diverse thalamic projections to the prefrontal cortex in the rhesus monkey

THE JOURNAL OF COMPARATIVE NEUROLOGY 313~65-94 (1991)

Diverse Thalamic Projections to the Prefrontal Cortex in the Rhesus Monkey

H. BARBAS, T.H. HASWELL HENION, AND C.R. DERMON Department of Health Sciences, Boston University, and Department of Anatomy and Neurobiology, Boston University School of Medicine, Boston, Massachusetts 02215

ABSTRACT We studied the sources of thalamic projections to prefrontal areas of nine rhesus monkeys

with the aid of retrograde tracers (horseradish peroxidase or fluorescent dyes). Our goal was to determine the proportion of labeled neurons contributing to this projection system by the mediodorsal (MD) nucleus compared to those distributed in other thalamic nuclei, and to investigate the relationship of thalamic projections to specific architectonic areas of the prefrontal cortex. We selected areas for study within both the basoventral (areas 11, 12, and ventral 46) and the mediodorsal (areas 32, 14, 46, and 8) prefrontal sectors. This choice was based on our previous studies, which indicate differences in cortical projections to these two distinct architectonic sectors (Barbas, '88; Barbas and Pandya, '89). In addition, for each sector we included areas with different architectonic profiles, which is also relevant to the connectional patterns of the prefrontal cortices.

The results showed that MD included a clear majority (over 80%) of all thalamic neurons directed to some prefrontal cortices (areas 11,46, and 8); it contributed just over half to some others (areas 12 and 321, and less than a third to area 14. Clusters of neurons directed to basoventral and mediodorsal prefrontal areas were largely segregated within MD: the former were found ventrally, the latter dorsally. However, the most striking findings establish a relationship between thalamic origin and laminar definition of the prefrontal target areas. Most thalamic neurons directed to lateral prefrontal cortices, which are characterized by a high degree of laminar definition (areas 46 and 81, originated in the parvicellular and multiform subdivisions of MD, and only a few were found in other nuclei. In contrast, orbital and medial cortices, which have a low degree of laminar differentiation, were targeted by the magnocellular subdivision of MD and by numerous other limbic thalamic nuclei, including the midline and the anterior. Thus topographic specificity in the origin of thalamic projections increased as the laminar definition of the target area increased. Moreover, the rostrocaudal distribution of labeled neurons in MD and the medial pulvinar also differed depending on the degree of the laminar definition of the prefrontal target areas. The rostral parts of MD and the medial pulvinar projected to the eulaminate lateral prefrontal cortices, whereas their caudal parts projected to orbital and medial limbic cortices. Selective destruction of caudal MD is known to disrupt mnemonic processes in both humans and monkeys, suggesting that this thalamic- limbic prefrontal loop may constitute an important pathway for memory.

Key words: thalamus, prefrontal architecture, orbital cortex, limbic system, mediodorsal nucleus, memory, macaque monkeys, medial cortex, cortical lamination, thalamic evolution

Classic studies have suggested that the mediodorsal (MD) thalamic nucleus and the prefrontal cortex are closely associated (see Le Gros Clark, '32; Nauta, '71; Reep, '84 for reviews).This view was based on observations that the elaboration of MD is associated with an expansion of the frontal cortex to which it projects (Le Gros Clark, '32; Walker, '36, '38; Von Bonin and Green, '49; Pribram et al., '53; Nauta, '62; Akert, '64; Locke, '69; Tobias, '75; Tanaka, '76). In fact, Rose and Woolsey ('48) suggested that the

prefrontal cortex should be defined as the region receiving projections from MD. With the advent of new and more sensitive tracing procedures, however, it became apparent that neurons in thalamic nuclei other than MD project to prefrontal areas as well (Carmel, '70; Jones and Leavitt,

Accepted July 24, 1991. Address reprint requests to H. Barbas, Boston University, 635 Common-

wealth Ave. #431, Boston, MA 02215.

o 1991 WILEY-LISS. INC.

66 H. BARBAS ET AL.

'74; Trojanowski and Jacobson, '74; Bos and Benevento, '75; Kievit and Kuypers, '77; Potter and Nauta, '79; Barbas and Mesulam, '81; Asanuma et al., '85; Baleydier and Mauguiere, '85; Goldman-Rakic and Porrino, '85; Ilinsky et al., '85, '87). The relative density of neurons in MD and in other nuclei directed to prefrontal cortices is, however, not known.

In this study we used a simple quantitative method to address the issue of diversity in the origin of thalamic projections to prefrontal cortices. This was accomplished by noting the relative proportion of retrogradely labeled neurons in MD compared to those distributed in other thalamic nuclei after injecting retrograde tracers in specific prefrontal areas. In addition, we investigated some hitherto unre- solved aspects of the topography of thalamic projections. We noted particularly the distribution of prefrontally directed neurons along the rostrocaudal extent of MD and the medial pulvinar, both of which are large nuclei. This information is important because mnemonic deficits have been associated with preferential destruction of the caudal parts of MD in both humans and monkeys (Victor et al., '71; Isseroff et al., '82; Zola-Morgan and Squire, '85).

Finally, the present study was conducted in the context of our previous findings, which have shown differences in the cortical and some subcortical projections to specific prefrontal areas (Barbas, '88; Barbas and De Olmos, '90). Those studies established that the degree of laminar definition of individual areas (Barbas and Pandya, '89) is relevant to the pattern of their connections. Therefore, we sought to determine whether the thalamic projections to the same prefrontal cortices varied along these lines. The present study provides evidence consistent with this view.

METHODS Experiments were conducted on nine rhesus monkeys

(Macaca mulatta). They were anesthetized with ketamine hydrochloride (10 mg/kg, i.m.) followed by sodium pentobar- bital administered intravenously through a femoral cathe- ter until a surgical level of anesthesia was achieved. Addi- tional anesthetic was administered during surgery as needed. Surgery was performed under aseptic conditions. The monkey's head was firmly positioned in a holder, which left the cranium unobstructed for surgical approach. A bone defect was made, the dura was retracted and the cortex was exposed.

Horseradish peroxidase (HRP) experiments In eight animals, injections of a solution containing 8%

HRP conjugated to wheat germ agglutinin (Sigma) were made with a microsyringe (Hamilton, 5 p1) mounted on a microdrive. The needle was lowered to the desired site under microscopic guidance. Small amounts (0.05 ~ 1 ) of the injectate were delivered 1.5 mm below the pial surface at each of two adjacent sites separated by 1-2 mm over a 30-minute period.

HRP was injected in orbital areas 11 and 12, lateral area 12, and ventral area 46 (cases 1-4). These areas belong to the basoventral prefrontal sector defined on the basis of its architecture and intrinsic connections (Barbas and Pandya, '89). In addition, injections were placed in areas within the mediodorsal prefrontal sector: medial area 32 (case 5), dorsal area 46 (case 71, and dorsal area 8 (cases 8 and 9).

Following a 40-48-hour survival period, the monkeys were re-anesthetized and perfused through the heart with

A AN AD AM AV Cau cc Cdc c g Cif Cim c1 Clc CM c s Csl GM GMmc GMpc H IL LD LF LGN Li LO LP MD MDdc MDmc MDmf, mf MDpc ML MO MTT

arcuate sulcus anterior nuclei anterior dorsal nucleus anterior medial nucleus anterior ventral nucleus caudate corpus callosum central densocellular nucleus cingulate sulcus central inferior nucleus central intermediate nucleus central lateral nucleus central latocellular nucleus centromedian nucleus central superior nucleus central superior lateral nucleus medial geniculate nucleus medial geniculate nucleus, magnocellular sector medial geniculate nucleus, parvicellular sector habenula intralaminar nuclei lateral dorsal nucleus lateral fissure dorsal lateral geniculate nucleus limitans nucleus lateral orbital sulcus lateral posterior nucleus mediodorsal nucleus mediodorsal nucleus, densocellular sector mediodorsal nucleus, magnocellular sector mediodorsal nucleus, multiform sector mediodorsal nucleus, parvicellular sector midline nuclei medial orbital sulcus mamillothalamic tract

Abbreuiations

OLF P Pa Pac PAll Pcn Pf Pi PI Pm Po Pro ProM Pt R Re Ro Sf SG SM,Sm ST THI VA VAmc VL VLC VLm VLO

