Corticocortical connections of visual, …brainmap.wustl.edu/resources/Lewis_JCN00b.pdfCorticocortical Connections of Visual, Sensorimotor, ... Ri retroinsula S2cx ... Corticocortical

Corticocortical Connections of Visual,Sensorimotor, and Multimodal

Processing Areas in the Parietal Lobe ofthe Macaque Monkey

JAMES W. LEWIS1* AND DAVID C. VAN ESSEN2

1California Institute of Technology, Pasadena, California 168252Washington University School of Medicine, St. Louis, Missouri 63110

ABSTRACTWe studied the corticocortical connections of architectonically defined areas of parietal

and temporoparietal cortex, with emphasis on areas in the intraparietal sulcus (IPS) that areimplicated in visual and somatosensory integration. Retrograde tracers were injected intoselected areas of the IPS, superior temporal sulcus, and parietal lobule. The distribution oflabeled cells was charted in relation to architectonically defined borders throughout thehemisphere and displayed on computer-generated three-dimensional reconstructions and oncortical flat maps. Injections centered in the ventral intraparietal area (VIP) revealed acomplex pattern of inputs from numerous visual, somatosensory, motor, and polysensoryareas, and from presumed vestibular- and auditory-related areas. Sensorimotor projectionswere predominantly from the upper body representations of at least six somatotopicallyorganized areas. In contrast, injections centered in the neighboring ventral lateral intrapa-rietal area (LIPv) revealed inputs mainly from extrastriate visual areas, consistent withprevious studies. The pattern of inputs to LIPv largely overlapped those to zone MSTdp, anewly described subdivision of the medial superior temporal area. These results, in conjunc-tion with those from injections into other parietal areas (7a, 7b, and anterior intraparietalarea), support the fine-grained architectonic partitioning of cortical areas described in thepreceding study. They also support and extend previous evidence for multiple distributednetworks that are implicated in multimodal integration, especially with regard to area VIP.J. Comp. Neurol. 428:112–137, 2000. © 2000 Wiley-Liss, Inc.

Indexing terms: ventral intraparietal; primate neocortex; vestibular; auditory; polysensory

A major proposed role of parietal cortex in primates is toestablish and maintain a spatial reference system derivedfrom multiple sensory sources, in order to guide eye, limb,and/or body movements (for review, see Hyvarinen, 1982;Andersen, 1987; Mesulam, 1998; Colby and Goldberg,1999). Within parietal cortex of the macaque, the intrapa-rietal sulcus (IPS) contains several functionally distinctdomains that are implicated in different aspects of sen-sory, polysensory, and sensorimotor function, and also inspatial attention (Andersen et al., 1985b; Colby and Du-hamel, 1991; Andersen, 1995; Bremmer et al., 1997; Du-hamel et al., 1998; Snyder et al., 1998). Given the evidencefrom the preceding study (Lewis and Van Essen, 2000)that the architectonic subdivisions of the IPS are evenmore numerous than previously appreciated, it is impor-tant to understand the detailed connectivity of the constit-uent areas and architectonically distinct zones of the IPS.

The present study examines the connectivity of severalsubdivisions, concentrating on those in and near the fun-dus of the IPS, where multimodal integration appears tobe strongest.

Previous connectivity studies have demonstrated that:(1) the lateral bank of the IPS is extensively intercon-nected with known visual areas (Maunsell and Van Essen,1983; Colby et al., 1988; Boussaoud et al., 1990); (2) themedial bank is connected with known somatosensory andsomatomotor areas (Jones et al., 1978; Vogt and Pandya,

Grant sponsor: NIH; Grant number: EY02091.*Correspondence to: James W. Lewis, Ph.D., Department of Cellular

Biology, Neurobiology and Anatomy, Medical College of Wisconsin, 8701Watertown Plank Rd., Milwaukee, WI 53226. E-mail: [email protected]

Received 17 June 1999; Revised 27 June 2000; Accepted 31 July 2000

THE JOURNAL OF COMPARATIVE NEUROLOGY 428:112–137 (2000)

© 2000 WILEY-LISS, INC.

1978; Pearson and Powell, 1985; Pons and Kaas, 1986;Burton and Fabri, 1995); and (3) the fundus of the IPS hasconnections with both visual and somatosensory areas,suggestive of a major integrative role (Pandya andKuypers, 1969; Jones and Powell, 1970; Seltzer and Pan-dya, 1980; Pandya and Seltzer, 1982). In addition, IPSconnections with motor and premotor areas in the frontallobe suggest involvement in eye and upper limb move-ments (Leichnetz, 1986; Andersen et al., 1990; Blatt et al.,1990; Tanne et al., 1995). Although many of these pub-lished connections have been assigned to particular sourceand target areas (cf. Felleman and Van Essen, 1991),many others cannot be specifically linked to recently iden-tified architectonic subdivisions, including those reportedin the preceding study. Thus, we have only a fragmentaryunderstanding of the relative strengths and precise distri-butions of connections between architectonically definedregions of the IPS and the numerous different areasthroughout the hemisphere.

Interest in understanding the detailed connectivity (andarchitecture) of the IPS is heightened by physiologicalstudies that reveal significant functional specializationswithin different regions of the IPS. Previous studies sug-gest that the middle to anterior thirds of the IPS have arole in visually guided reaching and grasping movements(Leinonen and Nyman, 1979; Taira et al., 1990; Tanne etal., 1995), complex arm and hand movements (Sakata etal., 1973, 1992; Mountcastle et al., 1975), reaching-relatedmovements with the head and lips (Leinonen and Nyman,1979), visual and vestibular integration that involve com-parisons of eye and head movements (Graf et al., 1996;Bremmer et al., 1997), the analysis of stimulus motion ofobjects in the near extrapersonal space (Colby and Du-hamel, 1991; Duhamel et al., 1998), and perhaps eye-handor eye-hand and mouth coordination, as occurs duringfeeding behaviors (Colby et al., 1993a).

Many of the visual and somatosensory inputs to the IPSare from areas that contain topographically organized

Abbreviations

Cortical areasA1 primary auditory cortexAIP anterior intraparietal areaER entorhinal cortexFST floor of superior temporal areaG gustatory cortexIa agranular insular cortexId dysgranular insular cortexIg granular insular cortexIPa area IPa from Cusick et al. (1995)LIPd lateral intraparietal (dorsal)LIPv lateral intraparietal (ventral)M2 motor area 2MDP medial dorsal parietal areaMIP medial intraparietal areaMST medial superior temporal (complex)MT middle temporal areaPi parainsular areaPIP posterior intraparietal areaPO parietal-occipital areaPrCO precentral opercular cortexProS prostriate cortexR auditory area RRi retroinsulaS2cx second somatosensory area (complex)SMA supplementary motor areaTAa area TAa from Cusick et al. (1995)TEa/m area TEa plus TEm of Seltzer and Pandya (1978)TE1-3 subdivisions TE1, TE2, and TE3 (Seltzer and Pandya,

1978)TF, TH temporal areas F and HTpt temporoparietal areaTPOc temporal parietal occipital (caudal)TPOi temporal parietal occipital (intermediate)TPOr temporal parietal occipital (rostral)V1 visual area 1V2 visual area 2V3 visual area 3V3A visual area V3AV4 visual area 4V4t V4 transitional area (complex)VIP ventral intraparietal (complex)VOT ventral occipitotemporal areaVP ventral posterior area1, 2 somatosensory areas 1 and 23a, 3b somatosensory areas 3a and 3b4 primary motor cortex5 somatosensory area 5 (complex)7a visual area 7a

7b somatosensory area 7b7op operculator area 78–14 areas 8 thru 14 from Carmichael and Price (1994)23, 24 cingulate areas 23 and 2429, 30 cingulate areas 29 and 3035, 36 perirhinal areas 35 and 3645 Walker area 4546 visual area 46Cortical zonesDP dorsal prelunate areaLOP lateral occipital parietalMSTda medial superior temporal area (dorsal, anterior)MSTdp medial superior temporal area (dorsal, posterior)MSTm medial superior temporal (medial)RS retrosplenial cortexST superior temporal cortical regionTE1-3(d) dorsal subdivision of TE1, TE2, plus TE3TE1-3(v) ventral subdivision of TE1, TE2, plus TE3Toc temporal opercular caudalVIPI ventral intraparietal (lateral)VIPm ventral intraparietal (medial)V4ta V4 transitional area (anterior)V4tp V4 transitional area (posterior)4C 4C of Barbas and Pandya (1987)5V area 5 (ventral)5D area 5 (dorsal)6D subdivision of area 6D6Ds subdivision of area 6D6DC subdivision of area 6D6Val area 6Va (lateral)6Vam area 6Va (medial)7t subdivision of 7 near tip of IPS8Am,s subdivisions of area 8 from PGR (1991a)46v,p ventral and posterior subdivisions of 46Cortical sulciASd arcuate sulcus, dorsal limbASv arcuate sulcus, ventral limbCaS calcarine sulcusCgS cingulate sulcusCeS central sulcusIOS inferior occipital sulcusIPS intraparietal sulcusLaS lateral (Sylvian) sulcusLuS lunate sulcusOTS occipital temporal sulcusPOS parieto-occipital sulcusPS principal sulcusSTS superior temporal sulcus

113CORTICAL CONNECTIONS OF PARIETAL CORTEX

maps of the retina and body surface, respectively (Maun-sell and Van Essen, 1983; Godschalk et al., 1984; Pearsonand Powell, 1985; Ungerleider and Desimone, 1986b). Ad-ditionally, the ventral intraparietal area (VIP) is known tocontain cells with distinct polysensory receptive fields,responsive to visual and somatosensory (Colby and Du-hamel, 1991) or visual and vestibular sensation (Graf etal., 1996). However, it is unclear whether or how theseinputs are organized within the IPS with regard to topog-raphy. Tactile response properties of VIP cells, which pre-dominantly include the head and face (Colby and Du-hamel, 1991; Colby and Goldberg, 1999), are oftencorrelated with visual receptive fields in their respectivelocation, size, and direction selectivity (Duhamel et al.,1998). For instance, cells with visual receptive fields in thesuperior hemifield often have tactile receptive fields onupper portions of the head and face, whereas the cells withinferior visual receptive fields have tactile receptive fieldson lower portions of the face and neck, and/or the arm andhand. Visual and vestibular receptive fields are also cor-related in terms of direction selectivity (Graf et al., 1996).Beyond these physiological correlations, no clear topo-graphical organization within VIP has been reported.

A number of technical factors have impeded the attain-ment of a more detailed understanding of the connectivityof area VIP and other intraparietal areas in the macaque.One is the general difficulty of making restricted injec-tions in areas that are small in dimensions and burieddeep in sulci that often contain irregular folds. Corticalfolds also hinder analysis and visualization of the distri-bution of complex connectivity patterns. Additionally, pos-terior parietal areas are most strongly connected withareas which themselves have been incompletely chartedand often lack robust architectonic boundaries (Andersenet al., 1990; Blatt et al., 1990; Van Essen et al., 1990;Felleman and Van Essen, 1991).

The analysis of corticocortical connectivity presentedhere capitalizes on the use of multiple retrograde tracersplus the revised and refined architectonic partitioningscheme for parietal cortex and other regions of the hemi-sphere described in the previous study (Lewis and Van

Essen, 2000). In addition, we have analyzed connectionsby using computerized surface reconstructions that allowinterchangeable visualization in multiple formats, includ-ing slices, three-dimensional reconstructions, and flatmaps, which facilitates comparisons across animals. Withthis approach, we studied the corticocortical connectionsof area VIP and newly identified architectonic zoneMSTdp, and also provide a more detailed and/or supple-mental analysis of the connections of areas previouslystudied, including the ventral lateral intraparietal area(LIPv), the anterior intraparietal area (AIP), the middletemporal area (MT), and areas 7a and 7b. Some of theseresults have been presented as abstracts (Lewis and VanEssen, 1994, 1996).

MATERIALS AND METHODS

Results from six cases were analyzed in detail (eachwith multiple injections) and were supplemented by re-sults from an additional five cases. All were from juvenilemacaque fascicularis monkeys (6 females and 5 males)weighing 2.0 to 3.8 kg. After mapping the sulcal geometryof the IPS and/or STS with recording electrodes, restrictedinjections of two or three retrograde neuronal tracers wereplaced into separate locations, with most of the injectionsaimed at buried cortical areas VIP and LIPv, and/or MT(see Table 1 below). The location of injection sites andcorresponding labeled cells throughout the ipsilateralhemisphere and portions of the contralateral hemispherewere charted in relation to architectonic subdivisionsidentified in nearby sections stained for myelin, Nissl, andCAT-301 and SMI-32 immunoreactivity (Lewis and VanEssen, 2000).

Surgical procedures

Invasive procedures included an initial chamber im-plant, physiological mapping of cortical gray matter plusthe injection of tracers, and in some cases electrophysio-logical mapping of portions of somatosensory cortex (seeLewis et al., 1999). All recovery surgical procedures wereconducted with strict aseptic techniques. For surgical

TABLE 1. Data on Tracer Injections (19 Injection Sites in 11 Hemispheres)1

TracerCore inj.

