Efferent association pathways originating in the caudal prefrontal cortex in the macaque monkey
Post on 12-Jun-2016
Efferent Association PathwaysOriginating in the Caudal Prefrontal
Cortex in the Macaque Monkey
M. PETRIDES1,2* AND D.N. PANDYA351Montreal Neurological Institute, McGill University, Montreal, Quebec H3A 2B4, Canada
2Department of Psychology, McGill University, Montreal, Quebec H3A 1B1, Canada3Departments of Anatomy and Neurology, Boston University School of Medicine,
Boston, Massachusetts 021184Department of Neurology, Beth Israel Deaconess Medical Center,
Boston, Massachusetts 022155ENR Memorial VA Hospital, Bedford, Massachusetts 01730
ABSTRACTThe efferent association bers from the caudal part of the prefrontal cortex to posterior
cortical areas course via several pathways: the three components of the superior longitudinalfasciculus (SLF I, SLF II, and SLF III), the arcuate fasciculus (AF), the fronto-occipitalfasciculus (FOF), the cingulate fasciculus (CING F), and the extreme capsule (Extm C).Fibers from area 8Av course via FOF and SLF II, merging in the white matter of the inferiorparietal lobule (IPL) and terminating in the caudal intraparietal sulcus (IPS). A group ofthese bers turns ventrally to terminate in the caudal superior temporal sulcus (STS). Fibersfrom the rostral part of area 8Ad course via FOF and SLF II to the IPS and IPL and via theAF to the caudal superior temporal gyrus and STS. Some bers from the rostral part of area8Ad are conveyed to the medial parieto-occipital region via FOF, to the STS via Extm C, andto the caudal cingulate gyrus via CING F. Fibers from area 8B travel via SLF I to thesupplementary motor area and area 31 in the caudal dorsal cingulate region and via theCING F to cingulate areas 24 and 23 and the cingulate motor areas. Fibers from area 9/46dcourse via SLF I to the superior parietal lobule and medial parieto-occipital region, via SLFII to the IPL. Fibers from area 9/46v travel via SLF III to the rostral IPL and the frontopa-rietal opercular region and via the CING F to the cingulate gyrus. J. Comp. Neurol. 498:227251, 2006. 2006 Wiley-Liss, Inc.
Indexing terms: monkey; frontal cortex; prefrontal cortex; association pathways;autoradiography; connections
The caudal part of the lateral prefrontal cortex, consist-ing of the cortex lying immediately in front of the arcuatesulcus (i.e., the prearcuate cortex), receives its main inputfrom the posterior parietal, posterior temporal, and extra-striate cortical regions (e.g., Kuypers et al., 1965; Jonesand Powell, 1970; Chavis and Pandya, 1976; Barbas andMesulam, 1981; Petrides and Pandya, 1984, 1988; Barbas,1988; Cavada and Goldman-Rakic, 1989; Ungerleider etal., 1989; Andersen et al., 1990; Hackett et al., 1999;Romanski et al., 1999a,b). The bers originating from theparietal cortex course as part of the superior longitudinalfasciculus (SLF), which has been demonstrated to consistof three distinct ber bundles: SLF I leading to the cau-dodorsal lateral and medial frontal cortex, SLF II to thedorsal premotor, prearcuate, and caudal mid-dorsolateralfrontal cortex, and SLF III to the ventral premotor and
ventrocaudal lateral prefrontal cortex (Petrides and Pan-dya, 1984). Fibers originating from the caudal superiortemporal gyrus and adjacent superior temporal sulcusterminate predominantly in dorsal area 8 and area 9/46d,
Grant sponsor: Canadian Institutes of Health Research (CIHR); Grantnumber: MOP-14620; Grant sponsor: Natural Sciences and EngineeringResearch Council of Canada (NSERC); Grant sponsor: ENR Memorial VAHospital, Bedford, Massachusetts.
*Correspondence to: Michael Petrides, Montreal Neurological Institute,3801 University St., Montreal, Quebec H3A 2B4, Canada.E-mail: firstname.lastname@example.org
Received 7 October 2005; Revised 1 February 2006; Accepted 3 April2006.
DOI 10.1002/cne.21048Published online in Wiley InterScience (www.interscience.wiley.com).
THE JOURNAL OF COMPARATIVE NEUROLOGY 498:227251 (2006)
2006 WILEY-LISS, INC.
while input from the middle superior temporal regiontargets primarily more anterior prefrontal areas (Petridesand Pandya, 1988; Hackett et al., 1999; Romanski et al.,1999a,b). The bers from the caudal superior temporalregion run via the arcuate fasciculus, while those from themiddle superior temporal region course predominantlythrough the extreme capsule (Petrides and Pandya, 1988).In addition, connections from the extrastriate region tothe caudolateral prefrontal cortex have consistently beenreported (e.g., Kuypers et al., 1965; Pandya and Kuypers,1969; Jones and Powell, 1970; Ungerleider et al., 1989).