VPI VPLC VPLO VPM WMpc x

VLPS

olfactory cortex principal sulcus paraventricular nucleus paraventricular nucleus, caudal sector limbic periallocortex paracentral nucleus parafascicular nucleus pulvinar nucleus, inferior sector pulvinar nucleus, lateral sector pulvinar nucleus, medial sector pulvinar nucleus, oral sector proisocortex rostral portion of the ventral premotor cortex parataenial nucleus reticular nucleus reuniens nucleus rostral sulcus subfascicular nucleus suprageniculate nucleus stria medullaris superior temporal sulcus habenulo-interpeduncular tract ventral anterior nucleus ventral anterior nucleus, magnocellular sector ventral lateral nuclei ventral lateral caudal nucleus ventral lateral medial nucleus ventral lateral oral nucleus ventral lateral postrema nucleus ventral posterior inferior nucleus ventral posterior lateral caudal nucleus ventral posterior lateral oral nucleus ventral posterior medial nucleus ventral posterior medial nucleus, parvicellular sector area X

THALAMIC PROJECTIONS TO THE PREFRONTAL CORTEX 67

-("- o/ *. 25

- - - -._ ~ _ _ _ _ _ _ ~ - ---/ - - 2 4 *.A-

5 mm 9 C

Fig. 1. Composite of injection sites shown on the medial (A), lateral (B), and basal (C) surfaces of the cerebral hemisphere. The injection sites are superimposed on an architectonic map of the prefrontal cortex (Barbas and Pandya, '89). Mediodorsal areas appear above and basoventral below the heavy dashed line (B). Within the mediodorsal sector, stepwise increases in laminar definition are observed in a direction from the medial periallocortex (PAll) to proisocortical areas 24, 25, and 32, and then through medial areas 14, 10, and 9, to dorsal areas 9, 10, and rostral 46, and finally through caudal areas 46 and 8. Within the basoventral sector, gradual increases in the number of layers and their delineation are observed from the orbital PAll to the proisocortex (Pro), orbital area 25, area 13, then to orbital areas 12, 14, and 11, and then through ventrolateral areas 10, 12, rostral 46, and finally to caudal areas 46 and 8. Small dashed lines demarcate architectonic areas; large dashed lines demarcate sulci. Large numbers designate architectonic areas; small numbers refer to cases. Injection sites in mediodorsal prefrontal cortices are shown by the striped area; in basoventral sites they are shown in black.

saline followed by 2 liters of fixative (1.25% glutaraldehyde, 1% paraformaldehyde in 0.1 M phosphate buffer at pH 7.4). The fixative was followed by perfusion with 2 liters of cold (4°C) phosphate buffer (0.1 M, pH 7.4).

The brain was then removed from the skull, photographed, placed in glycerol phosphate buffer (10% glycerol and 2% DMSO in 0.1 M phosphate buffer at pH 7.4) for 1 day and in 20% glycerol phosphate buffer for another 2 days. The brain was frozen in -75°C isopentane, transferred to a freezing microtome, and cut in the coronal plane at 40 pm in 10 series. One series of sections was treated to visualize HRP (Mesulam et al., '80). The tissue was mounted, dried, and counterstained with neutral red. Adja- cent series of sections were stained for Nissl bodies, myelin, acetylcholinesterase, or cytochrome oxidase to aid in delineating architectonic borders (Geneser-Jensen and Blacks- tad, '71; Gallyas, '79; Wong-Riley, '79).

Fluorescent tracer experiment With the surgical procedures outlined above, one animal

received an injection of the fluorescent tracer fast blue (3%, 0.4 pl) in medial area 14 (case 6). After a survival period of 10 days, the animal was deeply anesthetized and perfused with 6% paraformaldehyde in 0.1 M cacodylate buffer (pH 7.4). The brain then was placed in a solution of 6% paraformaldehyde with glycerol and 2% DMSO for 3 days, frozen, and cut (as described above). Sections were mounted on subbed slides.

Data analysis Brain sections were viewed microscopically under bright-

and darkfield illumination for HRP cases, or fluorescence illumination for case 6. Outlines of brain sections, the location of retrogradely labeled neurons ipsilateral to the injection site, and the site of blood vessels used as landmarks were transferred from the slides onto paper by using an X-Y recorder (Hewlett Packard, 7044B), which was electronically coupled to the stage of the microscope. In later experiments data were transferred from the slides onto paper by means of a digital plotter (Hewlett Packard, 7475A) coupled to the stage of the microscope (as described above) and to a computer (Compaq 386). In this system the analog signals are converted to digital signals via an analog-to-digital converter (Data Translation). Software developed in this laboratory allowed each labeled neuron to be recorded only once. Both procedures allow accurate topographic presentation of labeled neurons within the thalamus.

All sections in one series that contained labeled neurons in the thalamus were charted. In initial experiments labeled neurons (represented as dots on the charted sections) were counted by using a grid inserted in the microscope eyepiece. When the data were stored in a computer, labeled neurons were counted by outlining the area of interest (e.g., one nucleus) by moving the X and Y axes of the stage of the microscope. The number of labeled neurons within the enclosed area was calculated by means of an algorithm written for this purpose. A designated region was then subdivided automatically along its dorsoventral or mediolateral extent into five sectors and the number of labeled neurons per sector recorded. In initial experiments this operation was conducted manually. Labeled neurons in each dorsoventral, mediolateral, or rostrocaudal sector of a given nucleus were counted from serial sections as de-

68 H. BARBAS ET AL.

Fig. 2. Left: Brightfield photomicrographs o f coronal sections from rostra1 (A) through caudal (E) sectors of the thalamus showing the location of the various nuclei and their cytoarchitecture. Right: Brightfield photomicrographs of matched sections stained for myelin are shown in B, D, and F. (Frozen cut tissue, cresyl violet stain (A,C,E) and Gallyas myelin stain (B,D,F). Scale bar = 1 mm.


scribed previously (Barbas and De Olmos, '90). This procedure allowed a more detailed topographic analysis.

The size of the injection site and histochemical variables inadvertently differ from case to case. To minimize the above extrinsic factors, we expressed the number of labeled neurons in serial sections through each nucleus as a percentage of the total number of labeled neurons in the entire thalamus in that case. This procedure makes it possible to demonstrate the projection profile of each case and some qualitative differences in the pattern of regional labeling among cases.

Anatomic definition of injection sites The cortical areas containing the injection sites were

reconstructed serially using the sulci as landmarks and are shown on diagrams of the surface of the cortex. The latter were drawn from photographs of each brain showing the external morphology of the experimental hemispheres.

Seven of the nine cases of this study were used previously to study corticocortical connections (Barbas, '88). Cases 1-5 in the present study were denoted by the same number in the previous study; cases 8 and 9 (area 8) of the present study correspond, respectively, to cases 7 and 8 of the previous study. The architectonic boundaries of the above seven cases have been determined from matched sections stained for Nissl, myelin, or acetylcholinesterase (AChE). They are illustrated in diagrams of coronal sections through the injection site in the previous study (Barbas, '88; see Figs. 2-6,8-9). Figure 10F shows the injection site in area 8 in case 8; the boundaries of area 8 are shown in a matched section stained with thionin in Figure 10G. In case 6, not hitherto described, the extent of the injection site of fast blue was plotted from coronal sections, which were then counterstained with cresyl violet and returned to the microscope to determine architectonic borders (see Fig. 81,J). The injedion was within area 14 (Fig. 8A,H). Case 7 (dorsal area 46) of this study was also used previously to study corticocortical connections (Barbas and Mesulam, '85; see Fig. 4).

References to architectonic areas of the prefrontal cortex are according to a classification described previously (Bar- bas and Pandya, '89). On the basal and lateral surfaces, the injections were in architectonic areas designated by the same numbers by Walker ('40b); on the medial surface our map corresponds more closely to that of Brodmann ('05).

Thalamic borders Architectonic borders of thalamic nuclei were determined

from series of matched sections stained with thionin, myelin, AChE, or cytochrome oxidase. This analysis is based on differences in cytoarchitecture and in the pattern of distribution of the histochemical markers among thalamic nuclei (see Fig. 2). The borders of nuclei were placed first on drawings or photographs of these matched series of sections, and then onto photographs and the charted drawings with the labeled neurons by using blood vessels as landmarks. Photographs were obtained by placing glass slides with mounted tissue on a photographic enlarger and projecting a negative image directly onto photographic paper (Kodak, panchromatic).

RESULTS Architecture of thalamic nuclei

The histochemical and cytoarchitectonic procedures used proved to be differentially useful in delineating the borders

of the various nuclei. For example, matched sections treated to visualize AChE facilitated the placement of borders within the subdivisions of MD. The magnocellular subdivision (MDmc) contains a light AChE reaction product and numerous AChE positive blood vessels (Olivier et al., '69; see Fig. 22A). The parvicellular subdivision (MDpc) has irregular AChE patches and is overall denser in reaction product than MDmc (see Fig. 22A). The multiform subdivision (MDmfl has a distinctive architecture consisting of a mixture of small cells that appear pale in the Nissl preparation with some interspersed large neurons. Cytochrome oxidase stains the large neurons in MDmf darkly and sets the entire nuclear subdivision apart from the adjacent MDpc, where neurons appear light and indistinct. Densocel- lular MD is situated posteriorly within the MD complex and borders the caudal extent of the central lateral nucleus. The boundaries of this MD subdivision are difficult to delineate as noted by Olszewski ('52). Moreover, there is controversy as to whether the densocellular area is in fact part of MD or should be considered as the caudal extension of the central lateral nucleus as suggested by Jones ('85). The Nissl preparation seems to be the most useful in delineating the borders of this region.