(mm3)Estimated

inj. size LayersEstimated uptake zones

(%)Survival

(days)Figure

no.3

Cases illustratedA (95DR) FB 0.7 md All AIP (70), 7b (?) 19 9

DY 0.5 md All VIPm (90), VIP1 (10) 2, 3AG 1.6 lg 1–3 MSTdp (85) 9

B (94CR) FB 1.5 sm 3–5 LIPv (95) 14 7DY 2.0 lg 1–4 VIP1 (80), LIPv (20) 4, 5AG 1.5 33 md All 7b (95) 7

C (94IR) FB 0.7 md All:2–3 VIP1 (50), LIPv (50) 19DY 0.2 md 1–4 VIP (80), AIP (?) 6AG 0.2 md 1–4 7a/7b (100)

D (93IR) FB 0.6 md 3–6 LIPv (85), VIP1 (15) 14 8DY 1.6 lg 1–4 LIPd (60), 7a/7b (40) 8

E (94EL) FB 6.12 lg All MT (100) 17 10DY 4.62 lg All MT (90), MSTdp (10) 10

Cases not illustratedF (93GR) FB 0.7 md All VIPm (50), 5V (50) 11G (94GR) FB 1.0 lg 2–6 VIP (45), LIPv (45), LIPd (10) 13H (95AR) DY 0.5 md All AIP (80), 7b (20) 21I (95BR) DY 0.9 lg 4–6 MT (95) 17J (95CR) DY 0.2 sm All MT (95) 16K (95GR) FB — sm 1–5 VIP1 (60), LIPv (40) 15

1DY, Diamidino Yellow; FB, Fast Blue; AG, WGAapoHRP-Gold; sm, small (90 to 140 nl); md, medium (140 to 190 nl); lg, large (200 to 300 nl).2Volume estimates affected by compromised tissue or tissue quality.3The last column indicates figures where the corresponding injection data are illustrated. See text for other details. For other abbreviations, see list.

114 J.W. LEWIS AND D.C. VAN ESSEN

preparations, animals were tranquilized with an initialdose of ketamine HCl (10 mg/kg i.m.), and premedicatedwith atropine (27 mg/kg i.m.) or glycopyrolate (11 mg/kgi.v.), dexamethasone (0.27 mg/kg i.v.), and cephalothin (25mg/kg i.m.). Surgical anesthesia was induced with isoflu-orane (2–5%) in air containing 2.5% CO2. The heart rate,expired CO2, and rectal temperature were monitored andmaintained within appropriate physiological limits forsurgical anesthesia throughout operative procedures. Fol-lowing recovery surgeries, animals were maintained forapproximately 1 week on a regimen of antibiotics (cepha-lexin 25 mg/kg twice daily) and analgesics (buprenex, 0.1mg/kg, and then aspirin). All procedures were performedaccording to protocols approved by the Animal Study Com-mittees at the California Institute of Technology and atthe Washington University School of Medicine.

Chamber implant and tracer injections.

A stainless steel chamber was implanted over the pari-etal lobe of one or both hemispheres, centered 1–2 cmlateral to the midline and 2–3 cm anterior to the occipitalridge (approximately over the IPS). In a separate surgery(after a 1-week recovery), cortex overlying the IPS, and insome cases the superior temporal sulcus (STS), was ex-posed with one to three skull openings (1–2 cm2 total).Upon reflecting the dura, the cortical surface was photo-graphed to facilitate cortical mapping. A custom-mademanipulator and microdrive were secured to the chamber.The chamber was filled with mineral oil and sealed toreduce brain pulsations. The sulcal folds of the IPS and/orSTS were mapped by electrophysiological recording ofspontaneous, multiunit activity, by using tungsten elec-trodes (A-M Systems, Inc., Everett, WA). This mappingallowed the determination of structural boundaries in-cluding gray versus white matter and lumen gaps in thesulci. After estimating the locations of desired corticaltarget sites, two or three retrograde tracers were injectedat separate locations in each animal. The retrograde trac-ers were Fast Blue (FB, Sigma Co., St. Louis, MO, 5% aq.),Diamidino Yellow (DY, Sigma Co. 3% aq.), and wheatgerm agglutinin (WGA) apo horseradish peroxidase(HRP)-Gold (AG, 0.7% aq., cf. Menetrey, 1985), which wasprepared as described by Basbaum and Menetrey (1987).Labeled cells and injections in all flat map figures arecolor-coded, with blue corresponding to FB-labeled cells,green for DY, and red for AG.

Tracers were injected by using customized triple-barreled glass pipets (A-M Systems). One barrel wasloaded with a tracer, and the other two with 0.5 M NaCl,which allowed physiological recording to verify the loca-tion of the pipet tip in cortex. Injection volumes rangedfrom 90 to 300 nl for each tracer. To reduce tracer leakagealong the pipet track, the glass pipet was held in place atthe cortical target site for 15–30 minutes prior to injection.Tracer was then slowly injected with an air pressure sys-tem (PV830 Pneumatic Picopump, World Precision Instru-ments, New Haven, CT), and allowed to stand for anadditional 30 minutes before retracting the pipet. Animalsrecovered for an 11- to 21-day period following the tracerinjections. In some cases an electrophysiological mappingexperiment (see Lewis et al., 1999) preceded euthanasiaand intracardial perfusion (refer to Lewis and Van Essen,2000).

Data analysis, architectonics, and unfolding

Sections were cut at 60-mm thickness on a freezingmicrotome, and stored refrigerated in 0.1 M phosphatebuffer at pH 7.4. Blocks were cut orthogonal to the IPS (3hemispheres) or approximately in the coronal plane (8hemispheres ipsilateral to injections, and 2 hemispherescontralateral). Every sixth section was used for viewingthe WGAapoHRP-Gold and fluorescent tracer labeling.The remaining sections were used for histological process-ing, as described in the accompanying paper.

In sections analyzed for labeled cells, gold particles fromthe WGAapoHRP-Gold tracer were developed in one ortwo cycles with a silver enhancement kit (Ted Pella Inc.,Redding, CA, silver enhancing kit). Free-floating sectionswere washed in distilled H2O (5 to 10 minutes), incubatedin the silver enhancement solution (10 to 30 minutes),washed again, returned to 0.1 M phosphate buffer, andmounted onto glass slides coated with gelatin-chromealum. This development procedure did not interfere withthe fluorescent tracers, allowing all three tracers to beviewed in the same tissue sections.

Retrogradely labeled cells were viewed by using a fluo-rescence microscope (Zeiss, Axioplan, Thornwood, NY). Byusing the computerized neuroanatomy software MDPlot(Minnesota DataMetrics, St. Paul, MN, version 3.0), datawere plotted onto a one-in-six set of sections (tracer sec-tions), including fiducial alignment marks, approximatesulcal and fundal boundaries, pial and layer 4 (viewedunder darkfield illumination) contours, retrogradely la-beled cells, and injection sites. Methods for transferringarchitectonic borders and generating three-dimensional(3-D) surface reconstructions and flat maps are describedin Lewis and Van Essen (2000). Retrogradely labeled cellswere assigned 3-D coordinates and an identity accordingto the type of label. Each cell was projected to the nearesttile in the three-dimensional surface reconstruction oflayer 4 (cortical blocks were represented by up to 90,000tiles). By preserving this geometrical relationship, cellscould be positioned on the flat maps relative to the nearesttile. In some regions of cortex, labeled fields from two orthree tracers would overlap, obscuring some labeled celltypes. For illustrations, the positions of cells over thecortical flat maps were layered, such that the cells of oneidentity (e.g., the FB-labeled cells in Fig. 7, blue) could lieabove all the cells of another identity (e.g., the AG-labeledcells in Fig. 7, red). Additionally, cells lying below corticallayer 4 (i.e., in infragranular layers) were raised abovethis surface to make them visible on the flat maps.

In Tables 2 and 3, a projection was considered “strong”if the densest region of labeling within a core region hadmore than 20 cells per mm2, typically saturating a portionof an area or zone on the flat maps (e.g., DY label in themedial intraparietal area, MIP, of Figs. 4F and 5A). Mod-erate projections contained fewer than 20 cells per mm2,and typically contained several dozen cells in a cluster(e.g., DY label in the medial dorsal parietal area, MDP, ofFig. 4F). Light projections either had more than 10 cellswithin a core region or had at least one small patch oflabeled cells with a density greater than 2 cells per mm2.A questionable designation (“?”) was applied when an evenlower density of label fell within a core region or thecortical area had compromised tissue and architectonicidentification. In some instances, a significant labeled fieldstraddled a border outline (1?), indicating that the pro-


jection zone may have a stronger assignment than thatbased just on the core region.

Tracer injection analysis

Table 1 summarizes the 19 tracer injections of thisstudy. The core injection volumes (column 3) are based onserial reconstructions of tracer present in the processedtissue sections. In most cases, this quantitative estimatewas positively correlated with our qualitative estimates(column 4), which are based on both the intended tracervolumes injected (see legend) and the resulting relativesize of the injection core plus the surrounding halo ofintense diffuse labeling. Table 1 also denotes the corticallayers containing the injection core, the cortical subdivi-sion(s) included in the estimated effective uptake regions(including possible tracer leakage along pipet tracks), sur-vival times for each case, and figures in which the injec-tions are represented.

For illustration on flat maps, outlines of the extent ofthe injection core, halo, and the pipet track damage wereprojected to the nearest layer 4 contour and processed inthe same manner as described for architectonic bound-aries. Tracks passing through white matter were not pro-jected onto the cortical flat maps, but rather are impliedby the presence of a gap between the point of entry and theinjection site (e.g., Fig. 2E).

RESULTS

Overview

Most of the desired target zones in this study aresmall (1- to 3-mm-wide), making it difficult to obtaininjections entirely restricted to a single architectonicsubdivision, even though preinjection physiologicalmapping aided in centering the injections within corti-cal gray matter. As indicated in Table 1, the results arederived from 7 injections in which the estimated extentof effective uptake was restricted to one architectonicarea or zone (at least 90%), plus 5 injections largelyrestricted to one area (estimated 80 –90%) and 7 tracerinjections that substantially involved more than onearea or zone. Our analysis concentrates on comparingthe corticocortical connections of deep lying areas of theIPS (notably area VIP) with those of other parietal orvisual areas. Five cases including 11 tracer injectionsare illustrated below (Table 1: Cases A–E), each yield-ing a complex pattern of label across large portions ofthe hemisphere. We present these injections in relationto overall cortical geography, and with respect to archi-tectonically identified subdivisions, with emphasis onconnections that were robust, consistent, or were infor-mative with regard to issues of topography or overlap-ping connections. As in the companion paper (Lewis andVan Essen, 2000), our terminological convention is toreserve the term “area” for domains whose distinctnessis warranted by the aggregate evidence available(typically including criteria other than just architecton-ics). A “zone” is an architectonically distinct regionwhose status as a separate area or as a portion of alarger area awaits further evidence. A “subdivision”encompasses both areas and zones.

To aid in assimilating the complex pattern of connectiv-ity revealed in this study, it is useful to briefly summarizeour major findings. First, area VIP had widespread con-

nections, including the upper body representation of mo-tor and somatosensory cortex (arm, hand, digit, and facerepresentations), multiple visual areas (with indicationsof a patchy visuotopic organization), vestibular-relatedcortex (supporting recent physiological findings), other bi-modal or polysensory areas, plus light connections withauditory areas. Second, area LIPv had many visual-related connections similar to those of neighboring areaVIP, except for stronger connections with visual area 3(V3), but few connections with somatosensory and motorareas. Additionally, the connections of LIPv were strik-ingly similar to those of the newly described zone MSTdp.Third, area AIP had connections with many sensorimotor-related cortical areas, similar but not identical to those ofneighboring area VIP, but AIP received little visual-related input, except from higher visual areas and possi-bly bimodal or polysensory areas.

Many of the connections reported below involved topo-graphically organized sensory or motor areas whose visuo-topic, somatotopic, or tonotopic organization has been re-ported in other studies. To facilitate comparisons oflabeled fields with the internal organization of these ar-eas, Figure 1 shows topographically organized subdivi-sions on lateral and medial 3-D views of most of thehemisphere (Fig. 1A,B) and on a cortical flat map (Fig. 1C)that includes areal boundaries superimposed on a map ofcortical folding (gyral crowns in white, sulcal fundi dark).Known patterns of sensory or motor topography are dis-played on the flat map as estimated from publicationscited in the figure legend. Visual topography is indicatedas central (C) or peripheral (P) and upper (1) and lower (-)field. Somatotopy is indicated by body parts as shown inthe key.

We first present three injections (from three differenthemispheres) placed along the fundus of the IPS, in ornear area VIP. Subsequent figures illustrate additionalinjections that were made in these three cases, plus twoother cases, allowing for a comparative analysis with ar-eas LIPv, AIP, 7a, 7b, MT, and MSTdp. These injectionswere useful for assessing reciprocal connections and ad-dressing issues of tracer leakage. All the connections il-lustrated below are also summarized in Table 3 for exam-ination of trends in connectivity in the last section.

Connections of area VIP

Case A. Figures 2 and 3 illustrate the pattern of labelresulting from a DY injection that was restricted to areaVIP, and included all cortical layers. The injection corewas centered in the medial portion of VIP (VIPm), and theinjection halo extended into supragranular layers of lat-eral portion of VIP (VIPl), as illustrated in a micrograph ofa myelin-stained section (Fig. 2A), on coronal sections(Figs. 2B, 3F,G), and on a flat map of the IPS region (Fig.2E). The gray wedges in Figure 2B indicate architectonictransition regions between zone 5V (medial bank), areaVIP (along the fundus), and areas LIPv and LIPd (lateralbank). The anteroposterior level of the DY injection site isindicated on lateral and medial views (vertical line) of a3-D reconstruction of much of the right hemisphere (Fig.2C,D), which also shows labeled cells visible in each view.Multiple clusters of labeled cells are evident, includingextensive labeling in parieto-occipital cortex, plus label inoccipitotemporal cortex and in frontal cortex. Labeled cellsin buried regions can be readily seen on the flat map (Fig.2F), which shows cells from infragranular as well as su-


pragranular layers projected onto the map. Figure 3 illus-trates label in 13 coronal sections, which illustrates thelaminar distribution of some of the labeled fields, andfacilitates orientation between labeled fields in coronalsections and the flat map representation.

