D10-Dl3 involve central arezs on the lateral aspect, while D14-Dl7 - c e ~ t r a l areas on the medial aspect

of the hemisphere. For the used names, see the list of abbreviations.

frontal lobe cortex in areas localized homotopically to the injection site (areas PR, PRL, PORd, XM in Fig. 4C and D). Less numerous cells were found in heterotopic areas situated around the previously mentioned ones, but also in the dorsal prefrontal and premotor areas (XP, PGd, CA, XC, ORBd in Fig. 4C and D). On the contrary, in the ventral pre- frontal areas only single HRP-positive neurons were found (areas, PGv, SG in Fig. 40).

Fig. 3. Localization of FAC injections included in the ventral group. Injections D18, Dl9 are localized most ventrally on both aspects of the frontal lobe. I n cases D20, D21 the injections cover ventro-c;;udal areas of the lateral aspect of the frontal lobe, while in cases D22 and D23 - ventral areas of the medial aspect.

Denotations as in Fig. 2.

Also in other cases of this group the largest accumulation of labeled cells in FAC was found exactly in areas homotopic to the injection sites, while more dispersed labeling was observed in areas situated heterotopic- ally to the injections, but wa.s restricted to the dorsal half of the frontal lobe cortex (Figs. 5 and 7). Significantly fewer HRP-positive neurons were observed in the ventral half of the contralateral frontal lobe. In all cases with dorsal injections heterotopic projections originating from cortical areas localized outside the FAC are far less numerous. That is why individual charts of these labeled neurons were pulled together in Fig. 7 (see localization of dotted areas and black circles). The labeled cell bodies were found in the cingular cortex (in its dorso-central (CN) and caudal e.g. retrosplenial (RSPL) parts); in the insular cortex (INS) of the caudal orbital gyrus and in the depth of the dorso-caudal sylvian sulcus (sS); in the auditory association cortex (AAC) of the dorsal sylvian gyrus and the depth of the ectosylvian sulcus (sEs); in area 7 of the dorsal edge of suprasylvian sulcus (sSs); in area 19 of the dorsal edge of the splenial sulcus (sSpl); in area 20 of the posterior suprasylvian gyrus, as well as in the perirhinal (PRH) cortex (Fig. 7C and D, black circles).

Fig. 4. Distribution of retrogradely labeled neurons in the contralateral FAC after dorsal FAC injection D6. A, lateral aspect of the frontal lobe; B, medial aspect of the frontal lobe; C, lateral aspect of the frontal lobe of the contralateral hemisphere; D, medial aspect of the frontal lobe of the contralateral hemisphere. Black areas indicate site of injection; broken lines show-zone of enzyme diffusion; black dots symbolize localization of contralaterally labeled neurons on the convex- ity of gyri; open circles, localization of labeled neurons in the depth of sulci. For

the used names, see the list of abbreviations.

The results of the ventral injections are illustrated by a represent- ative case Dl9 (Fig. 6). In this subject the injection was restricted to the anterior parts of the subproreal gyrus (SPR) situated ventrally on both aspects of the hemisphere, as well as to the ventrolateral areas SPRL, PORv, and part of the frontal pole cortex (POL) situated ventro-medial- ly (Fig. 6A and B). The zone of enzyme diffusion covered more posterior parts of the above mentioned areas and other ventral prefrontal areas, ORBv and PGv (Fig. 6A and B). In the case Dl9 the largest accumulation of labeled neurons was observed in the contralateral frontal lobe cortex

Fig. 5. Microphotograph of HRP labeled neurons in the cortex of the contralateral dorsal FAC, homotopic to the injection site. Case Dl5 with injection into dorsomedial FAC area - XM.

Fig. 6. Distribution of labeled neurons in the contralateral FAC after ventral FAC injection i n D19. Denotations as i n Fig. 4.

homotopically to the injection site (areas SPR, SPRL, POL, ORBv, PGv in Fig. 6C and D). Less numerous cells were found in ventral prefrontal areas, but situated heterotopically to the injection site, whereas only single neurons were observed in dorsal prefrontal areas (ORBd, PRL, PR, XM in Fig. 6C and D).