Matched sections treated to show AChE activity aided in delineating the subdivisions of the pulvinar as well. The oral pulvinar has a lower AChE content than its medial subdivision (Olivier et al., '69). Its lateral and inferior subdivisions are intersected by fiber tracts that distinguish them from the medial pulvinar (see Fig. 2E-F).

The ventral anterior (VA) nucleus has a magnocellular subdivision found medially and a parvicellular subdivision located laterally (Olszewski, '52; Ilinksy et al., '87; see Fig. 2A). In the AChE preparation the VA nucleus has a reticulated appearance composed of AChE positive plexuses enclosing zones with a very low AChE content. This pattern is more prominent in the magnocellular than in the parvicellular VA (see Fig. 24A).

The intralaminar nuclei can be delineated primarily on the basis of their AChE activity and cytoarchitecture. For example, the paracentral nucleus is characterized by a dense AChE network (see Figs. 22A, 24A). The more caudally situated central lateral nucleus can be distinguished from the paracentral by its cytoarchitecture (see Fig. 2C) and lighter AChE content. The parafascicular and centromedian nuclei have distinct cytoarchitectonic features (see Fig. 2C) and appear darker in the AChE stain than the surrounding nuclei. The medially situated parafascicular nucleus is the darker of the two and is characterized by the presence of AChE-positive neurons (see also Jones and Leavitt, '74; Jones, '85; Ilinksy and Kultas-Ilinksy, '87; see Fig. 2C). The central superior, central superior lateral, and parataenial are small nuclei distinguished primarily by their cytoarchitecture (Olszewski, '52) and their distribution of AChE.

The midline nuclei include the paraventricular, central latocellular, central densocellular, rotund, central intermediate, central inferior, caudal paraventricular, and reuniens, according to the terminology of Olszewski ('52). These nuclei are referred to as rhomboid, central medial, and medioventral by Jones ('85). All midline nuclei have a low myelin content, which sets them apart from the laterally situated MD nucleus (see Fig. 2B,D). Midline nuclei can be distinguished from each other using cytoarchitectonic criteria or by the distribution of AChE. For example, the paraventricular nucleus appears darker when

70 H. BARBAS ET AL.

stained for AChE than its surrounding nuclei. Reuniens lies above the third ventricle (see Fig. 2A,B). It can be distinguished from the neighboring midline nuclei by its some- what lighter AChE content (see Fig. 24A).

The three anterior nuclei have different cytoarchitectonic features (Olszewski, '52) and differ in their distribution of AChE (Olivier et al., '69). The anterior dorsal (AD) has a high AChE content and AChE positive blood vessels. The anterior medial (AM) has the lowest content of AChE, and the anterior ventral (AV) an intermediate amount (see Fig. 24A).

The terminology for the thalamus used in this study is according to the map of Olszewski ('52), with references to modifications described by Jones ('85).

The injections were in four basoventral (area 11, case 1; orbital 12, case 2; lateral 12, case 3; ventral area 46, case 41, and five mediodorsal (area 32, case 5; area 14, case 6; dorsal 46, case 7; dorsal area 8, cases 8-9) prefrontal areas. Eight of the cases appeared in previous studies on cortical projections as described above (see Anatomic definition of injection site). The relationship of the injection sites to architectonic areas of the prefrontal cortex is shown in Figure 1.

Injection sites.

Thalamic projections The distribution of labeled neurons in the thalamus after

HRP or fluorescent dye injections is shown in diagrams of cross sections in Figures 3-11 and in Table 1. Labeled neurons were noted in MD, medial pulvinar (Pm), intralaminar (IL), and ventral anterior (VA) nuclei in all cases studied. A few scattered labeled neurons were noted also in the ventral lateral (VL) nuclei in most cases. In addition, there were labeled neurons in midline (ML) and anterior (AN) nuclei, but these showed a preferential projection to some prefrontal sites. The distribution of labeled neurons within the various thalamic nuclei differed from case to case, as described below.

The MD nucleus included the majority of all labeled neurons from the thalamus directed to most prefrontal areas (Table 1). In five cases (cases 1, 4, 7-9, with HRP in areas 11, 46, and 8), labeled neurons in MD constituted a clear majority ( > 80%), in three (cases 2, 3, and 5, with HRP in areas 12 and 321, they accounted for approximately half, and in one (case 6, area 141, they made up only 29% of all labeled neurons.

MD projections: Their origin among its subdivisions. The distribution of labeled neurons in the subdivisions of MD differed markedly from case to case (Table 2). MDpc, which projected to all prefrontal areas, included the majority of labeled neurons directed to areas 12, 46, and 8 (Figs. 4-6, 9-11; see also Figs. 15C-F, 17C-F) and made up a substantial proportion of the labeled neurons projecting to area 11 (24.8%) and area 32 (16.8%). However, MDpc included only a small portion of the labeled neurons directed to area 14 (4.3%).

In contrast, MDmc and MDmf projected preferentially to some, though not the same, areas. MDmc included half of all labeled neurons directed to area 11 and substantial proportions (20.4-29.3%) of those projecting to medial areas 32 and 14 (Figs. 3, 7, 8; see also Figs. 15A,B, 17A,B). MDmc provided only a few of the labeled neurons projecting to area 12 and very few, if any, of those directed to areas 46 or 8. In contrast, MDmf included about a third of all labeled neurons directed to area 8 (cases 8,9; Figs. 10, 11; see also

MD projections: Overall differences.

TABLE 1. Distribution of Labeled Neurons in the Thalamus Following Iniection of Retrograde Tracers in Prefrontal Cortices

~~ ~

Thalamic projection zone'

Injection Total Case site MD Pm VA VL IL ML AN N

Basoventral 1 Area 11 80.1 3.7 1.9 0.1 3.3 4.1 7.1 3,337 2 012 59.6 16.6 6.8 1.1 3.7 5.4 7.0 2,359 3 L12 61.5 17.8 12.7 0.4 3.1 4.2 0.7 5,189 4 V46 82.2 10.7 2.6 - 3.9 0.6 - 1,494

Mediodorsal 5 Area32 53.5 7.9 5.8 0.9 17.3 13.6 1.0 6,263 6 Area14 29.0 13.1 9.4 4.1 18.7 24.7 1.0 679 7 D46 93.0 0.8 3.2 - 3.4 - - 402 8 D8 87.0 2.5 2.7 1.9 6.0 0.1 - 2,848 9 D8 87.0 3.9 1.7 0.7 7.0 - - 2,851

'Data in columns below nuclear designations are expressed in percentages. Total N column shows the total number of labeled neurons in the thalamus in each case. Abbreviations here and throughout the tables are as in list preceding Figure 1. - = no labeled neurons.

TABLE 2. Distribution of Labeled Neurons in Subsectors of the Thalamic Nucleus MD Following Injection of Retrograde Tracers

in Prefrontal Cortices

Case Injection

site

Basoventral 1 2 3 4

5 6 7 8 9

Mediodorsal

Area 11 012 L12 V46

Area 32 Area 14 D46 D8 D8

MD projection zone'

MDmc

50.5 6.6 7.6 1.8

29.3 20.4 - -

MDpc

24.8 52.4 51.0 71.6

16.8 4.3

93.0 52.4 55.5

MDmf

- 0.2 8.8

0.1 - -

34.4 29.1

MDdc

4.8 0.6 2.7 -

7.3 4.3

0.2 2.4

-

Total % -

80.1 59.6 61.5 82.2

53.5 29.0 93.0 87.0 87.0

'Data in columns below nuclear designations are expressed in percentages - _ - no labeled neurons.

Fig. 17E,F), and a smaller proportion of those projecting to ventral area 46 (case 4, 8.8%, Fig. 6C-D). MDmf included few, if any, of the labeled neurons directed to the rest of the prefrontal areas.

The densocellular subdivision of MD included a small portion (4.3-7.3%) of the labeled neurons directed to areas 11, 32 and 14 (cases 1, 5, 6; Figs. 3E, 7E, 8F); it included only a few, if any, of the neurons projecting to areas 12,46, or 8.

Be- cause MDmc and MDpc are large, we investigated the specific location of labeled neurons within each subdivision. The clearest topography was observed for projections originating in MDpc.