The overall pattern of DY label included a large swath oflabeled cells in occipitoparietal cortex (middle and upperleft on flat map), which included both banks of the IPS,plus moderate labeling along the medial wall, and extend-ing into the parietal operculum of the lateral sulcus (Syl-vian fissure). A moderate swath of label was present infrontal cortex around the arcuate sulcus (upper right), anda lighter swath in occipitotemporal cortex along the pos-terior STS (lower center).

In relation to major architectonic subdivisions detailedin the previous study (Lewis and Van Essen, 2000), thelabeled fields involved many areas of sensory, polysensoryand motor function. Heavy to moderate label resultingfrom the VIP injection included subdivisions known orpresumed to be visual (LIPv, LIPd, PO, LOP, MSTda, and

MSTdp), polysensory (MIP, MDP, 7b, Toc, TPOc, andTPOi), somatosensory (5V, 7op, and light label in area 2),and premotor related subdivisions (4C, M2, 24d, and 8Ac).(See Abbreviations list for full names.)

Among topographically organized visual areas (refer toFig. 1), there were moderate connections with area PO,mainly from the estimated upper peripheral visual fieldrepresentations (Colby et al., 1988), and light connectionsin ventral extrastriate cortex, perhaps area VP (Newsomeet al., 1986). Visual-related inputs also originated fromthe parieto-occipital confluence (near LOP and PIP), andfrom the posterior STS (especially the MST complex).

Within somatotopically organized areas, labeled cellswere predominantly located in regions of estimated upperbody representations (hand, arm, and possibly the facerepresentation). This included zone 4C and the surround-ing motor-related cortex (Woolsey et al., 1952; Rizzolatti etal., 1981a,b; Gentilucci et al., 1988), the anterior half ofM2 (see Fig. 3M: F3 of Luppino et al., 1991, 1993), andlighter label in the posterior tip of 24d (Luppino et al.,

Fig. 1. Summary representation of topographically organized sen-sory and motor areas. Lateral (A) and medial (B) three-dimensional(3-D) views of a right hemisphere depicting 22 (filled white regions) ofthe 35 core architectonic outlines of subdivisions that are depicted onthe cortical flat map (C). The hemisphere is that of case B (cf. Fig. 4:also see Lewis and Van Essen, 2000 for map details). Symbols indicatethe estimated locations of visuotopic, tonotopic, somatotopic, and mo-tor maps based on architecture and geography. Visuotopic maps were

derived from Van Essen et al. (1990). Somatosensory and motor mapswere derived from Woolsey et al. (1952), Leinonen et al. (1979),Leinonen and Nyman (1979), Sessle and Wiesendanger (1982), Ponset al. (1985), Luppino et al. (1993), Iwamura et al. (1994), Burton andFabri (1995), Krubitzer et al. (1995), and Manger et al. (1996). Tono-topic map derived from Morel et al. (1993). For abbreviations in thisand subsequent figures, see list.


Fig. 2. Case A: Diamidino Yellow (DY) injection centered in medialportion of VIP (VIPm). A: Myelin stained section containing the DYcore, and showing the LIPv/VIP boundary. B: Drawing of coronalsection containing the DY injection core (green), depicting a superim-position of the core from nearby sections. Gray wedges depict transi-tion regions separating 5V, VIP, and LIPv. C,D: Lateral and medialviews of three-dimensional (3-D) layer 4 surface reconstruction show-ing supragranular labeled cells. E: Flat map showing injection core(dark green), halo (medium green), and track damage (light green).F: Flat map showing the distribution of retrogradely labeled cells

(supra- plus infragranular layers). Each dot corresponds to one la-beled neuron (19,397 cells plotted). White oval indicates injection corelocation. Physiological recording sites in area 2 (blue dots), and areas1 and 5 (red dots) depict digit representations (digits 2 and 3–4).Yellow corresponds to microlesions at the posterior boundary of phys-iologically defined area 2. Light gray outlines near 7a and V4 depicttissue compromised by skull opening. Refer to text and Lewis and VanEssen (2000; Fig. 14) for other map details. G: Label in IPS and STSof the contralateral hemisphere (1,809 cells), on a flat map oriented asright hemisphere to facilitate comparison.

Fig. 3. Case A: Distribution of cells projecting to VIP (VIPm:Diamidino Yellow [DY] injection) shown on coronal sections. Upperleft inset corresponds to Figure 2C, and shows the approximatelocation of 13 sections (A–M) plus the distribution of DY label. Solid

lines represent the pial surface and gray/white matter boundary, anddashed lines represent layer 4. Each dot corresponds to one retro-gradely labeled cell. Some areal boundaries are indicated by bars inthe white matter (others excluded for clarity). Scale bar 5 1 cm.

TABLE 2. Connections for Two VIP Injections (Cases A and B)1

1Strength of connections range from strong (111) to light (1), or absent (2). Laminar distribution of labeled cells were either predominantly supragranular (S), infragranular(I), or bilaminar (B). Bold text indicates consistent and newly described pathways.(***) represent sites of core tracer injection. See text for other details. For other abbreviations, see text.


Fig. 4. Case B: Diamidino Yellow (DY) injection centered in lateralportion of area VIP (VIPl). A: Nissl-stained section including theinjection core. B: Section outline containing the DY injection core.C,D: three-dimensional (3-D) maps showing distribution of supra-granular labeled cells. E: Flat map illustrating injection site. Lightgray outline near area 7b depicts compromised tissue. F: Flat map

showing the distribution of retrogradely labeled DY cells (21,326cells). In the STS: da, MSTda; and m, MSTm. In the arcuate sulcus: m,8Am; s, 8As. See text and Lewis and Van Essen (2000; Fig. 16) forother map details. G: Distribution of contralateral DY label near IPSand STS. The map is oriented as a right hemisphere. Other conven-tions as in Figure 2.


1991). Light label was present in two portions of area 2 (cf.Fig 3H–J). In this hemisphere, parts of the digit represen-tation of areas 1 and 2 were mapped physiologically aspart of another study (Lewis et al., 1999). The mappedregion indicated by blue, red, and yellow dots in Figure 3Eand F represent recordings of digits 2–4; thus, the DY celllabel ventral to this site is near the boundary of the firstdigit and face representation, and the labeled cells indorsal portions of area 2 are in the approximate location ofthe arm representation (cf. Pons et al., 1985). Light labelwas also present along the S2 complex (within and outsidethe core outline), though the somatotopy was unclear.

Strong projections to VIP originated from a number ofpolysensory related areas in the posterior half of thebrain. This included a nearly continuous swath of cortexinvolving areas MDP, MIP, VIP itself, 7b, and TPOc,which together are closely positioned in three-dimensionalspace, lying between visual and somatosensory cortex.

Table 2 summarizes the projections to VIP identified incase A, and also case B (see below). In case A, projectionsoriginated from 32 architectonic subdivisions (refer to Ma-

terials and Methods). We scored 4 projections as dense(111), 9 as moderate (11), 19 as light (1) or (1?), and 5as sparse (fewer than 10 cells) or uncertain (both indicatedas “?”). The bold type indicates pathways that have notpreviously been reported (or were uncertain from the lit-erature) yet were consistent across most other injectionsthat included VIP in this study (also see Table 3). Threecharacteristic laminar patterns of retrograde label weredistinguished, several of which are illustrated in Figure 3.An “S” designates labeled fields that were concentratedpredominantly in supragranular layers (.70%), charac-teristic of feed-forward projections (Rockland and Pandya,1979; Maunsell and Van Essen, 1983). These included PO(Fig. 3A,B) and area 2 (Fig. 3H–J), LOP (Fig. 3C), the VPregion, and possibly zone 7t. An “I” designates labeledfields predominantly in infragranular layers (.70%),characteristic of feed-back projections, which included M2(Fig. 3M), the S2 complex, 6Vam, 46p, 24a/b, and TPOr.Most of the labeled fields shown in Figures 2 and 3 had abilaminar distribution, denoted by a “B,” with 30–70% ofthe cells in superficial layers and the rest in deep layers,

Fig. 5. Case B: Distribution of cells projecting to VIP (VIPl: Diamidino Yellow [DY] injection) shownon coronal sections. Upper left inset corresponds to Figure 4C, and shows the approximate location of the9 sections (A–I). Scale bar 5 1 cm. Other conventions as in Figure 3.


which is ambiguous with respect to hierarchical relation-ship (Felleman and Van Essen, 1991). In some areas thelaminar distribution was uncertain (?), either due tosparse data or inconsistencies of the pattern inside versusjust outside the core region.

In the contralateral hemisphere of case A (Fig. 2G), thedistribution of DY label was heaviest in cortex geograph-ically and architectonically homotopic with the injectionsite, spanning VIPm and VIPl. The heterotopic labeledfields included MDP, 7a/7b, LIPv, LIPd, TPOc (and sparselabel in Toc), but none from early visual area PO, despiteits strong ipsilateral projection.

Case B. Figures 4 and 5 illustrate the pattern of labelresulting from a relatively large DY injection centered inlateral portions of VIP (VIPl). The core was centered nearthe fundus of the IPS in supragranular layers, as shown ina micrograph of a Nissl-stained section (Fig. 4A), on coro-nal section drawings (Figs. 4B and 5F,G), and on a flatmap of the IPS region (Fig. 4E). Though the injection corewas restricted to VIP, a portion of the halo was present inthe transition region between VIPl and LIPv, and to aminor degree along the injection track through LIPv.

The patterns of label for the VIP injection of case B waslargely similar to that of the VIP injection of case A. Forinstance, labeled fields common to both VIP injectionsincluded multiple visual (LIPv, LIPd, PO, LOP, MSTda,and MSTdp), polysensory (MIP, MDP, 7b, Toc, TPOc, andTPOi), somatosensory (5V, 7op, area 2), and premotor-related subdivisions (4C, M2, 24d, and 8Ac). However, theVIPl injection of case B also resulted in several light tomoderate projections from visual cortex (V2, PIP, V4, V4t,MT, TEa/m), some of which might have resulted fromtracer uptake in the VIPl/LIPv transition region. Addi-tional projections originated from sensorimotor-relatedareas 4, 3a, and the S2 complex, which were unlikely toresult from tracer uptake from LIPv.

From visuotopically organized cortex, the projections toVIPl largely originated from the lower peripheral visualfield representations, which contrasted with the uppervisual field input to VIP in case A. This included moderatelabel in lateral portions of area PO (Colby et al., 1988),light label in medial V2 and dorsal V4 (Gattass et al.,1981,1988). Light label was also present just ventral tothe architectonic core of MT in or near its central fieldrepresentation (Desimone and Ungerleider, 1986; Maun-sell and Van Essen, 1987). Moderate label was present in8Ac and 6Vam, which are subdivisions of the frontal eyefields (Schall et al., 1995).

Within somatotopically organized cortex, labeled cellswere predominantly located in regions of estimated upperbody representations as with the VIP injection of case A.This included zone 4C and the surrounding motor-relatedcortex plus area 4 (cf. Sessle and Wiesendanger, 1982;Godschalk et al., 1984), the anterior half of M2, and theposterior portion of 24d. The sparse to light label in areas3a, 1, and 2 was mainly in the estimated digit represen-tation (Pons et al., 1985). Label in the vicinity of the S2complex included a small patch in the posterior half of thecore region (cf. Fig. 5I), in the approximate location of thedigit and hand representation (Burton et al., 1995), plusscattered label anterior to the core. However, unlike theVIPm injection of case A, projections to the VIPl injectionsuggest a bias toward stronger label in the estimatedhand/arm representation of 24d, the S2 complex, and M2,and lighter label in the estimated face representation of

M2. This difference, together with the more robust differ-ences in upper versus lower visual field connections be-tween these two cases, might relate to the correlated vi-sual and somatosensory receptive fields of VIP cellsobserved physiologically (Duhamel et al., 1998).

Table 2 summarizes the projections and laminar distri-bution from 42 architectonic subdivisions to cortex in ornear VIP of case B (also see Fig. 5). The laminar organi-zation was largely consistent with the VIP injection ofcase A, even though the case B injection core was re-stricted to supragranular layers. Both VIP injections re-vealed a feed-forward pattern (S in Table 2) of label forareas 2 and PO, and common feed-back pattern (I) fromarea M2, the S2 complex, and possibly TPOr.

In the contralateral hemisphere (Fig. 4G), the distribu-tion of DY label in occipito-parietal cortex was similar tothat observed in the ipsilateral hemisphere, though lessdense overall. The heaviest label was in the anterior halfof VIPl, geographically homotopic to the injection site.Strong to moderate label was present in the anterior por-tions of VIPm, LIPv, and LIPd. Light to moderate labelwas also present in TPOc, MSTdp, MIP, PO, ventral area5, LIPd, the 7a/7b region and the S2 complex, plus MDPand TPOi/r (not shown).