Although it is known that the caudal lateral prefrontalcortical areas project back to post-Rolandic associationcortex (e.g., Jones and Powell, 1970; Pandya and Vignolo,1971), there is no information regarding the precise courseand the pathways utilized by the efferent association -bers originating from the caudolateral prefrontal cortex.The present study examined the pathways originatingfrom this region by means of the autoradiographic tech-nique, which allows the precise delineation of the entirecourse of pathways and their terminations. A major ques-tion of interest is whether the efferent bers from thecaudal prefrontal region course via the same general path-ways that relay afferent information to this region?
The caudal lateral prefrontal region in interaction withposterior parietal, posterior temporal, and extrastriateregions comprises the main cerebral circuits underlying
visual and auditory spatial attentional processing, as wellas body-centered multimodal interactions. Furthermore,it is generally accepted that the prefrontal cortex exercisesa top-down control over posterior cortical regions (e.g.,Petrides 1996, 2005; Barcelo et al., 2000; Moore and Arm-strong, 2003). Thus, a clear understanding of the regula-tion by the frontal cortex of posterior cortical attentionalprocessing must be based on a solid understanding of thepathways via which this control is exerted. This is ofimportance because many disease processes involve dam-age to the white matter and the behavioral/cognitive ef-fects observed may reect damage to both the inputs andoutputs of particular areas.
MATERIALS AND METHODSSurgery
Injections of radioactively labeled amino acids wereplaced in different parts of the prefrontal cortex in sevenrhesus monkeys (Macaca mulatta). The animals were im-mobilized with ketamine hydrochloride (10 mg/kg) andthen deeply anesthetized with sodium pentobarbital (30mg/kg) (Sigma, St. Louis, MO) administered intrave-nously. A craniotomy was then performed, under asepticsurgical technique, over the target area in the frontal lobe.In each case, an attempt was made to place two juxta-posed isotope injections into an architectonic area accord-ing to the parcellation of Petrides and Pandya (1994) (Fig.1). The intracortical injections consisted of radioactivelylabeled amino acids (3H-leucine and/or proline; volumerange, 0.41.0 L; specic activity range, 4080 Cu,aqueous solution; New England Nuclear Brand Radio-chemicals from PerkinElmer, Norwalk, CT). After sur-vival periods ranging from 710 days the animals weredeeply anesthetized with sodium pentobarbital and per-fused transcardially with physiologic saline, followed by a10% formalin solution.
The brains were divided into two blocks by a coronal cutand photographed from all angles. They were subse-quently embedded in parafn and sectioned at 16 mthickness. The brains were then processed for autoradiog-raphy according to the technique described by Cowan etal. (1972). The exposure times varied between 3 and 6months. During this period, trial sections were developedat monthly intervals for identication of optimal radiola-beling. The sections were also counterstained with thi-onine to permit identication of the architectonic areas. Aseries of coronal sections of the hemisphere were exam-ined microscopically with darkeld illumination. The la-beled bers in the white matter and the terminal labelingin the cerebral cortex and subcortical structures wererecorded with the aid of an X-Y plotter (Hewlett Packard,Corvallis, OR) that was electronically coupled to the stageof the microscope (Leitz Aristoplan, Wetzlar, Germany).This information was then used to reconstruct the injec-tion and termination sites, as well as the course of thelabeled bers. The cytoarchitectonic boundaries of the pro-jection areas within the cerebral cortex, as well as thesites of origin of the pathways within the prefrontal cor-tex, were established in the experimental material underlight eld illumination.