Rostrocaudal, In both orbital cases (cases 1 and 2) and in medial case 5 (area 321, most labeled neurons were found in the caudal and central sectors of MDpc (Figs. 12A,B, 13A), whereas those directed to lateral areas (cases 3,4 and 7-9) were found in its central and rostral sectors (Figs. 12C,D, 13B-E). A similar relationship was observed for labeled neurons within MDmc, albeit not as clearly (cases 2,3,5,6; Figs. 12A-C, 13A,B).

Labeled neurons directed to orbital and medial areas were found mostly in the medial sectors of MDpc; those directed to lateral areas 12, 46, and 8 were found progressively in its lateral sectors (Figs. 14-17). The same relationship was observed for labeled neurons in MDmc, albeit not as clearly.

MDprojections: Their origin within a subdivision.

Mediolateral.


Fig. 3. The distribution of labeled neurons in the thalamus (represented by dots) in diagrams of coronal sections in rostra1 (A) through caudal (F) thalamic levels in case 1 after HRP injection in orbital area 11 (black area in G) . The number of dots (in this and in Figs. 4-11) represents the relative density of labeled neurons.

72 H. BARBAS ET AL.

3

- 1 rnrn

Fig. 4. The distribution of labeled neurons in rostra1 (A) through caudal (F) thalamic levels in case 2 after HRP injection in orbital area 12 (G).


Fig. 5. 7 area 12 (GI.

74 H. BARBAS ET AL.

Fig. 6. The distribution of labeled neurons in rostra1 (A) through caudal (F) thalamic levels in case 4 after HRP injection in ventral area 46 (GI.

75 THALAMIC PROJECTIONS TO THE PREFRONTAL CORTEX

Fig. 7. The distribution of labeled neurons in rostral (A) through caudal (F) thalamic levels in case 5 after HRP injection in medial area 32 (GI.

76 H. BARBAS ET AL.

Fig. 8. Left: The distribution of labeled neurons in rostral (B) through caudal (G) thalamic levels in case 6 after injection of fast blue in area 14 shown in a coronal section through the injection site (A, black area; stripes show the halo of the injection site in this case). Right: Photomicrograph of coronal section caudal to that depicted in A shows the injection site of fast blue (H, white area). The boundaries of area 14,

which contains the injection site, are shown in the same section (I) after the section was stained with cresyl violet; J shows the region around the injection site at higher magnification. The boundaries of area 14 and the surrounding architectonic areas are demarcated by small vertical lines through the cortex. Arrow in H, I and J point to the same blood vessel for reference. Scale bar in H, I, and J = 1 mm.

THALAMIC PROJECTIONS TO THE PREFRONTAL CORTEX

Figure 8 continued

78 H. BARBAS ET AL.

Fig. 9. The distribution of labeled neurons in rostral (A) through caudal (F) thalamic levels in case 7 after HRP injection in dorsal area 46 (GI.


D

Fig. 10. Left: The distribution of labeled neurons in rostral (A) through caudal (D) thalamic levels in case 8 after HRP injection in dorsal area 8 (El. Right: F (next page). The injection site in area 8 is shown in a brightfield photomicrograph (dark area). G. The boundaries of area 8 are shown in a matched section stained with thionin. Arrow in F and G points to the needle mark at the center of the injection site. Scale bar in G = 1 mm applies to F.

80 H. BARBAS ET AL.

TABLE 3. Distribution of Labeled Neurons in the Intralaminar Thalamic Nuclei Following Injection of Retrograde Tracers in Prefrontal Cortices Intralaminar projection zone'

~~

Case Injection site Pcn c1 c s Csl Pt CM Pf Li SG Total %

Basoventral 1 Area 11 2 012 3 L12 4 V46

5 Area 32 6 Area 14 7 D46 8 D8 9 D8

Mediodorsal

0.5 - 0.2 0.5 1.3 0.2 0.04 0.3 1.3 0.1 1.1 0.7 - 0.1

2.2 - 1.6 1.0 3.2 - 2.9 0.3 2.7 - - 0.7 2.4 1.8 - 0.04 3.3 0.6 - 0.3

- - 0.7 0.2 0.5 0.7 - 3.3 - - 1.4 0.4 0.1 3.7 - 0.2 0.7 0.4 0.4 3.1 - - 1.3 0.2 0.5 3.9

1.9 2.6 5.1 1.7 1.2 17.3 4.9 - 2.1 5.3 - 18.7

3.4 - - 0.2 0.4 1.2 6.0 - - 0.4 1.7 0.7 7.0

- - - - -

'Data in columns below nuclear designations are expressed in percentages - = no labeled neurons.

Figure 10 continued

Dorsouentral. The distribution of labeled neurons within the dorsoventral extent of MDpc and MDmc did not vary much among the four basoventral cases, or among the five mediodorsal. The data within each group of cortices, therefore, were pooled and are shown in Figure 18. Most labeled neurons projecting to basoventral prefrontal areas were found in the ventral and central portions of MDpc and MDmc (Figs. 15, MA), whereas those directed to the mediodorsal areas were concentrated dorsally (Figs. 17, 18B).

Medial pulvinar (Pm) Labeled neurons in Pm were found in all cases but their

proportion differed from case to case (Table 1). The highest proportions were observed in cases with HRP injections in area 12 (cases 2,3, 16.6%-17.8%; Figs. 4E,F, 5E,F; see also Fig. 22E,F). Substantial proportions of labeled neurons (7.9-13.1%) were noted also in cases 4-6 (ventral area 46, area 32, and area 14; Figs. 6E,F, 7E,F, 8G). In the rest of the cases only a few labeled neurons were noted in Pm (Table 1,O.S-3.9%).

Like MD, the medial pulvinar is large. We thus analyzed the relative location of labeled neurons along its rostrocaudal, mediolateral, and dorsoventral axes.

Rostrocaudal. Most labeled neurons directed to areas 11, 12, and 32 were found in the caudal and central sectors of Pm, whereas those projecting to areas 14,46 and 8 were found progressively in its more rostra1 parts (Figs. 19,20).

Labeled neurons projecting to areas 11, 32, and 14 were found in the medial portions of Pm, whereas those directed to areas 12,46, and 8 were found in its central and lateral portions (Figs. 21,22E,F, 23).

Labeled neurons were more prevalent in the dorsomedial tip of Pm in cases 1 and 5 (areas 11 and 32) and occupied a central position in the rest of the cases.

Ventral anterior (VA) and ventral lateral (VL) The VA issued projections to all prefrontal areas studied

(Table 1). The highest proportions of labeled neurons in VA (5.8-12.7%) were found in cases with injections in areas 12, 32, and 14 (cases 2,3,5,6; Figs. 4A,B, 5A,B, 7A, 8C, 22C,D). In the rest of the cases, only a few (1.7-3.2%) labeled neurons were noted in VA. In all except one (case 8, area 8; Fig. lOA), most of the labeled neurons were found in magnocellular VA (Fig. 24B), and only a few were noted in its parvicellular subdivision.

Pm projections: topography.

Mediolateral.

Dorsoventral.


Fig. 11. The distribution of labeled neurons in rostral (A) through caudal (D) thalamic levels in case 9 after HRP injection in dorsal area 8 (El.

ul

D

-b 1

b

0 b

82 H. BARBAS ET AL.

Fig. 12. Histogram showing the distribution of labeled neurons within the rostral (shown in category 1) and in progressively more caudal (5) sectors of MDpc and MDmc in cases with HRP injection in basoventral prefrontal areas: A: case 1, area 11; B: case 2, orbital area 12; C: case 3, lateral area 12; D: case 4, ventral area 46.

Fig. 13. Histogram showing the distribution of labeled neurons within the rostral (shown in category 1) and in progressively more caudal (5) sectors of MDpc and MDmc after injection of retrograde tracers in mediodorsal prefrontal areas: A: case 5, area 32; B: case 6, area 14; C: case 7, dorsal area 46; D: case 8, dorsal area 8; E: case 9, dorsal area 8.

t

40 I

i_ 3 1 4 _[L 5

30 L B -

40 I v)

$ 3 301 W 2

1

c 2 3 4 5

1 2 3 4 5

“ 1 D

1 2 3 4 5

MEDIAL---------) LATERAL

Fig. 14. Histogram showing the distribution of labeled neurons in the medial (represented in category l), and in progressively more lateral (category 5) sectors of MDpc (black bars) and MDmc (hollow bars) after HRP injection in basoventral prefrontal areas. A: case 1, area 11; B: case 2, orbital area 12; C: case 3, lateral area 12; D: case 4, ventral area 46.