Case C. Figure 6 illustrates a DY injection that waslocated along the fundus of the IPS, near the estimatedanterior border of VIP (and/or the transition region be-tween VIP and AIP). The resulting pattern of label wasmost similar to that shown for the VIPm injection of caseA (Fig. 2), but with a stronger emphasis on sensorimotorinputs than visual-related inputs. Labeled fields in visual-related cortex included light to moderate cell density inMSTda, MSTdp, MSTm, LIPd, LIPv, and polysensory areaMIP. Unlike the other VIP injection cases, label wassparse in MDP, LOP, and PO, and sparse or absent fromother early visual areas. In sensorimotor cortex, light tomoderate connections with somatotopically organized cor-tex was similar as in the previous two cases, includinglabel in the estimated upper body representations of area4, zone 4C, M2, and possibly the S2 complex, and theestimated arm and/or hand representation of 3a, 1, 2, and24d. Additionally labeled fields included four regions im-plicated in vestibular processing (cf. Fig. 14 of Akbarian etal., 1994). In particular, there was dense label in zone 7tnear the tip of the IPS (overlapping their 2v) and zone 4C(their 6pa), plus light label near the Ri/S2 transition re-gion (approximately their PIVC, also see Grusser et al.,1990) and area 3a (their 3aV). A similar pattern of con-nections with vestibular-related cortex was also evident inthe VIP injection of case B, and to a lesser extent in caseA. All three VIP injections also resulted in light label inand near caudal temporoparietal cortex (Toc, Tpt, andTPOc), a region implicated in auditory processing (seeDiscussion).

Connections of other parietal areasand the posterior STS

The ensuing four figures each illustrate two tracer in-jections, where the color identifies the tracer type (seeTable 1). Emphasis is placed on comparing labeling pat-terns with other injections, especially with area VIP.

Connections of LIPv and 7b (case B). Figure 7 illus-trates label resulting from a relatively small FB injection(blue) centered within area LIPv, located 2–3 mm poste-rior to the VIPl injection of case B (cf. Fig. 7F), and a


triplet of AG injections (red) in area 7b. In contrast to thepolymodal nature of VIP connections (cf. Fig. 4), the in-puts to neighboring area LIPv were mostly from visualrelated cortex, consistent with connections reported byBlatt et al. (1990). Label in topographically organizedvisual areas was predominantly located in lower periph-eral visual field representations (V2, V3, PO, V4, butthroughout most of MT). Other visual-related inputs werefrom ventral MIP, MSTdp, and MSTda (but not the inter-vening MSTm), plus light inputs from FST and IPa. Com-pared to the VIP inputs in this case, LIPv inputs werestronger with areas V3 and MT, and weaker with LOP andpolymodal areas TPOc and MIP. Very little FB label waspresent in somatosensory- and motor-related areas, withthe exception of moderate label in and around 8Ac (frontaleye fields) and sparse label in a few subdivisions such as5V and AIP.

The pattern of label resulting from the 7b injections inFigure 7 (red) was consistent with those reported in ear-lier studies (Stanton et al., 1977; Neal et al., 1986, 1987;Cavada and Goldman-Rakic, 1989b). Area 7b inputs orig-inated from sensorimotor and polymodal cortex, with littleinvolvement of visual-related cortex. There was little over-lap between label associated with the 7b injections (reddots) and the LIPv injection (blue dots) of this case, thougheach pattern overlapped markedly with inputs to the VIPinjection from the same case (cf. Figs. 4F and 7E).

Both 7b and VIP were strongly interconnected in thiscase, and both areas received moderate to strong overlap-ping inputs from MDP, LIPv, 7op, zone 4C, and otherportions of the arcuate sulcus (cf. Figs. 4F and 7E).Sparser overlap included estimated upper body represen-tations of area 2 and the S2 complex. Projections to 7b butnot VIP arose from 6Vb, PrCO, and the anterior extent ofIPa. Moderate AG label (7b inputs) overlapped sparse andscattered DY label (VIP inputs) in portions of Ig/Id andTPOr/TAa that were not found with other VIP injections.Thus, these inputs are likely to result from modest traceruptake at the initial penetration site for the DY injection(i.e., area 7b in Fig. 7F).

Overlap of all three tracers in case B included the an-terolateral bank of the IPS (AIP, LIP, and VIP), the dorsalhalf of the STS (TPOi and IPa), and cortex of or near thefrontal eye fields (8Ac, 6Vam). All three tracers were alsopresent in polysensory area TPOc, though the FB (LIPv)and AG (7b) labeled fields did not overlap. This is consis-tent with the clustering by modality reported physiologi-cally for area cSTP (Hikosaka et al., 1988), which overlapsmuch of architectonically defined area TPOc (Cusick et al.,1995).

Connections of LIPv and LIPd/7a (case D). Figure 8shows a FB injection core (blue) that was located withinLIPv, though the injection halo extended into layers 1 and2 of the transition region between LIPv and VIP. Thedistribution of label was similar to that of the LIPv injec-tion of case B, though the injection in case D produced agreater number of labeled cells (see figure legend). Labelwas present in the peripheral representation of a numberof visuotopically organized areas, including V2 and V3A aswell as the areas identified in the other LIPv injection(case B). The prominent projections from V2 implicate apathway (feed-forward pattern of label) to LIPv or theLIPv/VIP transition region. FB label was denser and moreextensive throughout area VIP in case D than in case B,perhaps owing to tissue damage from the DY injection of

VIP in case B. FB label in both cases was present in all ofthe MST complex, but was densest in zone MSTdp.

Figure 8 also shows results from a DY injection (green)that included areas LIPd plus cortex within the outline of7a and 7b, revealing a pattern of label largely consistentwith earlier studies (Mesulam et al., 1977; Stanton et al.,1977; Cavada and Goldman-Rakic, 1989a; Blatt et al.,1990; area PFG of Matelli et al., 1997). In contrast withthe LIPv injection (blue), which was strongly connectedwith V3, PIP, and PO, label from the DY injection wasnearly absent from these areas. Additionally, the DY in-jection resulted in moderate to strong inputs from V4, V4t,TEa/m, TPOr, and 45, which were weaker or absent forLIPv (and other injections of 7a and 7b). Together theseresults suggest that LIPv and LIPd are considerably dif-ferent in their connectivity.

Another notable difference in labeled fields in case D isthe segregation of DY (green) and FB (blue) label, whichmay relate to the organization of the saccade system. Forinstance, the DY injection included label largely from cen-tral visual field representations (i.e., V4) and ventral fron-tal eye fields (45, 6Vam), consistent with connectivity of asmall-amplitude saccade system, whereas the FB injec-tion (blue) included peripheral visual field inputs (i.e., V2)and dorsal frontal eye fields (8Ac, 8As), consistent with alarge-amplitude saccadic system (Bruce et al., 1985;Schall et al., 1995). Another result from the DY injection(green) was that it produced two large, separated labeledfields within the outline of areas TEa/m, which may cor-respond to the anatomically defined areas PITd and CITd(Van Essen et al., 1990; DeYoe et al., 1994).

Connections of AIP and MSTdp (case A). Figure 9illustrates connections of somatosensory area AIP (blue),and visual-related zone MSTdp (red). Projections to AIPwere predominantly from somatosensory and motor-related areas, directly overlapping some of the VIP pro-jection zones from this same case (cf. Fig. 2). The AIPinjection produced strong label within anterior and medialportions of the IPS and moderate label along the lateralsulcus and the posterior ridge of the arcuate sulcus. Incontrast to VIP injections, visual-related inputs to AIPwere only modest, including sparse to light inputs fromVIP, LIPd, and LIPv near the injection site, plus cortex inand near area MIP, perhaps overlapping area V6A ofGalletti et al. (1996a).

Projections to AIP from anterior portions of the hemi-sphere were concentrated along the posterior ridge ofthe arcuate sulcus (6Val, 4C, and 6Ds), which partiallyoverlapped the VIP projecting zones. VIP receivedstronger inputs from the dorsal arcuate region (e.g.,6Ds), whereas AIP received stronger inputs from ven-tral arcuate cortex (e.g., 6Val). This organization mayrelate to networks involved in arm reaching versusgrasping (see Discussion).

In somatosensory-related cortex, the heaviest label out-side of AIP was present in 5V, 7t, and 7b, with moderatelabel in 5D. The label in 7t and 7b overlapped the associ-ated face area described by Leinonen and Nyman (1979).In the lateral sulcus, moderate label was present along thedorsal bank of the S2 complex, which included estimatedupper body (digit and arm) representations (Burton et al.,1995). Curiously, contralateral projections from the S2complex to AIP (Fig. 9G) were absent despite its strongipsilateral projections.


Fig. 6. Case C: Diamidino Yellow (DY) injection centered near theanterior extent of VIP. A,B: Three-dimensional (3-D) maps showingdistribution of supragranular labeled cells. C: Section outline depict-ing the DY injection core. D: Flat map illustrating injection site. Lightgray outline in 7a, 2, and 5 depicts compromised tissue. E: Flat map

showing the distribution of retrogradely labeled DY cells (13,392cells). The architectonic boundaries of LIPv were divided into twoportions (see D), wherein the posterior half included a cortical bulge,whereas the anterior half did not. In the STS, i, TPOi. Other conven-tions as in Figures 2 and 4.


Fig. 7. Case B: Fast Blue (FB) injection (blue) centered in LIPvand three WGAapoHRP-Gold (AG) injections (red) centered in 7b. Thethree-dimensional (3-D) maps show the supragranular distribution ofFB- (A) and AG- (C) labeled cells, and section drawings depict the FB(B) and (D) injection cores. E: Flat map showing the distribution of

cells retrogradely labeled with FB (4,115 cells) and AG (10,899 cells).Blue (FB) occludes the red (AG) in regions of overlap. F: Flat mapillustrating FB, AG, and DY injection sites of this case. The FBinjection was centered 2–3 mm away from the DY injection core (cf.Fig. 4). Other conventions as in Figure 2.


Fig. 8. Case D: Fast Blue (FB) injection (blue) centered in LIPvand Diamidino Yellow (DY) injection (green) centered in LIPd/7a. Thethree-dimensional (3-D) maps show the supragranular distribution ofFB (A,B) and DY (C,D) retrograde cell label. E,F: Section outlinesdepicting FB and DY injection cores. Gray wedges in F depict borders

of LIPv, LIPd, and 7a. G: Distribution of FB- (13,542 cells) and DY-(19,986 cells) labeled cells on a flat map. Blue dots occlude green dotsin regions of overlap. Refer to Lewis and Van Essen (2000; their Fig.15) for map details. H: Flat map illustrating injection sites. Otherconventions as in Figure 2.


Fig. 9. Case A: Fast Blue (FB) injection (blue) centered in AIP, andWGAapoHRP-Gold (AG) injection (red) centered in MSTdp. A,C: Three-dimensional (3-D) maps showing supragranular FB and AG label,respectively. B,D: Respective drawings depicting FB and AG injectioncores. E: Flat map depicting the FB, AG, and DY (cf. Fig. 2) injections.

F: Flat map showing distribution of cells labeled with FB (11,330cells) and AG (30,704 cells). Blue dots occludes the red in regions ofoverlap. G: Distribution of contralateral cell label from FB (348 cells)and AG (1,152 cells) injections, oriented as a right hemisphere. Otherconventions as in Figure 2.


The AG injection of case A (Fig. 9; red) was centered inthe newly defined architectonic zone MSTdp. Label waspresent predominantly in visual-related regions of cortex,but was absent from all of area V1 (V1 not included onmap), indicating that there was no significant tracer up-take into area MT (Maunsell and Van Essen, 1983; Un-gerleider and Desimone, 1986a). Additionally, the patternof label was largely similar to results from MST injectionspreviously reported (cf. Boussaoud et al., 1990).

Projections to MSTdp included regions representingboth upper and lower fields of the visual periphery (V2,V3, V3A, dorsal V4, and the region near estimated ventralV4 and VP). Curiously, label in area PO was restricted toits lateral tip, indicative of inputs from its lower-field butnot upper-field representation (Colby et al., 1988). In theSTS, a notably strong projection from FST almost com-pletely filled the core extent of this area.

Projections from the IPS were largely restricted to areasVIP (including two clusters of label at the anterior andposterior ends of VIPm) and LIPv. The overall patterns ofMSTdp inputs and LIPv inputs were strikingly similar (cf.Fig. 9; red vs. Fig. 8; blue). One major difference was thepresence of strong inputs to MSTdp from ventral occipito-temporal regions (upper visual field representations), in-cluding VP and cortex near V4/VOT, and feed-back pro-jections from TF/TH.

Cell label from all three tracers of case A was inter-mixed in anterior portions of VIPm and LIPv, in the vi-cinity of cortex implicated in visual guidance of handmovements (Taira et al., 1990; Jeannerod et al., 1995),and sparse overlap was present along the posterior bankof the arcuate sulcus, and in 24b. An interdigitated pat-tern of label was evident in area MIP: projections to so-matosensory area AIP (blue) were concentrated along thedorsal core boundary of MIP (top portion on map); those tovisual area MSTdp (red) were concentrated in ventralportions; and those to bimodal area VIP (green) includedboth dorsal and ventral portions, consistent with the re-ported dorsoventral functional gradient within area MIP(Colby and Duhamel, 1991).