In all other cases of ventral group injections (Fig. 3), a general pat- tern of distribution of contralaterally labeled cells was similar to the one described above. It is worth emphasizing that neurons of this pro- jection were predominantly found in the ventral half of the frontal cortex. Moreover, in cases of ventral FAC injections, a smali number of labeled cells was found in non-frontal cortical areas, which is shown in

FAC CONTRALATERAL CONNECTIONS

Latera v i e w

Media l v i e w

Fig. 7. General scheme of the organization of contralateral homo - and hetero- topic projections to the FAC (results collected from all cases). Dotted areas in A and R symbolize dorsal FAC injections, while checkered areas - ventral FAC injections. Black circles in C and D represent areas which are the source of contra- lateral afferents to the dorsal FAC, while open squares indicate sources of afferents

to the ventral FAC. For the used names see the list of abbreviations.

a general scheme (Fig. 7, open squares). Labeled neurons were observed in the cingular cortex (predominantly in its ventro-rostra1 part - G); in the insular cortex (INS) of the caudal orbital gyrus and in the depth of the anterior sylvian sulcus (sS), as well as in the prepiriform (Ppir) and periamygdaloid (Pamg) areas of the olfactory cortex and in the sub- callosal (SC) area (Fig. 7, open squares).

It can be concluded that dorsal and ventral FAC areas are supplied by contralateral afferents originating from differently localized cortical areas in dorsal and ventral half of the hemisphere, respectively (Fig. 7, compare distribution of black circles versus open squares).

The neurons of contralateral projections to the FAC were predom- inantly found in cortical layers IIIb, V, IIIa, and sporadically in layers I1 and VI (Fig. 5), which is in accordance with a general pattern of laminar organization of cortical projections in the dog (8).

DISCUSSION

The obtained results provide a general rule of the distribution of callosal connections to the frontal cortex. The most numerous connect- ions link homotopic areas of the dog's frontal association cortex of both hemispheres, while more scarce connections run between areas situated heterotopically within the frontal cortices. Additionally, sporadic callosal afferents to the FAC arrive from many other neocortical, but also meso- cortical and allocortical regions situated outside the frontal lobe. This general rule and the sources of contralateral connections to the pre- frontal cortex were earlier demonstrated in investigations on other mam- malian species (1-7, 11, 13, 15, 16, 19). In the present study we described the topography of callosal FAC connections in the dog brain.

In addition to a general rule of strong connections between homotopic areas of both hemispheres and weaker connections between heterotopic areas, we found in the dog brain a differentiation in callosal connections reaching the dorsal and ventral FAC areas. Dorsal areas of both pre- frontal and premotor regions are predominantly connected with the contralateral homo- and heterotopic areas of the dorsal FAC and have weak connections with nonfrontal cortex of the dorsal half of the hemi- sphere, Nonfrontal afferents come from the cingular and insular cortices (their dorso-caudal parts), association cortex of the auditory, somatosens- ory and visual areas and from the perirhinal cortex. On the contrary, ventral FAC areas are supplied by numerous afferents from homo- and heterotopic areas of the ventral contralateral FAC and also by scarce nonfrontal heterotopical connections from the ventral half of the hemi-

sphere, namely from the ventro-rostral cingular and insular cortices, as well as from the subcallosal area and the pkiform cortex.

All the above mentioned areas giving rise to connections to the con- tralateral FAC have been also reported to be the sources of ipsilateral afferents to the FAC (12, 17). The organization of both ipsilateral and contralateral projections of the dog FAC seems to be similar. The dorsal FAC areas are connected ipsi- and contralaterally with neocortical and mesocortical areas involved in the processing of polysensory information, while the ventral FAC areas are supplied by afferents from allo- and mesocortical areas presumably related to the emotional behavior. Ipsi- lateral connections are however much more numerous than the contra- lateral ones.

Finally, we should like to emphasize that in the dog FAC the extent of dorsal and ventral zones, determined on the basis of differentiation in patterns of short and distal ipsilateral connections (12, 17), is in accord- ance with the topography of their contralateral connections.

This investigation was supported by Project CPBP 0401 of the Polish Academy of Sciences.

ABBREVIATIONS

A AC C A CJ CN CX I fG fPg fPs F AC G HRP INS MI MI1 ORB ORBd ORBv Pamg PG PGd PGv PM POL

auditory association cortex area composita anterior area compoeita internal gyrus cinguli. pars dorsalis area composita precruciata fissura fissura genualis fissura pregenualis fissura presylvia frontal association cortex area genualis horseradish peroxidase cortex insularis primary motor cortex secondary motor cortex gyrus orbitalis gyrus orbitalis pars dorsalis gyrus orbitalis pars ventralis cortex piriformis pars periamygdalaidea area pregenualis area pregenualis pars dorsalis area pregenualis pars ventralis premotor cortex area polaris