TABLE 4. Distribution of Labeled Neurons in the Midline Thalamic Nuclei Following Injection of Retrograde Tracers in Prefrontal Cortices

Midline projection zone

Injection Total Case site Pa.Pac Cdc Cim CiF Clc Re %

Basoventral 1 2 3 4

5 6 7 8 9

Mediadorsal

Area 11 0.3 0.4 1.2 1.4 0.2 0.6 4.1 012 0.1 0.3 0.8 2.5 0.2 1.5 5.4

2.1 4.2 L12 - 0.5 - 0.1 0.6 V46

Area 32 1.0 3.1 0.9 3.6 1.6 3.4 13.6 Area 14 1.8 5.6 4.3 3.4 2.8 6.8 24.7 D46

0.04 - 0.04 - - 0.1 D8 - D8

0.7 0.1 1.3 - - - -

- - - - - - -

- - - - - - -

‘Data in columns below nuclear designations are expressed in percentages. *According to the classification of Olszewski (’521; includes a region called central medial by Jones (‘85). _ _ - no labeled neurons

A small number of labeled neurons were found in VLm in all cases except those involving injections in area 46 (cases 4 and 7). In cases 8 and 9 (area 81, some labeled neurons were noted in the caudal and oral VL as well. In addition, a few labeled neurons were seen in nucleus X in all, except for cases 1 and 4.

Intralaminar The distribution of labeled neurons within individual

intralaminar nuclei is shown in Table 3. Though labeled neurons in intralaminar nuclei were found in all cases, the highest proportions (17.3-18.7%) were recorded for the medial cases 5 and 6 (areas 32 and 14; Figs. 7, 8, 22B), followed by cases 8 and 9 (dorsal area 8,6-7%; Figs. 10-11). In the rest of the cases, the intralaminar nuclei accounted for only a small proportion of the labeled neurons (3.1- 3.9%).

Midline and Anterior Both midline and anterior nuclei showed a preferential

projection to some, though not the same, prefrontal areas (Tables 1, 4). The midline nuclei included the highest proportions of labeled neurons directed to medial areas 32 and 14 (13.6-24.7%; Figs. 7, 8, 24B). Midline nuclei also included a small proportion (4.1-5.4%) of the labeled neurons projecting to areas 11 and 12 (cases 1-3, Table 4).

The anterior nuclei showed a pattern of projection essen- tially complementary to that noted for the midline: they contained labeled neurons mostly in the orbital cases 1 and 2 (7%), a few in case 3 (Table 1; Figs. 3-5,24C,D), and 1% of those directed to medial areas (cases 5,6). Most of the labeled neurons were found in the AM nucleus in all cases. Only occasional labeled neurons were noted in AV (cases 2 and 5), or AD (case 5). Neither midline nor the anterior nuclei showed significant labeling after injections in areas 46 or 8.

DISCUSSION In the present study using relative densities of labeled

neurons as an indicator, we found differences in the projection from MD to prefrontal cortices. MD provided a clear majority of thalamic neurons projecting to prefrontal cortices in most cases, but it contributed a lesser, though still a substantial proportion of neurons directed to others. These

84 H. BARBAS ET AL.

Fig. 15. Darkfield photomicrographs of coronal sections through the thalamus of tissue treated to visualize HRP, showing: (A) the location of labeled neurons primarily in the magnocellular (mc) portion of MD (white arrows), but also in MDpc (arrowhead) in case 1 with HRP injection in area 11; (C) in the ventral portion of MDpc (arrows), and to a lesser extent in MDmc in case 2 with HRP injection in orbital area 12;

(E) mostly in MDpc (arrows) in case 4 with HRP injection in ventral area 46. B, D, and F show the labeled neurons depicted in the opposite frame at higher magnification. White circles in the low and high magnification pairs show blood vessel for reference. Scale bar for A, C, and E = 1 mm, and 100 p m for B, D, and F.


:"i 20 A

1 2 3 4 5

1 2 3 4 5 1 2 3 4 5

601 A

1 2 3 4 5

1 2 3 4 5

2 4 0 c E 1 :" 1 2 3 4 5

MEDIAL -* LATERAL

Fig. 16. Histogram showing the distribution of labeled neurons in the medial (category 1) and in progressively more lateral (category 5) portions of MDpc and MDmc after injection of retrograde tracers in mediodorsal prefrontal areas: A case 5, area 32; B: case 6, area 14; C: case 7, dorsal area 46; D case 8, dorsal area 8; E: case 9, dorsal area 8.

results confirm and extend previous findings (Walker, '36, '40a; Mettler, '47; Von Bonin and Green, '49; Pribram et al., '53; Tobias, '75; Kievit and Kuypers, '77; Barbas and Mesulam, '81; Goldman-Rakic and Porrino, '85; Giguere and Goldman-Rakic, '88). The lowest percentage of labeled neurons in MD was noted in case 6 (area 14). This observation is consistent with previous findings reporting that some medial sites (situated dorsal to the ones reported in this study) had the lowest density of label in MD when compared with other prefrontal areas (Goldman-Rakic and Porrino, '85). However, more information is necessary to determine whether these findings apply to medial areas, in general, or even to all of area 14.

Relationship of thalamic projections to cortical architecture

The present investigation was conducted in the context of our previous architectonic study, where the prefrontal cortex was subdivided in two sectors on the basis of structural and intrinsic connectional criteria (Barbas and Pandya, '89). The two sectors were designated basoventral and mediodorsal to describe their anatomic locations (in Fig. 1 the heavy dashed line divides the two). Our architectonic analysis was founded on the observation that not all cortices have six layers and those that do vary in how distinct their layers are (Barbas and Pandya, '89). Each of the two prefrontal sectors is composed of a series of cortices characterized by different degrees of laminar definition. Following the above classification in the present study, basoventral areas 11,012, L12, and V46 are depicted in an ascending order of laminar differentiation, as are mediodorsal areas 32,14, D46, and D8 (Tables 1-4). The rationale for this ordering is based on our observations that the laminar differentiation of individual prefrontal areas, as well as the architectonic sector to which areas belong, seems to be related to the pattern of their connections (Barbas, '86, '88; Barbas and Pandya, '89; Barbas and De Olmos, '90). We now report that the thalamic projections to prefrontal areas also varied along these lines. Figure 25 summarizes the most important connectional patterns.

Our findings that MD neurons directed to basoventral prefrontal areas were found ventrally (Fig. 25 hollow symbols), whereas those directed to mediodorsal cortices were situated dorsally (Fig. 25 solid black symbols) provide evidence of some degree of topographic segregation in these projections. These findings are consistent with previous reports (Walker, '40a; Giguere and Goldman-Rakic, '88). However, the most striking connectional patterns between the thalamus and the prefrontal cortex were associated with the laminar definition of the prefrontal target areas. For example, areas with a low degree of laminar differentiation received widespread projections involving many thalamic nuclei (Fig. 25, triangles). In contrast, cortices with a high degree of laminar definition received projections from fewer thalamic sources (Fig. 25, squares). An example is provided by comparing the widespread projections of area 32 with the more restricted projections of area 8 (Table 1). These cortices occupy opposite poles within the mediodorsal sector: area 32 represents an area characterized by low and area 8 by high laminar definition.

The preferential projections of MDmc, midline, anterior, and MDmf seemed to be associated with the laminar definition of the prefrontal target areas as well. MDmc reached preferentially those orbital and medial cortices

86 H. BARBAS ET AL.

Fig. 17. Darkfield photomicrographs of coronal sections through the mediodorsal nucleus showing (A), tissue treated for the simulta- neous visualization of AChE and HRP showing labeled neurons (white dots between arrows) found mostly in MDmc in case 5 with an HRP injection in area 32. The adjacent parvicellular portion of MD has a higher AChE content (white network) than MDmc. The labeled neurons are shown at higher magnification in B. In C Labeled neurons are found mostly in the lateral portion of MDpc (arrows) in case 7 with an

HRP injection in dorsal area 46; these neurons are shown at higher magnification in D. E. Labeled neurons in the lateral portion of MDpc and in MDmf (region between arrowheads) in case 8 with an HRP injection in dorsal area 8; neurons found in MDmf are shown at higher magnification in F. White circle in the pairs A-B and C-D indicate blood vessels for reference. Scale bar = 1 mm for A, C, and E; 100 km for B, D, and F.

H. BARBAS ET AL. 87

15 I "1 A

z n : 20

3 -

8 1 0 -

1 2 3 4 5 0 -

" } B

1

1 2 3 4 5 DORSAL + VENTRAL

Fig. 18. Histograms showing the distribution of labeled neurons in the dorsal (shown in category 1) and in progressively more ventral (category 5) portions of MDpc and MDmc after injection of retrograde tracers in basoventral (A, cases 1-4) and mediodorsal (B, cases 5-9) prefrontal areas.

characterized by a low or moderate degree of laminar definition. Midline and anterior nuclei targeted these areas too, but showed regional specificity. Midline nuclei projected preferentially to medial sites (areas 32 and 14; Fig. 25 solid triangles), whereas the anterior nuclei projected selec- tively to orbital areas 11 and 12 (Fig. 25, hollow triangles). None of the above thalamic nuclei had significant links with the eulaminate lateral prefrontal areas. The latter (areas 46 and 8) received the majority of their thalamic projections from the parvicellular and multiform sectors of MD (Fig. 25, squares).