Connections of MT (case E). We were interested inanalyzing the pattern of MT connections in the IPS, giventhat area VIP was originally defined based on its connec-tions with area MT (Maunsell and Van Essen, 1983),whereas more recent studies reported that MT was alsoconnected with LIPv (Ungerleider and Desimone, 1986b;Blatt et al., 1990). Case E involved a pair of tracer injec-tions centered in area MT, as illustrated in Figure 10. TheFB injection was restricted to posterior portions of MT(peripheral visual field representation, mainly lowerfield); the nearby DY injection included an even moreperipheral representation in MT and probably encroachedinto zone MSTdp.

The global pattern of connections from both MT injec-tions was consistent with the studies mentioned above,including topographically organized projections from layer4B of V1, confirming that both injections included area MT(Maunsell and Van Essen, 1983). Topographic projectionsalso originated from areas V2, V3, and VP, and possiblyPO. Tracer uptake near the estimated boundary of MTand MSTdp (architecture was compromised by track dam-age) may account for the relatively stronger DY label inregions such as PO, cortex dorsal to MIP, and the presenceof DY label (but not FB label) in 7op, 24d, and zone Toc.

This adds to the evidence that MSTdp is a zone distinctfrom area MT.

Figure 10F and 10G shows an expanded view of the IPS,illustrating the distributions of DY- and FB-labeled cells,respectively. In LIPv, DY- and FB-labeled cells were in-termixed and distributed over most of its extent. A heavyDY-labeled patch was present in the posterior half ofVIPm, and a strong FB-labeled patch included VIPl butwas centered on the LIPv/VIPl transition zone. Beyondthis, MT projections to the IPS did not show any obvioustopographic relationship (including two other MT injec-tions, summarized in Table 3), though were suggestive ofa patchy topography. The consistent presence of label inLIPv and absence of significant label in LIPd adds to theevidence that these are distinct areas (see Discussion).

Summary of cortical connections

Results from the 11 injections illustrated above aresummarized in Table 3, along with results from six otherdeep IPS injections and two additional MT injections.These latter cases supported the main results illustratedabove and were useful for analyzing trends and consisten-cies in labeling patterns. This table is organized alonggeographic lines according to the injection sites (columns)and the labeled subdivisions (rows). The progression ofinjection sites starts with injections of anterior portions ofthe IPS (columns 1, 2, somatosensory dominated cortex),and is fofolfollowed by injections along midportions of thefundus of the IPS involving area VIP (columns 3–9,multimodal-dominated), injections of the lateral IPS andinferior parietal lobule (columns 10,11 involving LIPv,columns 12–14 involving 7a/7b), and injections within theSTS (columns 15–19 involving MT or MSTdp). The rowsinclude 63 architectonically distinct zones (highlighted),areas or areal complexes, and are arranged in 11 groups:dorsal and ventral occipital visual cortical areas or re-gions; areas of the STS; visual-related parietal cortex;somatosensory areas within the IPS and of parietal cortex;areas of the lateral sulcus; motor plus dorsal premotorareas; ventral premotor areas; cingulate cortex; and pre-frontal cortex.

The injection sites are depicted in each row by blackentries, with the estimated effective uptake strength forthe injected subdivision(s) indicated, ranging from strong(***) to light (*). Question marks in black boxes denotesites of possible tracer uptake (generally obfuscated bytissue damage along the pipet track). The remaining en-tries represent the strength of each projection, and rangein shading from dark gray (strong projection) to light gray(light projection). Projection strength is represented sym-bolically as well, for strong (111), moderate (11), andlight (1) connections (refer to Materials and Methods). Adistinction is also made between connections that weredemonstrably absent (-) and those that were not examinedin the tissue blocks available (blank entries).

In general, Table 3 provides a lower bound for theconnectivity associated with a given injection, takinginto account the sampling of sections, exclusion of somelabel lying outside core outlines, and exclusion of re-gions not analyzed due to compromised tissue. On theother hand, many injection sites involved more than onesubdivision, and are thus likely to overestimate theconnectivity of a single area or zone. Below, we summa-rize the connections of VIP, LIPv (vs LOP), AIP, and theSTS, taking into consideration all 19 tracer injections


(including reciprocal connections) and issues of uncer-tainty pertaining to tracer uptake into more than onecore region.

Connections of VIP. The significant and consistentcortical connections of area VIP (zones VIPm plus VIPl)

included at least 26 (possibly up to 35) architectonicallydistinct subdivisions (Table 3, columns 3–9), spanningseveral functional divisions (also see Fig. 11 in Discus-sion). Consistent connections of the VIP complex not pre-viously reported or well established included (roughly in

Fig. 10. Case E: Fast Blue (FB) injection (blue) centered in poste-rior MT, and Diamidino Yellow (DY) injection (green) in posterior MTthat encroached upon MSTdp. A,B: Three-dimensional (3-D) mapsshowing supragranular DY and FB cell label, respectively. C: Sectionoutline (cut orthogonal to the IPS) depicting the FB and DY injectioncores relative to MSTm, MSTdp, MT, and V4t. D: Flat map showingthe distribution of cells labeled with FB (11,230 cells) and DY (15,047

cells). Unlike previous maps, most of area V1 is included in the map,excluding tip of the occipital pole. E: Flat map depicting the injectionsites. The architectonic border at the MT/MSTdp transition was com-promised by the DY injection. F,G: Expanded view of the IPS showingthe distribution of DY- and FB-labeled cells, respectively. Other con-ventions as in Figure 2.


TABLE 3. Summary of Ipsilateral Projections for 19 Injections of Parietal Cortex and the STS1

1Relative strengths are indicated by shading and icons (see text). Injections restricted to one architectonic subdivision are indicated by densely bordered columns. Estimated sizeof injections from Table 1: S, small; M, medium; L, large. Vm, VIPm; Vl, VIPl; Lv, LIPv; Ld, LIPd. Note that some rows include two or more subdivisions, and some subdivisionsdepicted in the flat maps were excluded from this table.


order of robustness) subdivisions MIP, MDP, TPOc, 4C,MSTdp, TPOi, the S2 complex, 24d, 7t, LOP, Toc, M2, area4, and area 2. Connections with more variability includedareas 7op, PIP, TPOr, and 3a. Somatotopic inputs to VIPinvolved the upper, as opposed to lower, body representa-tion. This included the estimated digit, hand, or arm rep-resentation of the S2 complex, 24d, M2, areas 2, 1, andpossibly 3a, and the estimated face/head representation ofM2, and possibly the S2 complex, area 4, and 3a. Most ofthe injections involving VIP also resulted in sparse to lightlabel in cingulate area 24b.

Previous anatomical studies have identified areas thatreceive at least light projections from the fundus of theIPS near its middle one-third (i.e., near estimated areaVIP), further suggesting that most of the pathways weobserved are reciprocal. This included 12 visual-relatedareas, V2 (Seltzer and Pandya, 1980), V3 and VP (Felle-man et al., 1997), PO (Seltzer and Pandya, 1980; Colby etal., 1988), MT (Van Essen et al., 1980; Maunsell and VanEssen, 1983; Ungerleider and Desimone, 1986b), MSTda(DMZ), zone MSTm and possibly FST (Boussaoud et al.,1990), LIPd and LIPv (Blatt et al., 1990), MDP (7m) and7a (Stanton et al., 1977; Pandya and Seltzer, 1982; Seltzerand Pandya, 1986), polysensory area TPOc (TPO-4 of Selt-zer and Pandya, 1991); and eight sensorimotor-relatedareas, including 1 and 2 (area 1 hand representation,Burton and Fabri, 1995; arm representation of 1 and 2,Seltzer and Pandya, 1980; Pearson and Powell, 1985; alsosee Jones and Powell, 1969), zone 7t (Pandya and Seltzer,1982), 5V and 7b (Seltzer and Pandya, 1986; Neal et al.,1990), area 4 (Kunzle, 1978; Leichnetz, 1986), and premo-tor subdivisions 6Val and 4C (Godschalk et al., 1984).Whether all connections are indeed reciprocal andwhether there are pronounced or consistent differences inconnection strengths in opposite directions awaits furtherinvestigation using combined anterograde and retrogradetracer injections.

Connections of LIPv. To the list of previously re-ported LIPv connections (Blatt et al., 1990), we can addinputs from LOP, TPOc, and FST (Table 3, columns10,11). In some cases, additional sparse to moderateconnections with LIPv included V2, TPOi, 4C, M2, 24b,and 6Vam, but the consistency of these pathways awaitsfurther verification. Area V2 projections to the IPS wereonly observed with injections that included the VIPl/LIPv transition region (columns 6 –10), consistent witha study by Baizer et al. (1991). LIPv connections withvisuotopically organized areas predominantly involvedperipheral visual field representations (V2, V3, PO, andV4), with a strong bias towards lower peripheral visualfield inputs (compare top two clusters of rows in Table 3,columns 8 –11). However, this pattern might relate towhere the injections were located within LIPv ratherthan the area as a whole.

Area LIPv had strong, reciprocal connections with areaMT and zone MSTdp. In contrast, neighboring zone LOP,whose architecture is similar to LIPv (Lewis and VanEssen, 2000), sent few projections, if any, to the portions ofMT or MSTdp we injected (cf. Table 3, row “LOP” vs.“LIPv,” columns 15–19). Additionally, area LIPv and zoneLOP differed in their connections with the inferior pari-etal lobule (areas LIPd, 7a/7b) and area AIP.

Connections of AIP. In contrast with area VIP, con-nections of the anterior IPS (columns 1,2: mostly areaAIP) were largely devoid of visual-related inputs, with the

exception of areas VIP, LIP, and possibly MIP. Otherwise,injections of the anterior IPS showed a similar range ofsensorimotor connections to those of VIP, but with rela-tively stronger connections with zones 5V, 5D, 7t, 6Val,area 7b, and the S2 complex.

Connections of the STS. The connections of MT (Ta-ble 3, columns 16–19) were consistent with previous stud-ies (Maunsell and Van Essen, 1983; Ungerleider and Desi-mone, 1986b), but also included light projectionsoriginating from IPa in most cases. Of the three architec-tonic subdivisions of the IPS that projected to MT, connec-tions were strongest with LIPv, intermediate with VIPl,and sparsest with VIPm. The MT to LIPv pathway isnotable (and puzzling) in the view of physiological evi-dence reported, suggesting that direction selectivity isprominent in MT and VIP, but not in LIPv (Colby et al.,1993a,b). The MST complex was heavily interconnectedwith VIP and LIPv, consistent with earlier studies (Bous-saoud et al., 1990), and consistent with the similar re-sponse properties to optic flow for MSTdp and VIP(Schaafsma and Duysens, 1996).

DISCUSSION

Given the tremendous complexity of cortical circuitry, itis important to seek underlying principles of informationprocessing that are related to overall patterns of connec-tivity, rather than just the presence of one or anotherparticular pathway. The present study has provided infor-mation relevant to four broad issues. One issue is thedegree to which connectivity data can be used as a guide topartitioning the cortex. A second issue concerns the over-all degree of connectivity and possible regional differencesin the statistics of connectivity patterns. A third issueinvolves the concept of hierarchical organization (intra-and interhemispheric) based on laminar patterns of in-puts and outputs. A final issue regards the presence ofdistinct yet intertwined processing streams within sen-sory and motor systems. Our findings support the involve-ment of the VIP complex in several functional streams,including networks related to visual, vestibular, somato-motor, and auditory processing, and possibly a networkrelated to spatial attention.

Areal partitioning

Connectivity patterns provide an important basis foridentifying distinct cortical areas and determiningwhether neighboring regions belong to the same or differ-ent areas. Complexities such as tracer leakage and theinherent variability of connectivity patterns are routinelyobserved (cf. Perkel et al., 1986; Van Essen et al., 1986;Felleman and Van Essen, 1991), but despite these caveats,our data lend support for the distinctness of architectoni-cally defined area VIP, a distinction between areas LIPv,LIPd, and zone LOP, and between zone MSTdp and areaMT.

Area VIP. The pattern of inputs to architectonically de-fined area VIP differed markedly from those of neighboringareas LIPv and AIP. Additionally, injections placed in visualsubdivisions MT, MSTdp, and LIPv consistently producedmoderate to strong label in area VIP with little or no label inneighboring subdivisions 5V and AIP, thereby respecting themedial and anterior architectonic boundaries of VIP. To-gether, these results strongly support the architectonic (andphysiological) evidence summarized in the preceding study


(Lewis and Van Essen, 2000) for the distinctness of area VIPfrom surrounding areas.

The results of the present study, along with the com-panion study, raise but do not resolve the issue of whetherthe medial and lateral architectonic zones within VIP(VIPm and VIPl) are sufficiently distinct to be consideredseparate areas. VIPl tends to have stronger connectionswith visual-related areas, whereas VIPm is more stronglyconnected with sensorimotor-related areas. However,given the small number of injections and tracer leakageissues, the significance of the heterogeneity of VIP willneed to be resolved by future studies.

LIPv, LIPd, and zone LOP. LIPv has several connec-tions with motion processing pathways, including MT,whereas LIPd is connected more with “featural” processingpathways in inferotemporal cortex. This pattern is consis-tent with LIPv representing peripheral visual fields andLIPd representing central visual fields as reported by Blattet al. (1990) and Ben Hamed et al. (1999). It is also consistentwith physiological distinctions noted by Ben Hamed et al.(1999) and supports our overall conclusion that LIPd andLIPv warrant classification as distinct areas.