2 - Acta Neurobiol. Exp. 4/89

PORd PORv Ppir PR PRH PRL RSPL S

sCr sEs sG spg sRha sRhp s s ssp1 sss SC SG SPR SPRL XC XL m XP

area paraorbitalis dorsalis area paraorbitalis ventralis cortex piriformis pars prepiriformis gyrus proreus cortex perirhinalis area prorea lateralis area retrosplenialis sulcus sulcus cruciatus sulcus ectosylvius sulcus genualis sulcus pregenualis sulcus rhinalis anterior sulcus rhinalis posterior sulcus sylvius sulcus splenialis sulcus suprasylvius area subcallosa area subgenualis gyrus subproreus gyrus subproreus lateralis area precruciata centralis area precruciata lateralis area precruciata medialis area precruciata posterior

REFERENCES

1. AUDINAT, E., CONDE, F. and CREPEL, F. 1988. Cortico-cortical connections of the limbic cortex of the rat. Exp. Brain Res. 69: 439-443.

2. BECKSTEAD, R. M. 1979. An autoradiographic examination of the cortico- cortical and subcortical projections of the medio-dorsal-projection (pre- frontal) cortex in the rat. J. Comp. Neurol. 184: 43-62.

3. CAVADA, C. and REINOSO-SUAREZ, F. 1981. Interhemispheric cortico-cortical connections to the prefrontal cortex in the cat. Neurosci. Lett. 24: 211-214.

4. CAVADA, C. and REINOSO-SUAREZ, F. 1985. Topographical organization of the cortical afferent connections of the prefrontal cortex in the cat. J. Comp. Neurol. 242: 293-324.

5. GOLDMAN-RAKIC, P. S. and SCHWARTZ, M. L. 1982. Interdigitation of con- tralateral and ipsilateral columnar projections to frontal association cortex in primates. Science 216: 755-756.

6. JACOBSON, S. 1963. IntraIaminar, interIaminar, callosal and thalamocorticaI connections in frontal and parietal areas of the albino rat cerebral cortex. J. Comp. Neurol. 124: 131-146.

7. JACOBSON, S. and TROJANOWSKI, J. Q. 1977. Prefrontal granular cortex of the rhesus monkey. 11. Interhemispheric cortical afferents. Brain Res. 132: 235-246.

8. KOSMAL, A., STEPNIEWSKA, I. and MARKOW, G. 1983. Laminar organiza- tion of efferent connections of the prefrontal cortex in the dog. Acta Neu- robiol. Exp. 43: 115-127.

9. KREINER, J. 1961. Myeloarchitectonics of the frontal cortex in dog. J. Comp. Neurol. 116: 401-414.

10. KREINER, J. 1966. Reconstruction of neocortical lesions within the dog's brain: Instructions. Acta Biol. Exp. 26: 221-243.

11. LUTTENBERG, J. 1974. Heterotropic contralateral projection of the neocortical spheres of the cat brain. I. Frontal cortex. A. Interhemispheric association of the frontal spheres. Acta Univ. Carol. Med. 20: 225-249.

12. MARKOW-RAJKOWSKA, G. and KOSMAL, A. 1987. Organization of cortical afferents to the frontal association cortex in dogs. Acta Neurobiol. Exp. 47: 137-161.

13. MARKOWITSCH, H. J. and GULDIN, W. 0. 1983. Heterotopic interhemispheric cortical connections in the rat. Brain Res. Bull. 10: 805-810.

14. MESULAM, M. M. 1978. Tetramethylbenzidine for horseradish peroxidase neu- rochemistry: a non-carcinogenic blue reaction product with superior sen- sitivity for visualizing neuronal afferents and efferents. J. Histochem. Cytochem. 26: 106-117.

15. MIZUNO, N., CLEMENTE, D. C. and SAURLAND, E. K. 1969. Projections from the orbital gyrus in the cat. J. Comp. Neurol. 136: 130-136.

16. PANDYA, D. N. and VIGNOLO, L. A. 1971. Intra- and interhemispheric pro- jections of the precentral, premotor, arcuate area in the rhesus monkey. Brain Res. 26: 217-233.

17. RAJKOWSKA, G. and KOSMAL, A. 1988. Intrinsic connections and cytoarchi- tectonic data of the frontal association cortex in the dog. Acta Neurobiol. EXP. 48: 169-192.

18. SCHWARTZ, M. L. and GOLDMAN-RAKIC, P. S. 1982. Single cortical neuro- nes have axon collaterals to ipsilateral and contralateral cortex in fetal and adult primates. Nature 299: 154-155.

19. VONEIDA, T. J. and TREVARTHEN, C. R. 1069. An experimental study of transcallosal connections between the proreus gyri of the cat. Brain Res. 12: 384-395.

Accepted 2 February 1989