The association of thalamic origin with the laminar definition of the prefrontal target area extended to the specific topography within one nucleus, or nuclear subdivision, as well. For example, orbital and medial cortices were represented medially within MD or the medial pulvinar (Fig. 25 triangles), whereas lateral areas were represented laterally (Fig. 25, squares). These findings are consistent with previous reports (Walker, '40a; Pribram et al., '53;

3 10 z lA 9 5 [

3 $U 1 2 3 4 5

'5 1 v) z 3 10 z 0

8

Y g 5 8

1 2 3 4 5

'5 I

1 2 3 4 5 15 I

2 10 W !$ z ID

1 2 3 4 5 ROSTRAL -* CAUDAL

Fig. 19. Histogram showing the distribution of labeled neurons in the rostral (category 1) and in progressively more caudal (category 5) portions of the medial pulvinar in cases with HRP injection in basoventral prefrontal sites: A: case 1, area 11; B: case 2, orbital area 12; C: case 3, lateral area 12; D: case 4, ventral area 46.

Tobias, '75; Gower, '81; Goldman-Rakic and Porrino, '85; Giguere and Goldman-Rakic, '88).

In addition, we now report that the rostrocaudal position of labeled neurons in MD varied systematically depending on the degree of the laminar differentiation of the prefrontal target areas. Within MD this relationship was particu-

88

2

H. BARBAS ET AL.

l5 I 15 I

A

I5 t l5 I

j o 1 0 5

6

lzL 1

Z l 0

W 3 4 m

aQ

t

- 4 5 2 3

1 2 3 4 5 15

v) z

10 2

9 z 5 4 *

C C

1 2 3 4 5 15 I

lo% 0 5

l= I

l 0 I 5

D

zp m 1 2 3 4 5

MEDIAL ------b LATERAL HUSIRAL ~ CAUDAL

Fig. 21. Histogram showing the distribution of labeled neurons in the medial (category 11, and in progressively more lateral (category 5) sectors of the medial putvinar in cases with HRP injection in basoventral prefrontal areas: A: case 1, area 11; B, case 2, orbital area 12; C: ease 3, lateral area 12; D, case 4, ventral area 46.

Fig. 20. Histogram showing the distribution of labeled neurons in the rostral (category 1) and in progressively more caudal (category 5 ) sectors of the medial pulvinar in cases with injection of retrograde tracers in mediodorsal prefrontal areas: A: case 5, area 32; B: case 6, area 14; C: case 8, dorsal area 8; D: case 9, dorsal area 8.


Fig. 22. A. Brightfield photomicrograph of a coronal section through the intralaminm nuclei highlighted in tissue treated to visualize ACh" (dark reaction product) which stains these nuclei darkly. B. Darkfield photomicrograph of section matched to that shown in A and treated to visualize HRP-labeled neurons in the nucleus paracentralis in case 5, with HRP injection in area 32. The labeled neurons are found in the vicinity of blood vessels used as landmarks, which border the paracentral nucleus (arrowheads in A and B). C. Darkfield photomicrograph of a coronal section through the anterior thalamus showing HRP labeled neurons in VAmc (small white arrows) in case 2 with an HRP injection in orbital area 12; the labeled neurons are found in the vicinity of the mamillothalamic tract bounded by two blood vessels dorsally (white

circles). Labeled neurons are seen as well in parvicellular VA, which is situated dorsolaterally (arrowhead). The labeled neurons in VAmc are shown in D at higher magnification. E. Darkfield photomicrograph of a coronal section showing labeled neurons in the medial pulvinar (w- rows) in case 2 with an HRP injection in orbital area 12; these neurons are shown in F at higher magnification. White circle in E and F indicate blood vessel for reference. Labeled neurons in the above nuclei were found in all cases. In addition, some labeled neurons are seen in the anterior medial nucleus (C, large arrow silhouette), observed in only some of the cases. Scale bar in E = 1 mm applies to A and C; scale bar in F = 100 km applies to B and D.

90 H. BARBAS ET AL.

'5 I v)

H w 3 10 I* z

0 5 1 / 1

15 I

3 10 W z i" I 2 3 4 5

v)

H 2 10 / c z

4 8

n - I 2 3 4 5

l5 I

bp

1 2 3 4 5 MEDIAL 4 LATERAL

Fig. 23. Histograms showing the distribution of labeled neurons in the medial (category 11, and in progressively more lateral portions of the medial pulvinar in cases with injection of retrograde tracers in mediodorsal prefrontal areas: A: case 5, area 32; B: case 6, area 14; C: case 8, dorsal area 8; D case 9, dorsal area 8.

larly evident for the parvicellular sector. Thus the caudal sectors of MDpc projected to areas characterized by a low or moderate degree of laminar differentiation, such as area 32 or area 11 (Fig. 25D,E, triangles). In contrast, the rostral portions of MDpc projected to areas with a high degree of laminar definition, such as caudal areas 46 and 8 (Fig. 25B,C, squares). We made parallel observations for projections originating in the medial pulvinar.

The above rostrocaudal relationship in the thalamofrontal projections has not been hitherto described. Although this may seem surprising in view of the robustness of this gradient, it may be attributed to the technical limitations of the ablation-degeneration procedure in studies where this issue was addressed (Walker, '40a; Pribram et al., '53). This finer topographic analysis adds a new dimension to the pattern of thalamofrontal connections and may have implications for the evolution of these thalamic nuclei. Caudal MD and Pm, which were consistently connected with the limbic prefrontal cortices, may represent phylogenetically old districts of the nuclei. The mediodorsal and medial pulvinar nuclei, which are large in primates (Jones, '85), may have expanded in a caudorostral direction in concert with the evolution of the prefrontal cortex. In this regard, caudal MD and Pm may join MDmc, midline, anterior, the lateral dorsal, and intralaminar nuclei, which have been affiliated with the limbic system on the basis of their architecture and connections (Rose and Woolsey; '48; Yak- ovlev et al., '60; Locke et al., '64; Vogt et al., '79, '87; Yeterian and Pandya, '88).

Functional implications of projection patterns The rostrocaudal topography of projections from MD to

prefrontal cortices may be significant from a functional point of view as well. Whereas a number of behavioral studies have implicated MD in mnemonic processes in primates (Markowitsch, '82 for review; Aggleton and Mishkin, '83a,b), disturbances in memory are more prevalent with posterior MD lesions in both humans (Victor et al., '71) and monkeys (Isseroff et al., '82; Zola-Morgan and Squire, '85). Moreover, damage to basal and medial prefrontal cortices results in visual recognition deficits in monkeys (Bachevalier and Mishkin, '86). Our data demonstrate an anatomic linkage between these cortical and diencephalic memory-related structures. Further experiments are necessary to determine which aspects of memory may depend on this loop.

In a previous study using most of the same cases described here, we found that cortical input to basoventral and mediodorsal prefrontal sectors is relayed via pathways that are not only segregated anatomically but may also be functionally distinct (Barbas, '88). For example, mediodorsal prefrontal areas receive input from premotor and sensory cortices associated with postural mechanisms and spatial functions (Barbas and Pandya, '87; Barbas, '88). In contrast, basoventral prefrontal areas receive projections from areas concerned with the processing of stimulus features and their significance (Barbas, '88). Whether the differences in the thalamic input to basoventral and mediodorsal prefrontal cortices have Comparable functional implications is not known. The difficulties in addressing this issue are compounded by the fact that little is known about the physiologic properties of neurons in thalamic nuclei that project to prefrontal cortices. However, several lines of evidence suggest that the thalamic input to prefron-

THALAMIC PROJECTIONS TO THE PREFRONTAL CORTEX

Fig. 24. Labeled neurons in midline and anterior nuclei were seen only in a few of the cases. A. Section treated to visualize AChE shows the distribution of this enzyme (dark reaction product) in the anterior (AV and AM), paracentral (Pcn), midline (Clc, large arrow silhouette, reuniens, white arrowhead), and the VAmc nuclei (silhouette arrowhead, in the vicinity of the mamillothalamic tract). B. In a section matched to that depicted in A, HRP-labeled neurons are seen in the

midline nuclei clc (large arrow silhouette), reuniens (Re, white arrowhead), and in VAmc (silhouette arrowhead) in case 5 with an HRP injection in area 32. C. Labeled neurons are seen in AM in case 1 (area 11); these neurons are shown in D at higher magnification. Small arrow in C and D point to the same blood vessel for reference. Scale bar = 1 mm for A and C, and 100 pm for B and D. A was photographed using brightfield, and B-D darkfield illumination.