In the companion article, a distinction was made be-tween area LIPv and neighboring zone LOP, though thearchitectonic differences are modest (Lewis and Van Es-sen, 2000). Connectivity patterns supported this distinc-tion, in that injections of MT and MSTdp consistentlyproduced moderate to strong label in LIPv, but little or nolabel in zone LOP. Schall et al. (1995) reported connec-tions between the frontal eye fields (FEF) and LIPv, whichincluded what we consider to be LIPv plus zone LOP. Theyreported differences in connections between rostral andcaudal portions of their boundaries for LIPv, and showedexamples of labeled fields that respected or were discon-tinuous near the estimated boundary between LOP andLIPv of the present study. Finally, Sakata et al. (1997)have reported a high concentration of cells selective for3-D surface orientation in the caudal intraparietal sulcus(their “cIPS”), a region that appears to overlap substan-tially with LOP. Although the aggregate evidence is sug-gestive, we do not feel that it is strong enough to designateLOP unequivocally as a distinct area.

MST complex. The companion article identified archi-tectonic distinctions between area MT and zone MSTdpthat are strongly supported by the present connectivitydata. For instance, an injection of MSTdp (Fig. 9, red)resulted in a wide range of dense label in early visualcortex comparable to MT (Fig. 10), with the notable excep-tion of area V1 (Van Essen et al., 1981; Ungerleider andDesimone, 1986a). MSTdp appears to overlap the “periph-eral V2-recipient zone” adjacent to MT described by Un-gerleider and Desimone (1986a; their Fig. 11), and par-tially overlaps physiologically described MSTd (Komatsuand Wurtz, 1988).

Degrees of connectivity

A striking aspect of connectivity patterns in our data setis the wide range in the total number of architectonicsubdivisions labeled after different injections (see Table 3,bottom row). Even after accounting for various tracer up-take issues (i.e., leakage along the pipet track, injectionsize and spread into neighboring areas), injections intoarea VIP consistently revealed more labeled cortical path-ways than did the injections of largely unimodal areasLIPv, AIP, and MT. An intriguing possibility is that mul-

timodal areas may have an intrinsically greater degree ofconnectivity in terms of number of subdivisions (thoughthis would not imply that the total density of inputs andoutputs differs across areas). The fact that VIP lies alonga fundus and is widely interconnected relative to cortexalong the banks of the IPS appears consistent with twoearlier hypotheses regarding the relationship betweencortical function and geography. One is a tension-basedtheory of morphogenesis (Van Essen, 1997), which pre-dicts that mechanical tension generated along axons ofcorticocortical pathways is responsible for determiningcortical folding patterns. Specifically, VIP is heavily con-nected with both posterolateral and anteromedial portionsof the hemisphere (cf. yellow and blue in Fig. 11), and itsposition along a fundus makes it well situated geograph-ically (in 3-D space) to have relatively low aggregate ax-onal connection lengths with both regions. Another hy-pothesis is that of “fundal cognition,” which proposes thatthe fundi of the cortex, serving as “hubs of cross-corticaltraffic,” might be particularly suitable for higher cognitiveprocessing, including multimodal integration (Markow-itsch and Tulving, 1994).

Hierarchical relationships

By using similar criteria for hierarchical placement, ourresults largely support the placement of VIP at the sev-enth level of the cortical hierarchy proposed by Van Essenet al. (1990; also see Felleman and Van Essen, 1991), andthey are also consistent with the various alternate hier-archical schemes reported by Hilgetag et al. (1996). How-ever, the S2 complex (both ipsilateral and contralateral)consistently showed a feed-back pattern of connectionwith VIP, suggesting that it is higher than VIP ratherthan lower in the hierarchy. This is at odds with theplacement of S2 by Van Essen et al. (1990) but is consis-tent with a more recent somatosensory hierarchicalscheme (H. Burton, personal communication; also see Bur-ton et al., 1996).

Interhemispheric connections

Interhemispheric connections were generally strongestin regions that were geographically and architectonicallyhomotopic to each injection site, consistent with earlierstudies of parietal cortex (Hedreen and Yin, 1981; Cavadaand Goldman-Rakic, 1989a,b; for review see Kennedy etal., 1990). However, in other species, the densest callosalconnections can occur in anatomically heterotopic portionsof the corresponding areas in the contralateral hemi-sphere (rat, Lewis and Olavarria, 1995; cat, Shatz, 1977).Thus, we did not use callosal connection patterns to inferthe precise location of injection sites. As expected, theheterotopic patterns of contralateral label generally in-volved a subset of the labeled regions found in the ipsilat-eral hemisphere, though with considerably lower densityof labeled cells (Pandya and Vignolo, 1969; Andersen etal., 1985a; Cavada and Goldman-Rakic, 1989a,b).

Functional streams involving VIP

Physiological studies suggest that area VIP is involvedin a variety of functions, including multisensory analysisof stimulus location and motion; the construction of amultisensory, head-centered representation of near ex-trapersonal space; and perhaps eye/hand and mouth coor-dination (Colby et al., 1993a; Duhamel et al., 1997, 1998).Consistent with this diversity of functions, the wide range


of connections observed in the present study implicatesVIP in multiple cortical networks that are known or sus-pected to have distinct functional specializations. Theseare summarized in Figure 11, by using an expanded por-tion of the cortical map for case B, but drawing on resultsacross several cases. Colored regions represent consistentprojections to area VIP observed in the present study,with each color corresponding to a different proposed func-tional network. Filled areas represent relatively strongerand/or more consistent connections, whereas ovals andcolored text represent lighter connections. The visual spa-

tial analysis network (yellow) and related oculomotor net-work (orange), are largely consistent with earlier studiesdescribing the “dorsal stream” pathway pertaining to vi-sual spatial analysis (Ungerleider and Mishkin, 1982;Andersen, 1987; Felleman and Van Essen, 1991). Polysen-sory areas (visual and somatosensory response propertiesor inputs) are colored green; these include several of thestrongest and most consistent connections of VIP. In ad-dition, there is evidence for a somatosensory- and motor-related network involving the upper body representations(blue), a vestibular-related network (purple), an auditory-

Fig. 11. Schematic summary of proposed cortical networks involv-ing area VIP. Sulci are labeled and colored gray (buried cortex), and aselect set of core architectonic areas (from case B) are indicated.Colored areas, ovals, and text depict regions projecting to VIP, and arecoded according to presumed function (see Figure key and text).Multiple colors in LOP, 7b, TPO, 4C depict possible involvement inmultiple networks. LOP was partially included in the oculomotornetwork based on a connectivity study by Schall et al. (1995). Hatched

oval in area 24b may be related to a spatial attention network (seetext). Bimodal or polymodal areas colored green include 4C (Rizzolattiet al., 1981a,b; Fogassi et al., 1996), MIP (Colby and Duhamel, 1991),MDP (PEc/PGm of Pandya and Seltzer, 1982), 7b (Leinonen et al.,1979; Hyvarinen, 1981; Dong et al., 1994), and TPOc (Bruce et al.,1981). S2cx, S2 complex (cf. Burton et al., 1995; Krubitzer et al.,1995). Refer to text for other details.


related network (red), and possibly an attention-relatednetwork, each of which merits brief discussion.

Topography of VIP. Inputs to VIP from visuotopi-cally organized areas generally involved lower field, and toa lesser extent upper field representations of the visualperiphery. However, across the core extent of VIP therewas no obvious visuotopy, consistent with physiologicalobservations (Colby and Duhamel, 1991; Duhamel et al.,1998). Inputs from somatotopically organized areas gen-erally originate from the upper body (digits, arm, and face)representations, but without any obvious somatotopic or-ganization in VIP. Interestingly, VIP connections withsome subdivisions (e.g., MT, MSTdp, area 4, TPOr, 5D,and zone 7t) were moderate or strong in some cases,whereas sparse or entirely absent in others. This variabil-ity was not easily accounted for by technical artifacts,such as errors in parcellation or tracer leakage. Rather,the evidence is suggestive of either individual variabilityin the presence of major cortical projections, or a patchytopographic organization for VIP with regard to visual andsomatomotor connections.

Somatosensory inputs to VIP. The majority of VIPcells have tactile receptive fields representing the headand face (where the mouth represents the tactile “fovea”),whereas a minority of cells represent other portions of theupper body including the arms and hands (Colby andDuhamel, 1991; Duhamel et al., 1998; Colby and Gold-berg, 1999). Consistent with this physiology, VIP is con-nected with the estimated upper body representation ofseveral areas that have at least a crude degree of soma-totopic organization (blue in Fig. 11). However, an unre-solved issue concerns the origins of predominant face-related somatosensory inputs to VIP. Projections fromsomatosensory areas 2, 1, and possibly 3a were mainlyfrom the digit or digit/face boundary representations (de-termined physiologically for area 2 in one case; see Lewiset al., 1999), but notably absent from the large expanse ofestimated face and head representations (Pons et al.,1985). Candidate regions more likely to convey face-related inputs to VIP include: (1) postarcuate cortex (nearzone 4C) which has similar response properties to those ofarea VIP (Rizzolatti et al., 1981a,b; Gentilucci et al.,1988); (2) areas 5 and 7b, though they have only a coarsesomatotopic organization (Mountcastle et al., 1975;Leinonen et al., 1979; Leinonen and Nyman, 1979; Robin-son and Burton, 1980; Hyvarinen, 1981; Rizzolatti et al.,1981a; Gentilucci et al., 1988; Fogassi et al., 1996); and (3)M2 and S2, though these were relatively light and pre-dominantly of a feed-back pattern of connection.

Inputs to VIP from the estimated hand and arm repre-sentations of somatotopically organized areas is consis-tent with its proposed involvement in a network of corticalareas that subserve reaching (as opposed to grasping)movements with the limbs (Tanne et al., 1995, 1996; alsosee Jeannerod et al., 1995). This network includes dorsalpremotor cortex (overlapping 6Ds and dorsal 4C of thisstudy), and visual areas PO, MDP (cf. V6A of Galletti etal., 1996b), caudal MIP, and LIP (cf. Taira et al., 990),which were all interconnected with VIP in the presentstudy. These connections are consistent with VIP’s pro-posed involvement in eye-hand and mouth coordination(Colby et al., 1993a). The grasping-related network, in-cluding ventral premotor cortex 6Val, lateral portions ofzone 4C, plus AIP, 7b, and area 5, fits with the results ofour AIP injections. However, there is a considerable de-

gree of cross-talk between these two hypothesized func-tional networks.

A vestibular network. Our anatomical data supportprevious physiological evidence that VIP is involved inprocessing vestibular information (Graf et al., 1996;Bremmer et al., 1997). Specifically, area VIP (especiallyanterior and medial portions) receives inputs from severalcortical areas (purple in Fig. 11) that also have directconnections with vestibular nuclei (cf. Akbarian et al.,1994; Faugier-Grimaud et al., 1997). These include mod-erate connections with zone 7t, near to or overlapping area2v, where vestibular responsiveness has been reportedphysiologically (Schwarz and Fredrickson, 1971; Buttnerand Buettner, 1978; but see Grusser et al., 1990), and zone4C (6pa of Akbarian et al., 1994). Lighter and more vari-able vestibular-related connections included: area 3a(overlapping 3aV of Akbarian et al., 1994); cortex near theretro-insular area Ri (overlapping PIVC of Grusser et al.,1990), which is a region that is responsive to vestibular,somatosensory, and visual stimulation; and possibly theMST complex (Faugier-Grimaud et al., 1997). Interest-ingly, all the above-mentioned vestibular-related corticalregions connected with VIP are characterized by moderateto high SMI-32 immunoreactivity (Lewis and Van Essen,2000). The vestibular responses of VIP may be involved inthe analysis of self-motion, and/or modifying the referenceframe in which visual and other sensory inputs are rep-resented (Bremmer et al., 1997; Faugier-Grimaud et al.,1997).

Auditory inputs to the IPS. Auditory informationmay reach the IPS via light projections (red in Fig. 11) toVIP from presumed auditory-related zone Toc (overlap-ping area C of Morel et al., 1993) and area Tpt (Leinonenet al., 1980; Hikosaka et al., 1988). In addition, VIP andLIPv receive inputs from part of the polysensory TPOcomplex (Mistlin and Perrett, 1990), and possibly the au-ditory belt region CM (for review see Kaas et al., 1999).Auditory inputs to LIP may account for cells reported toencode the remembered spatial location of auditory stim-uli as well as the less frequent cells that are directlyresponsive to auditory stimuli (Mazzoni, 1994; Linden etal., 1996; also see Andersen, 1995). Physiological tests forauditory inputs to VIP have yet to be examined system-atically. If present, they might convey the type of memory-related responsiveness reported for LIP. Alternatively,they might involve polysensory responsiveness to near-field auditory stimuli that are congruent with somatosen-sory and/or visual receptive fields, as has been reported forcells in the ventral premotor cortex (Graziano et al., 1999).Furthermore, cells of the VIP complex may be involved inthe analysis of auditory motion, consistent with visualmotion and tactile response properties of VIP (cf. Duhamelet al., 1998), and consistent with the convergence of audi-tory motion and visual motion systems in and near theIPS of humans (Lewis et al., 2000).

An attention network. The control of spatial atten-tion in human and nonhuman primates has been proposedto be coordinated by a large-scale network including pari-etal cortex, the frontal eye fields plus surrounding cortex,and cingulate cortex (Mesulam, 1981; Rizzolatti et al.,1987; Morecraft et al., 1993). Consistent with this model,all injections involving VIP, LIP, and 7a revealed connec-tions with the frontal eye field region (8Ac, plus portions of45 and 6Vam), consistent with earlier studies (Andersenet al., 1985a; Cavada and Goldman-Rakic, 1989b). Addi-


tionally, a small portion of cingulate area 24b (hatchedoval in Fig. 11) showed reasonably consistent projections(though relatively light) to a wide variety of parietal (andtemporal) areas, including polysensory areas VIP and 7b,somatosensory-related area AIP, and visual-related areas,LIPv, LIPd, 7a, and zone MSTdp. These pathways may beimportant for coordinating or directing attention towardmotivationally relevant events (Mesulam et al., 1977).