The distribution of the terminations of the labeled berswas transferred onto two-dimensional reconstructions(i.e., attened views) of the lateral, medial, and ventral
Area 6DC dorsocaudal area 6Area 6DR dorsorostral area 6AF arcuate fasciculusAS arcuate sulcusASi arcuate sulcus (inferior limb)ASu arcuate sulcus (upper limb)CING S cingulate sulcusCING F cingulate fasciculusCC corpus callosumCF calcarine ssureCS central sulcusEC external capsuleExtm C extreme capsuleFOF fronto-occipital fasciculusIC internal capsuleILF inferior longitudinal fasciculusLF lateral ssureLS lunate sulcusIPL inferior parietal lobuleIPS intraparietal sulcusIOS inferior occipital sulcusLOS lateral orbital sulcusMB Muratoff bundleMdLF middle longitudinal fasciculusMOS medial orbital sulcusOTS occipitotemporal sulcusPOMS medial parieto-occipital sulcusPreSMA presupplementary motor areaPresub presubiculumPro M proisocortical motor areaPS sulcus principalisRS rhinal sulcusRsp retrosplenial cortexSB subcortical bundleSLF I superior longitudinal fasciculus (bundle I)SLF II superior longitudinal fasciculus (bundle II)SLF III superior longitudinal fasciculus (bundle III)SMA supplementary motor area (secondary motor area MII)St B striatal bundleSTS superior temporal sulcusUF uncinate fasciculus
The Journal of Comparative Neurology. DOI 10.1002/cne
228 M. PETRIDES AND D.N. PANDYA
surfaces of the examined cerebral hemispheres. These 2Dreconstructions of the hemispheres were achieved using aprecision technical drawing software program (Au-toSketch, Release 7, Autodesk, San Rafael, CA). In orderto minimize the distortion in the normal view of the cere-
bral hemisphere that inevitably follows such attenedmaps, we unfolded the cortex lying within the major sulciseparately. These unfolded sulci are presented next to thelateral and medial views of the hemispheres so that thereader can appreciate the details of the terminationswithin the sulci. On the coronal sections of the hemispherethat was to be reconstructed, we traced the distance fromthe midline (i.e., the border of the lateral with the medialsurface of the hemisphere) to the rst sulcus encounteredlaterally. We then measured the distance from that sulcusto the next sulcus and so on until the lateral-to-ventraledge of the hemisphere was reached. These measurementswere used for the lateral surface reconstruction. For thereconstruction of the medial surface of the frontal lobe, wemeasured the distance from the dorsalmost part of themidline to the rst sulcus encountered ventrally and thento the next sulcus until the ventralmost part of the medialsurface was reached. The orbital surface of the frontal lobewas reconstructed by measuring the distance, on eachcoronal section, from the midline to the rst sulcus en-countered laterally (i.e., the medial orbital sulcus) andthen from there to the lateral orbital sulcus and then tothe ventral-to-lateral edge of the hemisphere. The medial/ventral surface of the temporal lobe was included with thereconstruction of the medial surface of the hemisphereand the origin of the measurements was the hippocampalsulcus. The measurements obtained for the lateral, me-dial, and orbital surfaces, as well as the cortex within thesulci, were thus a series of line segments (the y coordi-nates) arranged in the anteroposterior direction (x coordi-nates). In separate spreadsheets, the points (x, y coordi-nates) were plotted and joined together in order toreconstruct the 2D attened outlines of surfaces of thehemisphere and the sulci.
The labeled long association bers that were traced oncoronal sections of selected cases were transferred ontothe corresponding coronal sections of a monkey brain(male, Macaca mulatta) that was scanned with magneticresonance imaging (MRI). A high-resolution T1-weighted3D MRI anatomical image of this monkey brain was ac-quired from a 1.5 T Siemens (Erlangen, Germany) SonataVision scanner (TR 22 ms, TE 9.2 ms, ip angle 30,1.0 1.0 1.0 mm, coronal slices, 1.0 mm in thickness).3D surface reconstructions of this brain were obtainedusing an automated, model-based, surface deformationalgorithm (MacDonald et al., 1994). The position of thelabeled bers were marked on the MRI monkey brainusing DISPLAY, an interactive 3D imaging software pack-age (MacDonald, 1996) that allows the MRI le to beviewed and marked simultaneously in the coronal, hori-zontal, and sagittal planes, as well as in one other arbi-trary plane of the investigators choice. All views are au-tomatically updated as the cursor is moved to any point ina given section and the investigator can mark any point inthe MRI volume. The markings of the charted bers on aseries of contiguous 2D sections could then be transferredonto comparable coronal sections of the standard MRIbrain using local details in the anatomy of the sections.Once the marked bers from the charted case were trans-ferred into the MRI brain, they could be reconstructed andviewed on the 3D rendering of the standard MRI brain,permitting a clear visualization of the course of the entirepathway from various angles. This 3D reconstructionhelped provide a clearer appreciation of the course of thevarious pathways within the white matter of the cerebral
Fig. 1. Cytoarchitectonic map of the lateral (A), orbital (B), andmedial (C) surfaces of the frontal lobe of the monkey according toPetrides and Pandya (1994).
The Journal of Comparative Neurology. DOI 10.1002/cne
229CAUDAL PREFRONTAL CORTICAL CONNECTIONS
hemisphere and is illustrated in summary diagrams (Figs.1012). The 3D hemispheric view provides an impressionof the trajectories of the identied association pathways,whereas the coronal sections indicate the location of thesepathways at different levels.
The care and use of the animals were in accordance withthe guidelines of the National Institutes of Health.
Terminology, architectonic parcellation, andborders
The architectonic delineation of the prefrontal cortexwas determined according to the criteria of Petrides andPandya (1994) (Fig. 1), the posterior parietal cortex ac-cording Pandya and Seltzer (1982), the superior temporalgyrus according to Pandya and Sanides (1973), the infero-temporal cortex and the cortex of the superior temporalsulcus according to Seltzer and Pandya (1978), and theposterior parahippocampal gyrus according to Rosene andPandya (1983). The parcellation of the cingulate cortex,and that of the retrosplenial cortex, which lies on thedorsal aspect of the corpus callosum, follows Vogt et a...