tal cortices may be functionally specific as well. For example, neurons in the upper portions of the central lateral and paracentral thalamic nuclei respond to visual stimuli and in association with eye movements (Schlag-Rey and Schlag, '84; Schlag and Schlag-Rey, '84). The same portions of these intralaminar nuclei project to area 8, whose connectional and physiologic properties also suggest visual and visuomotor functions (Barbas, '88 for discussion and references). Moreover, area 8 receives some projections from the suprageniculate and limitans nuclei and a substantial projection from MDmf. All of these nuclei receive projections from the deep layers of the superior colliculus (Benevento and Fallon, '75; Benevento et al., '77) whose responses are also related to eye movements (Wurtz and Albano, '80 for review). The possible involvement of MDmf in eye move-

ment is supported further by its connections with the lateral part of the substantia nigra (Ilinsky et al., '85; Ilinsky and Kultas-Ilinsky, '87), which has been implicated in eye movement activities as well (Hikisaka and Wurtz, '83a-d). Collectively, the intralaminar and MDmf nuclei made up a higher proportion of the neurons directed to the mediodorsal than to the basoventral prefrontal sector (Ta- ble 1; Fig. 25, solid symbols). The above evidence suggests that at least some of the projections from the thalamus to mediodorsal areas may be related to visuomotor functions in a way that the cortical input to these same areas also suggests (Barbas, '88; see also Barbas and Mesulam, '81; Schwartz and Goldman-Rakic, '84; Andersen et al., '85; Huerta et al., '87). It should be noted that the caudal part of the lateral bank of the intraparietal sulcus, a major visuomo-

92 H. BARBAS ET AL.

A . L \

Basal Ventral Medlal mDorsal

Fig. 25. Composite diagram summarizing the major projection patterns between the thalamus and prefrontal areas. A-G represent rostra1 to caudal thalamic levels. Hollow symbols represent labeled neurons directed to the basoventral prefrontal sector, solid symbols to

tor region (Lynch, '80; Hyvarinen, '821, receives projections from intralaminar nuclei and from MDmf as well (Schmah- mann and Pandya, '90).

The physiologic properties of other thalamic nuclei that provide the majority of neurons directed to orbital and medial prefrontal areas are not known. However, the role of sensory input, which may reach limbic thalamic nuclei that target preferentially basal and medial prefrontal cortices, will probably be associated with mnemonic processes as the behavioral data suggest (see Squire and Zola-Morgan, '88

the mediodorsal. Within both sectors triangles represent neurons directed to cortices that have a low or moderate degree of laminar differentiation, squares depict neurons projecting to eulaminate areas characterized with a high degree of laminar definition.

for review). Further experiments are necessary to address whether the specific topographic relationships of thalamofrontal projections will have specific behavioral and physiologic attributes as well.

ACKNOWLEDGMENTS We thank Mr. Brian Butler and Ms. Michelle Richmond

Kalajian for excellent technical assistance. We also thank Dr. Robert Sikes for giving us the initial computer program,


and Mr. Shuwan Xue for writing the algorithms and specific software used for the analysis of the data. This study was supported by NIH grant NS24760 and NSF grant BNS 83-15411. C.R.D. was supported by a Fulbright fellowship while on leave from the University of Crete, Greece.

LITERATURE CITED Aggleton, J.P., and M. Mishkin (1983a) Visual recognition impairment

following medial thalamic lesions in monkeys. NeuropsychologiaSl: 189- 197.

Aggleton, J.P., and M. Mishkin (1983b) Memory impairments following restricted medial thalamic lesions in monkeys. Exp. Brain Res. 52199- 209.

Akert, K. (1964) Comparative anatomy of frontal cortex and thalamofrontal connections. In J.M. Warren and K. Akert (eds): The Frontal Granular Cortex and Behavior. New York: McGraw-Hill, pp. 372-396.

Andersen, R.A., C. Asanuma, and W.M. Cowan (1985) Callosal and prefrontal associational projecting cell populations in area 7a of the macaque monkey: A study using retrogradely transported fluorescent dyes. J. Comp. Neurol. 232t443-455.

Asanuma, C., R.A. Andersen, and W.M. Cowan (1985) The thalamic relations of the caudal inferior parietal lobule and the lateral prefrontal cortex in monkeys: Divergent cortical projections from cell clusters in the medial pulvinar nucleus. J. Comp. Neurol. 241:357-381.

Bachevalier, J.B., and M. Mishkin (1986) Visual recognition impairment follows ventromedial but not dorsolateral prefrontal lesions in monkeys. Behav. Brain Res. 20:249-261.

Baleydier, C., and F. Mauguiere (1985) Anatomical evidence for medial pulvinar connections with the posterior cingulate cortex, the retrosple- nial area, and the posterior parahippocampal gyrus in monkeys. J. Comp. Neurol. 232t219-228.

Barbas, H. (1986) Pattern in the laminar origin of corticocortical connections. J. Comp. Neurol. 252:415-422.

Barbas, H. (1988) Anatomic organization of basoventral and mediodorsal visual recipient prefrontal regions in the rhesus monkey. J. Comp. Neurol. 276:313-342.

Barbas, H., and J. De Olmos (1990) Projections from the amygdala to basoventral and mediodorsal prefrontal regions in the rhesus monkey. J. Comp. Neurol. 300:549-571.

Barbas, H., and M-M. Mesulam (1981) Organization of afferent input to subdivisions of area 8 in the rhesus monkey. J. Comp. Neurol. 200:407- 431.

Barbas, H., and M-M. Mesulam (1985) Cortical afferent input to the principalis region of the rhesus monkey. Neurosci. 15:619-637,1985.

Barbas, H., and D.N. Pandya (1987) Architecture and frontal cortical connections of the premotor cortex (area 6) in the rhesus monkey. J. Comp. Neurol. 256:211-218.

Barbas, H., and D.N. Pandya (1989) Architecture and intrinsic connections of the prefrontal cortex in the rhesus monkey. J. Comp. Neurol. 286:353-375.

Benevento, L.A., and J.H. Fallon (1975) The ascending projections of the superior colliculus in the rhesus monkey (MCUXXR nulatta). J. Comp. Neurol. 160:339-362.

Benevento, L.A., M. Rezak, and R. Santos-Anderson (1977) An autoradio- graphic study of the projections of the pretectum in the rhesus monkey (Macaca rnulatta): Evidence for sensorimotor links to the thalamus and oculomotor nuclei. Brain Res. 127t193-218.

Bos, J., and L.A. Benevento (1975) Projections of the medial pulvinar to orbital cortex and frontal eye fields in the rhesus monkey (Macaca rnulatt~). Exp. Neurol. 49t487-496.

Brodmann, K. (1905) Beitrage zur histoligischen Localisation der Grosshirn- rinde: 111. Mitteilung: Die Rindernfelder der niederen Men. J. Psychol. Neurol. 4:177-266.

Carmel, P.W. (1970) Efferent projections of the ventral anterior nucleus of the thalamus in the monkey. Am. J. Anat. 128t159-184.

Gallyas, F. (1979) Silver staining of myelin by means of physical develop- ment. Neurol. Res. 1:203-209.

GeneserJensen, F.A., and T.W. Blackstad (1971) Distribution of acetyl cholinesterase in the hippocampal region of the guinea pig. Z. Zellforsch. Mikrosk. Anat. 114t460-481.

Giguere, M., and P.S. Goldman-Rakic (1988) Mediodorsal nucleus: Areal, laminar, and tangential distribution of afferents and efferents in the frontal lobe of rhesus monkeys. J. Comp. Neurol. 277t195-213.

Goldman-Rakic, P.S., and L.J. Porrino (1985) The primate mediodorsal (MD) nucleus and its projection to the frontal lobe. J. Comp. Neurol. 242535-560.

Gower, E.C. (1981) Architectonic and Thalamic Connections of Macaque Temporal Polar Cortex. Doctoral dissertation, Massachusetts Institute of Technology.

Hikosaka, O., and R.H. Wurtz (1983a) Visual and oculomotor functions of monkey substantia nigra pars reticulata. I. Relation of visual and auditory responses to saccades. J. Neurophysiol. 49:1230-1253.

Hikosaka, O., and R.H. Wurtz (1983b) Visual and oculomotor functions of monkey substantia nigra pars reticulata. 11. Visual responses related to fixation of gaze. J. Neurophysiol. 49t1254-1267.