Towards probabilistic maps of corticalconnectivity

In the preceding study (Lewis and Van Essen, 2000), wediscussed the importance of establishing probabilisticsurface-based atlases, in which each location in the cortexcan be assigned a likelihood of belonging to one or anothercortical subdivision according to any of several partition-ing schemes. A natural extension of this approach is todevelop a database of cortical connectivity, which wouldprovide probabilistic assessments of the distribution andstrength of the connections associated with each locationon the cortical surface. The present study represents onestep in this direction, as it includes an archival repositoryof quantitative data on connection patterns that are rep-resented precisely in relation to architecture and the cor-tical surface. A variety of technical hurdles must be over-come in order to bring data into a common framework thatwill allow systematic analysis of complex connectivity pat-terns derived from numerous individual cases using avariety of pathway-tracing techniques. Nonetheless, givenrecent progress in computational neuroanatomy (cf. VanEssen et al., 1998), the time is ripe for developing suchmethods. Their advent will greatly enhance the utility andease of access of connectivity information that is neededfor a variety of purposes, including formulating models ofcortical organization and function that respect the under-lying cortical anatomy.

ACKNOWLEDGMENTS

We thank Ms. Heather Drury for making possible thecomputational flattening of cortex, Dr. Harold Burton foradvice and technical assistance with the physiologicalmapping of somatosensory cortex, Dr. Thomas Coogan forsurgical assistance, and Mrs. Shashi Kumar for histolog-ical assistance. This work was supported by NIH grant EY02091 to D.V.E.

LITERATURE CITEDAkbarian S, Grusser O-J, Guldin WO. 1994. Corticofugal connections be-

tween the cerebral cortex and brainstem vestibular nuclei in the ma-caque monkey. J Comp Neurol 339:421–437.

Andersen RA. 1987. Inferior parietal lobule function in spatial perceptionand visuomotor integration. In: Mountcastle VB, Plum F, Geiger SR,editors. Handbook of physiology. The nervous system. Higher functionsof the brain. Bethesda, MD: American Physiological Society, sect 1, volV, chapt 12. p 483–518.

Andersen RA. 1995. Encoding of intention and spatial location in theposterior parietal cortex. Cereb Cortex 5:457–469.

Andersen RA, Asanuma C, Cowan M. 1985a. Callosal and prefrontal as-sociated projection cell populations in area 7a of the macaque monkey:a study using retrogradely transported fluorescent dyes. J Comp Neu-rol 232:443–455.

Andersen RA, Essick GK, Siegel RM. 1985b. Encoding of spatial location byposterior parietal neurons. Science 230:456–458.

Andersen RA, Asanuma C, Essick G, Siegel RM. 1990. Corticocortical

connections of anatomically and physiologically defined subdivisionswithin the inferior parietal lobule. J Comp Neurol 296:65–113.

Baizer JS, Ungerleider LG, Desimone R. 1991. Organization of visualinputs to the inferior temporal and posterior parietal cortex in ma-caques. J Neurosci 11:168–190.

Basbaum AI, Menetrey D. 1987. Wheat germ agglutinin-apoHRP gold: anew retrograde tracer for light- and electron-microscopic single- anddouble-label studies. J Comp Neurol 261:306–318.

Ben Hamed S, Bihouee A, Deneve S, Duhamel J-R. 1999. Neural correlatesof visual fixation in the dorsal lateral intraparietal area of the macaque(LIPd). Soc Neurosci Abstr 25:1163.

Blatt GJ, Andersen RA, Stoner GR. 1990. Visual receptive field organiza-tion and cortico-cortical connections of the lateral intraparietal are(area LIP) in the macaque. J Comp Neurol 299:421–445.

Boussaoud D, Ungerleider LG, Desimone R. 1990. Pathways for motionanalysis: Cortical connections of the medial superior temporal andfundus of the superior temporal visual areas in the macaque. J CompNeurol 296:462–495.

Bremmer F, Duhamel J-R, Ben Hamed S, Graf W. 1997. The representa-tion of movement in near extra-personal space in the macaque ventralintraparietal area (VIP). In: Their P, Karnath H-O, editors. Parietallobe contributions to orientation in 3D space. Heidelberg: Springer-Verlag. p 619–631.

Bruce C, Desimone R, Gross CG. 1981. Visual properties of neurons in apolysensory area in superior temporal sulcus of the macaque. J Neu-rophys 53:603–635

Bruce CJ, Goldberg ME, Bushnell MC, Stanton GB. 1985. Primate frontaleye fields. II. Physiological and anatomical correlates of electricallyevoked eye movements. J Neurophys 54:714–734.

Burton H, Fabri M. 1995. Ipsilateral intracortical connections of physio-logically defined cutaneous representations in areas 3b and 1 of ma-caque monkeys: projections in the vicinity of the central sulcus. J CompNeurol 335:508–538.

Burton H, Fabri M, Alloway K. 1995. Cortical areas within the lateralsulcus connected to cutaneous representations in areas 3b and 1: arevised interpretation of the second somatosensory area in macaquemonkeys. J Comp Neurol 355:539–562.

Buttner U, Buettner UW. 1978. Parietal cortex (2v) neuronal activity in thealert monkey during natural vestibular and optokinetic stimulation.Brain Res 153:392–397.

Cavada C, Goldman-Rakic PS. 1989a. Posterior parietal cortex in rhesusmonkey: I. parcellation of areas based on distinctive limbic and sensorycorticocortical connections. J Comp Neurol 287:393–421.

Cavada C, Goldman-Rakic PS. 1989b. Posterior parietal cortex in rhesusmonkey: II. evidence for segregated corticocortical networks linkingsensory and limbic areas with the frontal lobe. J Comp Neurol 287:422–445.

Colby CL, Duhamel J-R. 1991. Heterogeneity of extrastriate visual areasand multiple parietal areas in the macaque monkey. Neuropsychology29:517–37.

Colby CL, Goldberg ME. 1999. Space and attention in parietal cortex. AnnuRev Neurosci 22:319–349.

Colby_2 CL, Gattass R, Olson CR, Gross CG. 1988. Topographical organi-zation of cortical afferents to extrastriate visual area PO in the ma-caque: a dual tracer study. J Comp Neurol 269:392–413.

Colby CL, Duhamel J-R, Goldberg E. 1993a. Ventral intraparietal area ofthe macaque: Anatomic location and visual response properties. J Neu-rophysiol 69:902–914.

Colby CL, Duhamel J-R, Goldberg E. 1993b. The analysis of visual space bythe lateral intraparietal area of the monkey: the role of extraretinalsignals. Prog Brain Res 95:307–316.

Cusick CG, Seltzer B, Cola M, Griggs E. 1995. Chemoarchitectonics andcorticocortical terminations within the superior temporal sulcus of therhesus monkey: evidence for subdivisions of superior temporal polysen-sory cortex. J Comp Neurol 360:513–535.

Desimone R, Ungerleider LG. 1986. Multiple visual areas in the caudalsuperior temporal sulcus of the macaque. J Comp Neurol 248:164–189.

DeYoe EA, Felleman DJ, Van Essen DC, McClendon E. 1994. Anatomicalbasis of concurrent processing streams in visual area V4 and infero-temporal cortex of the macaque monkey. Nature 371:151–154.

Dong WK, Chudler EH, Sugiyama K, Roberts VJ, Hayashi T. 1994. So-matosensory, multisensory, and task-related neurons in cortical area7b (PF) of unanesthetized monkeys. J Neurophys 72:542–564.

Duhamel J-R, Bremmer F, Ben Hamed S, Graf W. 1997. Spatial invariance


of visual receptive fields in parietal cortex neurons. Nature 389:845–848.

Duhamel J-R, Colby CL, Goldberg ME. 1998. Ventral intraparietal area ofthe macaque: Convergent visual and somatic response properties.J Neurophysiol 79:126–136.

Faugier-Grimaud S, Baleydier C, Magnin M, Jeannerod M. 1997. Directcortical projections to the vestibular complex in macaque monkey. In:Their P, Karnath H-O, editors. Parietal lobe contributions to orienta-tion in 3D space. Heidelberg: Springer-Verlag. p 57–75.

Felleman DJ, Van Essen DC. 1991. Distributed hierarchical processing inthe primate cerebral cortex. Cereb Cortex 1:1–47.

Felleman DJ, Burkhalter A, Van Essen DC. 1997. Cortical connections ofareas V3 and VP of macaque monkey extrastriate visual cortex. J CompNeurol 379:21–47.

Fogassi L, Gallese V, Fadiga L, Luppino G, Matelli M, Rizzolati G. 1996.Coding of peripersonal space in inferior premotor cortex (area F4).J Neurophysiol 76:141–157.

Galletti C, Fattori P, Battaglini PP, Shipp S, Zeki S. 1996a. Functionaldemarcation of a border between area V6 and V6A in the superiorparietal gyrus of the macaque monkey. Eur J Neurosci 8:30–52.

Galletti C, Fattori P, Kutz DF, Battaglini PP. 1996b. Arm movement-related neurons in the visual area V6A of monkey superior parietallobule. Soc Neurosci Abstr 22:2026

Gattass R, Gross CG, Sandell JH. 1981. Visual topography of V2 in themacaque. J Comp Neurol 201:519–539.

Gattass R, Sousa APB, Gross CG. 1988. Visuotopic organization and extentof V3 and V4 of the macaque. J Neurosci 8:1831–1845.

Gentilucci M, Fogassi L, Luppino G, Matelli M, Camarda R, Rizzolatti G.1988. Functional organization of inferior area 6 in the macaque mon-key. Exp Brain Res 71:475–490.

Godschalk M, Lemon RN, Kuypers HGJM, Ronday HK. 1984. Corticalafferents and efferents of monkey postarcuate area: an anatomical andelectrophysiological study. Exp Brain Res 56:410–424.

Graf W, Bremmer F, Ben Hamed S, Duhamel J-R. 1996. Visual-vestibularinteraction in the ventral intraparietal area (VIP) of macaque mon-keys. Soc Neurosci Abstr 22:1692.

Graziano MSA, Reiss LAJ, Gross CG. 1999. A neural representation of thelocation of nearby sounds. Nature 397:428–430.

Grusser O-J, Pause M, Schreiter U. 1990. Localization and responses ofneurones in the parieto-insular vestibular cortex of awake monkeys(macaca fascicularis). J Physiol 430:537–557.

Hedreen JC, Yin TCT. 1981. Homotopic and heterotopic callosal afferentsof caudal inferior parietal lobule in macaca mulatta. J Comp Neurol197:605–621.

Hikosaka K, Iwai E, Saito H-A, Tanaka K. 1988. Polysensory properties ofneurons in the anterior bank of the caudal superior temporal sulcus ofthe macaque monkey. J Neurophysiol 60:1615–1637.

Hilgetag C-C, O’Neill MA, Young MP. 1996. Indeterminate organization ofthe visual system. Science 271:776–777.

Hyvarinen J. 1981. Regional distribution of functions in parietal associa-tion area 7 of the monkey. Brain Res 206:287–303.

Hyvarinen J. 1982. Posterior parietal lobe of the primate brain. PhysiolRev 62:1060–1129.

Iwamura Y, Tanaka M, Sakamoto M, Hikosaka O. 1994. Rostrocaudalgradients in the neuronal receptive field complexity in the finger regionof the alert monkey’s postcentral gyrus. Exp Brain Res 92:360–368.

Jeannerod M, Arbib MA, Rizzolatti G, Sakata H. 1995. Grasping objects:the cortical mechanisms of visuomotor transformation. Trends Neuro-sci 18:314–320.

Jones EG, Powell TPS. 1969. Connexions of the somatic sensory cortex ofthe rhesus monkey. I. Ipsilateral cortical connexions. Brain 92:477–502

Jones EG, Powell TPS. 1970. An anatomical study of converging sensorypathways within the cerebral cortex of the monkey. Brain 93:793–820.

Jones EG, Coulter JD, Hendry SHC. 1978. Intracortical connectivity ofarchitectonic fields in the somatic sensory, motor and parietal cortex ofmonkeys. J Comp Neurol 181:291–348.

Kaas JH, Hackett TA, Tramo MJ. 1999. Auditory processing in primatecerebral cortex. Curr Opin Neurobiol 9:164–170.

Kennedy H, Meissirel C, Dehay C. 1990. Callosal pathways and theircompliancy to general rules governing the organization of corticocorti-cal connectivity. In: Dreher B, Robinson S, editors. Vision and visualdysfunction: vol 3, neuroanatomy of the visual pathways and theirdevelopment. London: MacMillan. p 324–359.

Komatsu H, Wurtz RH. 1988. Relation of cortical areas MT and MST to

pursuit eye movements. I. Localization and visual properties of neu-rons. J Neurophys 60:580–603.

Krubitzer L, Clarey J, Tweedale R, Elston G, Calford M. 1995. A redefini-tion of somatosensory areas in the lateral sulcus of macaque monkeys.J Neurosci 15:3821–3839.

Kunzle H. 1978. Cortico-cortical efferents of primary motor and somato-sensory regions of the cerebral cortex in macaca fascicularis. Neuro-science 3:25–39.

Leichnetz GR. 1986. Afferent and efferent connections of the dorsolateralprecentral gyrus (area 4, hand/arm region) in the macaque monkey,with comparisons to area 8. J Comp Neurol 254:460–492.

Leinonen L, Nyman G. 1979. II. Functional properties of cells in antero-lateral part of area 7 associative face area of awake monkeys. ExpBrain Res 34:321–333.

Leinonen L, Hyvarinen J, Sovijarvi ARA.1980. Functional properties ofneurons in the temporo-parietal association cortex of awake monkey.Exp Brain Res 39:203–215.

Lewis JW, Olavarria JF. 1995. Two rules for callosal connectivity in striatecortex of the rat. J Comp Neurol 361:119–137.

Lewis JW, Van Essen DC. 1994. Connections of area VIP with area MIPand other architectonically identified areas of the intraparietal sulcusin macaque monkeys. Soc Neurosci Abstr 20:724.

Lewis JW, Van Essen DC. 1996. Connections of visual area VIP withsomatosensory and motor areas of the macaque monkey. Soc NeurosciAbstr 22:398.

Lewis JW, Van Essen DC. 2000. Architectonic parcellation of parieto-occipital cortex and interconnected cortical regions in the macaquemonkey. J Comp Neurol 428:79–111.

Lewis JW, Burton H, Van Essen DC. 1999. Anatomical evidence for theposterior boundary of area 2 in the macaque monkey. SomatosensMotor Res 16:382–390.

Lewis JW, Beauchamp M, DeYoe EA. 2000. A comparison of visual andauditory motion processing in human cerebral cortex. Cereb Cortex10:873–888.

Linden JF, Grunewald A, Andersen RA. 1996. Auditory sensory responsesin area LIP? Soc. Neurosci Abstr 22:1062.

Luppino G, Matelli M, Camarda RM, Gallese V, Rizzolatti G. 1991. Mul-tiple representations of body movements in mesial area 6 and theadjacent cingulate cortex: an intracortical microstimulation study inthe macaque monkey. J Comp Neurol 311:463–482.

Luppino G, Matelli M, Camarda R, Rizzolatti G. 1993. Corticocorticalconnections of area F3 (SMA-proper) and area F6 (pre-SMA) in themacaque monkey. J Comp Neurol 338:114–140.

Manger PR, Woods TM, Jones EG. 1996. Representation of face and intra-oral structures in area 3b of macaque monkey somatosensory cortex.J Comp Neurol 371:513–521.

Markowitsch HJ, Tulving E. 1994. Cognitive processes and cerebral corti-cal fundi: findings from positron-emission tomography studies. ProcNatl Acad Sci USA 91:10507–10511.

Matelli M, Luppino G, Govoni P, Geyer S. 1997. Arm fields of both superiorand inferior area 6 in the macaque monkey are targets of area PFG. SocNeurosci Abstr 23:1273.

Maunsell JHR, Van Essen DC. 1983. The connections of the middle tem-poral visual area (MT) and their relationship to a cortical hierarchy inthe macaque monkey. J Neurosci 3:2563–2586.

Maunsell JHR, Van Essen DC. 1987. Topographic organization of themiddle temporal visual area in the macaque monkey: representationalbiases and the relationship to callosal connections and myeloarchitec-tonic boundaries. J Comp Neurol 266:535–555.

Mazzoni P. 1994. Spatial perception and movement planning in the pos-terior parietal cortex. MIT Thesis.

Menetrey D. 1985. Retrograde tracing of neural pathways with a protein-gold complex. I. Light microscopic detection after silver intensification.Histochemistry 83:391–395.

Mesulam M-M. 1981. A cortical network for directed attention and unilat-eral neglect. Ann Neurol 10:309–325.

Mesulam M-M. 1998. From sensation to cognition. Brain 121:1013–1052.Mesulam M-M, Van Hoesen GW, Pandya DN, Geschwind N. 1977. Limbic

and sensory connections of the inferior parietal lobule (area PG) in therhesus monkey: a study with a new method for horseradish peroxidasehistochemistry. Brain Res 136:393–414.

Mistlin AJ, Perrett DI. 1990. Visual and somatosensory processing in themacaque temporal cortex: The role of “expectation”. Exp Brain Res82:437–450.


Morecraft J, Geula C, Mesulam M-M. 1993. Architecture of connectivitywithin a cingulo-fronto-parietal neurocognitive network for directedattention. Arch Neurol 50:279–284.

Morel A, Garraghty PE, Kaas JH. 1993. Tonotopic organization, architec-tonic fields, and connections of auditory cortex in macaque monkeys.J Comp Neurol 335:437–459.

Mountcastle VB, Lynch JC, Georgopoulos A, Sakata H, Acuna C. 1975.Posterior parietal association cortex of the monkey: command functionsfor operations within extrapersonal space. J Neurophysiol 38:871–908.

Neal JW, Pearson RCA, Powell TPS. 1986. The organization of the corti-cocortical projection of area 5 upon area 7 in the parietal lobe of themonkey. Brain Res 381:164–167.

Neal JW, Pearson RCA, Powell TPS. 1987. The cortico-cortical connectionsof area 7b, PF, in the parietal lobe of the monkey. Exp Brain Res419:341–346.

Neal JW, Pearson RCA, Powell TPS. 1990. The ipsilateral cortico-corticalconnections of area 7b, PF, in the parietal and temporal lobes of themonkey. Brain Res 524:119–132.

Newsome WT, Maunsell JHR, Van Essen DC. 1986. Ventral posteriorvisual area of the macaque: visual topography and areal boundaries.J Comp Neurol 252:139–153.

Pandya DN, Kuypers HGJM. 1969. Cortico-cortical connections in therhesus monkey. Brain Res 13:13–36.

Pandya DN, Seltzer B. 1982. Intrinsic connections and architectonics ofposterior parietal cortex in the rhesus monkey. J Comp Neurol 204:196–210.

Pandya DN, Vignolo LA. 1969. Interhemispheric projections of the parietallobe in the Rhesus monkey. Brain Res 15:49–65.

Pearson RCA, Powell TPS. 1985. The projection of the primary somaticsensory cortex upon area 5 in the monkey. Brain Res Brain Res Rev9:89–107.

Perkel DJ, Bullier J, Kennedy H. 1986. Topography of the afferent connec-tivity of area 17 in the macaque monkey: a double-labelling study.J Comp Neurol 253:374–402.

Pons TP, Kaas JH. 1986. Corticocortical connections of area 2 of somato-sensory cortex in macaque monkeys: a correlative anatomical andelectrophysiological study. J Comp Neurol 248:313–335.

Pons TP, Garraghty PE, Cusick CG, Kaas JH. 1985. The somatotopicorganization of area 2 in macaque monkeys. J Comp Neurol 241:445–466.

Rizzolatti G, Scandolara C, Matelli M, Gentilucci M. 1981a. Afferent prop-erties of periarcuate neurons in macaque monkeys. I. Somatosensoryresponses. Behav Brain Res 2:125:146.

Rizzolatti G, Scandolara C, Matelli M, Gentilucci M. 1981b. Afferent prop-erties of periarcuate neurons in macaque monkeys. II. Visual re-sponses. Behav Brain Res 2:147–163.

Rizzolatti G, Riggio L, Dascola I, Umilta C. 1987. Reorientating attentionacross the horizontal and vertical meridians: evidence in favor of apremotor theory of attention. Neuropsychologia 25:31–40.

Robinson CJ, Burton H. 1980. Organization of somatosensory receptivefields in cortical areas 7b, retroinsula, postauditory and granular in-sula of M. fascicularis. J Comp Neurol 192:69–92.

Rockland KS, Pandya DN. 1979. Laminar origins and terminations ofcortical connections of the occipital lobe in the Rhesus monkey. BrainRes 179:3–20.

Sakata H, Takaoka Y, Kawarasaki A, Hidetoshi S. 1973. Somatosensoryproperties of neurons in the superior parietal cortex (area 5) of therhesus monkey. Brain Res 64:85–102.

Sakata H, Murata A, Luppino G, Kaseda M, Kusunoki M. 1992. Selectivityof hand-movement-related neurons of the parietal cortex for shape, sizeand orientation of objects and hand grips. Soc Neurosci Abstr 18:504.

Sakata H, Taira M, Kusunoki M, Murata A, Tanaka Y. 1997. The parietalassociation cortex in depth perception and visual control of hand action.Trends Neurosci 20:350–357.

Schaafsma SJ, Duysens J. 1996. Neurons in the ventral intraparietal areaof awake macaque monkey closely resemble neurons in the dorsal partof the medial superior temporal area in their responses to optic flowpatterns. J Neurophys 76:4056–4068.

Schall JD, Morel A, King DJ, Bullier J. 1995. Topography of visual cortexconnections with frontal eye field in macaque: convergence and segre-gation of processing streams. J Neurosci 15:4464–4487.

Schwarz DWF, Fredrickson JM. 1971. Rhesus monkey vestibular cortex: abimodal primary projection field. Science 172:280–281.

Seltzer B, Pandya DN. 1980. Converging visual and somatic sensory cor-tical input to the intraparietal sulcus of the rhesus monkey. Brain Res192:339–351.

Seltzer B, Pandya DN. 1986. Posterior parietal projections to the intrapa-rietal sulcus of the rhesus monkey. Exp Brain Res 62:459–469.

Seltzer B, Pandya DN. 1991. Post-Rolandic cortical projections of thesuperior temporal sulcus in the rhesus monkey. J Comp Neurol 312:625–640.

Sessle BJ, Wiesendanger M. 1982. Structural and functional definition ofthe motor cortex in the monkey (Macaca fascicularis). J Physiol 323:245–265.

Shatz CJ. 1977. Anatomy of interhemispheric connections in the visual sys-tem of Boston Siamese and ordinary cats. J Comp Neurol 173:497–518.

Snyder LH, Grieve KL, Brotchie P, Andersen RA. 1998. Separate body- andworld-referenced representations of visual space in parietal cortex.Nature 394:887–891.

Stanton GB, Cruce WLR, Goldberg ME, Robinson DL. 1977. Some ipsilat-eral projections to areas PF and PG of the inferior parietal lobule inmonkeys. Neurosci Lett 6:243–250.

Taira M, Mine S, Georgopoulos AP, Murata A, Sakata H. 1990. Parietalcortex neurons of the monkey related to the visual guidance of handmovement. Brain Res 83:29–36.

Tanne J, Boussaoud D, Boyer-Zeller N, Rouiller EM. 1995. Direct visualpathways for reaching movements in the macaque monkey. Neurore-port 7:267–272.

Tanne J, Boussaoud D, Bover-Zeller N, Moret V, Rouiller EM. 1996. Pari-etal inputs to dorsal vs. ventral premotor areas in the macaque mon-key: a multiple anatomical tracing study. Soc Neurosci Abstr 22:1084.

Ungerleider LG, Desimone R. 1986a. Projections to the superior temporalsulcus from the central and peripheral field representations of V1 andV2. J Comp Neurol 248:147–163.

Ungerleider LG, Desimone R. 1986b. Cortical connections of visual areaMT in the macaque. J Comp Neurol 248:190–222.

Ungerleider LG, Mishkin M. 1982. Two cortical visual systems. In: IngleDJ, Goodale MA, Mansfield RJW, editors. Analysis of visual behavior.Cambridge, MA: MIT Press. p 549–586.

Van Essen DC. 1997. A tension-based theory of morphogenesis and com-pact wiring in the central nervous system. Nature 385:313–318.

Van Essen DC, Maunsell JHR, Bixby JL. 1981. The middle temporalvisual area in the macaque: myeloarchitecture, connections, func-tional properties and topographic organization. J Comp Neurol 199:293–326.

Van Essen DC, Newsome WT, Maunsell JHR, Bixby JL. 1986. The projec-tions from striate cortex (V1) to areas V2 and V3 in the macaquemonkey: asymmetries, areal boundaries, and patchy connections.J Comp Neurol 244:451–480.

Van Essen DC, Felleman DM, DeYoe EA, Olavarria J, Knierim JJ. 1990.Modular and hierarchical organization of extrastriate visual cortex in themacaque monkey. Cold Spring Harbor Symp Quant Biol 55:679–696.

Van Essen DC, Felleman DF, DeYoe EA, Knierim JJ. 1991. Probing theprimate visual cortex: pathways and perspectives. In: Valberg A, LeeBB, editors. From pigments to perception: advances in understandingvisual processes. New York: Plenum. p 227–237.

Van Essen DC, Drury HA, Joshi S, Miller MI. 1998. Functional andstructural mapping of human cerebral cortex: solutions are in thesurfaces. Proc Natl Acad Sci USA 95:788–795.

Vogt BA, Pandya DN. 1978. Cortico-cortical connections of somatic sensorycortex (areas 3, 1 and 2) in rhesus monkey. J Comp Neurol 177:179–192.

Woolsey CN, Settlage PH, Meyer DR, Sencer W, Pinto Hamuy T, TravisAM. 1952. Patterns of localization in precentral and “supplementary”motor areas and their relation to the concept of a premotor area. ResPubl Assoc Nerv Ment Dis 30:238–264.


Documents

Corticocortical connections of visual, …brainmap.wustl.edu/resources/Lewis_JCN00b.pdfCorticocortical Connections of Visual, Sensorimotor, ... Ri retroinsula S2cx ... Corticocortical