Hikosaka, O., and R.H. Wurtz (1983~) Visual and oculomotor functions of monkey substantia nigra pars reticulata. 111. Memory contingent visual and saccade responses. J. Neurophysiol. 49: 1268-1284.

Hikosaka, O., and R.H. Wurtz (1983d) Visual and oculomotor functions of monkey substantia nigra pars reticulata. IV. Relation of substantia nigra to superior colliculus. J. Neurophysiol. 49:1285-1301.

Huerta, M.F., L.A. Krubitzer, and J.H. G a s (1987) Frontal eye field as defined by intracortical microstimulation in squirrel monkeys, owl monkeys, and Macaque monkeys 11. Cortical connections. J. Comp. Neurol. 265332361.

Hyvarinen, J. (1982) Posterior parietal lobe of the primate brain. Physiol. Rev. 621060-1129.

Ilinsky, LA., and K. Kultas-Ilinsky (1987) Sagittal cytoarchitectonic maps of the Macaca mulatta thalamus with a revised nomenclature of the motor-related nuclei validated by observations on their connectivity. J. Comp. Neurol. 262:331-364.

Ilinsky, LA., M.L. Jouandet, and P.S. Goldman-Rakic (1985) Organization of the nigrothalamocortical system in the rhesus monkey. J. Comp. Neurol. 236:315-330.

Isseroff, A,, H.E. Rosvold, T.W. Galkin, and P.S. Goldman-Rakic (1982) Spatial memory impairments following damage to the mediodorsal nucleus of the thalamus in rhesus monkeys. Brain Res. 23297-113.

Jones, E.G. (1985) The Thalamus. New York: Plenum Press. Jones, E.G., and R.Y. Leavitt (1974) Retrograde axonal transport and the

demonstration of non-specific projections to the cerebral cortex and striatum from thalamic intralaminar nuclei in the rat, cat, and monkey. J. Comp. Neurol. 154:349-378.

Kievit, J., and H.G.J.M. Kuypers (1977) Organization of the thalamocortical connexions to the frontal lobe in the rhesus monkey. Exp. Brain Res. 29:299-322.

Le Gros Clark, W.E. (1932) The structure and connections of the thalamus. Brain 35:406470.

Locke, S. (1969) Projection of the magnocellular medial dorsal nucleus of the monkey. Modern Neurology. Boston: Little, Brown, pp. 55-67.

Locke, S., J.B. Angevine, Jr., and P.I. Yakovlev (1964) Limbic nuclei of thalamus and connections of limbic cortex. Arch. Neurol. 11:l-12.

Lu, X.-C.M., J. Bachevalier, D. Kowalska, and L.G. Ungerleider (1988) Thalamic input to medial prefrontal cortex in macaques. Neurosci. Abstr. 14:1228.

Lynch, J.C. (1980) The functional organization of posterior parietal association cortex. Behav. Brain Sci. 3,485-534.

Markowitsch, H.J. (1982) Thalamic mediodorsal nucleus and memory: A critical evaluation of studies in animals and man. Neurosci. Biobehav. Rev. 6:351-380.

Mesulam, M-M., E. Hegarty, H. Barbas, K.A. Carson, E.C. Gower, A.G. Knapp, M.B. Moss, and E.J. Mufson (1980) Additional factors influ- encing sensitivity in the tetramethyl benzidine method for HRP neuro- histochemistry. J. Histochem. Cytochem. 28: 1255-1259.

Mettler, F.A. (1947) Extracortical connections of the primate cerebral cortex. J. Comp. Neurol. 86t95-117.

Nauta, W.J.H. (1962) Neural associations of the amygdaloid complex in the monkey. Brain 85:505-520.

Nauta, W.J.H. (1971) The problem of the frontal lobe: a reinterpretation. J. Psychiat. Res. 8,167-187.

Olivier, A,, A. Parent, and L.J. Poirier (1969) Identification of the thalamic nuclei on the basis of their cholinesterase content in the monkey. J. Anat. 106t37-50.

Olszewski, J. (1952) The Thalamus of the Macaca Mulatta. An Atlas for Use with the Stereotaxic Instrument. Basel: S. Karger.

94

Potter, H., and W.J.H. Nauta (1979) A note on the problems of olfactory associations of the orbitofrontal cortex in the monkey. Neurosci. 4:361- 367.

Pribram, K.H., K.L. Chow, and J. Semmes (1953) Limit and organization of the cortical projection from the medial thalamic nucleus in monkey. J. Comp. Neurol. 98:433-448.

Reep, R. (1984) Relationship between prefrontal and limbic cortex: A comparative and anatomical review. Brain, Behav. Evol. 2 5 - 8 0 ,

Rose, J.E., and C.N. Woolsey (1948) The orbitofrontal cortex and its connections with the mediodorsal nucleus in rabbit, sheep and cat. Res. Publ. Assoc. Nerv. & Ment. Dis. 27:210-232.

Schlag, J., and M. Schlag-Rey (1984) Visuomotor functions of central thalamus in monkey. 11. Unit activity related to visual events, targeting, and fixation. J. Neurophysiol. 5lt1175-1195.

Schlag-Rey, M., and J. Schlag (1984) Visuomotor functions of central thalamus in monkey. I. Unit activity related to spontaneous eye movements. J. Neurophysiol. 51t1149-1174.

Schmahmann, J.D., and D.N. Pandya (1990) Anatomical investigation of projections from thalamus to posterior parietal cortex in the rhesus monkey: A WGA-HRP and fluorescent tracer study. J. Comp. Neurol. 295.299326.

Schwartz, M.L., and P.S. Goldman-Rakic (1984) Callosal and intrahemi- spheric connectivity of the prefrontal association cortex in rhesus monkey: Relation between intraparietal and principal sulcal cortex. J. Comp. Neurol. 226:403420.

Squire, L.R., and S. Zola-Morgan (1988) Memory: Brain systems and behavior. Trends in Neurosci. 11: 170-175.

Tanaka, D., Jr. (1976) Thalamic projections of the dorsomedial prefrontal cortex in the rhesus monkey (Macaca mulatta). Brain Res. 110:21-38.

Tobias, T.J. (1975) Merents to prefrontal cortex from the thalamic mediodorsal nucleus in the rhesus monkey. Brain Res. 83:191-212.

Trojanowski, J.Q., and S. Jacobson (1974) Medial pulvinar afferents to

H. BARBAS ET AL. frontal eye fields in rhesus monkey demonstrated by horseradish peroxidase. Brain Res. 80:395411.

Victor, M.R., R. Adams, and G. Collins (1971) The Wernicke-Korsakoff Syndrome. Philadelphia: F.A. Davis.

Vogt, B.A., D.N. Pandya, and D.L. Rosene (1987) Cingulate cortex of the rhesus monkey: I. Cytoarchitecture and thalamic afferents. J. Comp. Neurol. 262256-270.

Vogt, B.A., D.L. Rosene, and D.N. Pandya (1979) Thalamic and cortical f ierents differentiate anterior from posterior cingulate cortex in the monkey. Science 204.205-207.

Von Bonin, G., and J.R. Green (1949) Connections between orbital cortex and diencephalon in the macaque. J. Comp. Neurol. 90:243-254.

Walker, A.E. (1936) An experimental study of the thalamocortical projection of the macaque. J. Comp. Neurol. 64:l-39.

Walker, A.E. (1938) The Primate Thalamus. Chicago: University of Chicago Press.

Walker, A.E. (1940a) The medial thalamic nucleus. J. Comp. Neurol. 73237-115.

Walker, A.E. (1940b) A cytoarchitectural study of the prefrontal area of the macaque monkey. J. Comp. Neurol. 98:59-86.

Wong-Riley, M. (1979) Columnar cortico-cortical interconnections within the visual system of the squirrel and macaque monkeys. Brain Res. 162:ZOl-2 17.

Wurtz, R.H., and J.E. Albano (1980) Visual-motor function of the primate superior colliculus. In W.M. Cowan, J.W. Hall, and E.R. Kandel (eds): Ann. Rev. Neurosci., Vol. 3, Palo Alto, Annual Reviews Inc., pp. 189-226.

Yakovlev, P.I., S. Locke, D.Y. Koskoff, and R.A. Patton (1960) Limbic nuclei of thalamus and connections of limbic cortex. AMA Arch. Neurol. 3:62O-641.

Yeterian, E.H., and D.N. Pandya (1988) Corticothalamic connections of paralimbic regions in the rhesus monkey. J. Comp. Neurol. 269:130-146.

Zola-Morgan, S., and L.R. Squire (1985) Amnesia in monkeys after lesions of the mediodorsal nucleus of the thalamus. Ann. Neurol. 17:558-564.

diverse thalamic projections to the prefrontal cortex in the rhesus monkey

Documents