Skip Navigation

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (26)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Morecraft, R. J.
Right arrow Articles by Schoolfield, M. W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Morecraft, R. J.
Right arrow Articles by Schoolfield, M. W.
Social Bookmarking
 Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Brain, Vol. 125, No. 1, 176-198, January 1, 2002
© 2002 Oxford University Press

Localization of arm representation in the corona radiata and internal capsule in the non-human primate

Robert J. Morecraft1, James L. Herrick1, Kimberly S. Stilwell-Morecraft1, Jennifer L. Louie1, Clinton M. Schroeder1, Jonovan G. Ottenbacher1 and Matt W. Schoolfield1

1Division of Basic Biomedical Sciences, The University of South Dakota School of Medicine, Vermillion, SD 57069, USA Correspondence to: Dr Robert J. Morecraft, Division of Basic Biomedical Sciences, The University of South Dakota School of Medicine, Vermillion, SD 57069, USA E-mail: rmorecra{at}usd.edu

Received April 5, 2001. Revised August 8, 2001. Second revision August 16, 2001. Accepted September 10, 2001. .


    Summary
 Top
 Summary
 Introduction
 Material and methods
 Results
 Discussion
 References
 
Localization of the corticofugal projection in the corona radiata (CR) and internal capsule (IC) can assist in evaluating a patient’s residual motor capacity following subtotal brain damage and predicting their potential for functional restitution. To advance our understanding of the organization of the corticofugal projection in this critical brain region, we studied the trajectories of the projection arising from six different cortical arm representations in rhesus monkeys. They included the arm representation of the primary (M1), ventral lateral pre- (LPMCv), dorsolateral pre- (LPMCd), supplementary (M2), rostral cingulate (M3) and caudal cingulate (M4) motor cortices. In the CR, each pathway was segregated as medial motor area fibres arched over the caudate and lateral motor area fibres arched over the putamen. In the IC, the individual corticofugal pathways were found to be widespread, topographically organized and partially overlapping. At superior levels of the IC, the corticofugal projection from the arm representation of M3 coursed through the middle and posterior portion of the anterior limb (ICa). The projection from M2 passed through the posterior portion of the ICa and the genu (ICg). The projection from LPMCv travelled through the genu and anterior portion of the posterior limb (ICp). The projection from LPMCd occupied the anterior portion of the ICp. The projection from M4 descended through the mid-portion of the ICp. Fibres from M1 also travelled in the ICp, positioned immediately posterior to the M4 projection. As each fibre system progressed inferiorly within the IC, all fibres shifted posteriorly to occupy the ICp. Within the ICp, the projections from M3, M2, LPMCv, LPMCd, M4 and M1 maintained their anterior to posterior orientation, respectively. M2, LPMCd and LPMCv fibres overlapped extensively, as did fibres from M4 and M1. Our data suggest that CR and superior capsular lesions may correlate with more favourable levels of functional recovery due to the widespread nature of arm representation. In contrast, the extensive overlap and comparatively condensed organization of arm representation at inferior capsular levels suggest that lesions seated inferiorly are likely to correlate with poorer levels of recovery of upper limb movement. Based on the relative density of corticospinal neurones associated with the motor areas studied, our findings also suggest that motor deficit severity is likely to increase as a lesion occupies progressively more posterior regions of the IC.

Keywords: cingulate cortex; frontal lobe; limbic lobe; motor cortex; striatocapsular infarction; stroke, subcortical white matter

Abbreviations: ABC= avidin–biotin; BDA = biotinylated dextran amine; CR = corona radiata; DAB = 3,3'-diaminobenzidine tetrahydrochloride; FD = fluorescein dextran; FR = fluoro ruby; IC = internal capsule; ICa = anterior limb of the internal capsule; ICg = genu of the internal capsule; ICp = posterior limb of the internal capsule; LYD = lucifer yellow dextran; LPMCd = dorsal lateral premotor cortex; LPMCv = ventral lateral premotor cortex; M1 = primary motor cortex; M2 = supplementary motor cortex; M3 = rostral cingulate motor cortex; M4 = caudal cingulate motor cortex; PHA-L = Phaseolus vulgaris leucoagglutinin


    Introduction
 Top
 Summary
 Introduction
 Material and methods
 Results
 Discussion
 References
 
Although the mechanisms underlying motor recovery following subtotal brain damage remain obscure, it has been suggested that intact corticofugal projections arising from cortex located ipsilateral to the lesion may contribute to the initial regression of hemiplegia and eventual recovery of motor function (Bucy, 1957Go; Chollet et al., 1991Go; Fries et al., 1993Go; Furlan et al., 1996Go; Dettmers et al., 1997Go). In support of this notion are recent clinical observations suggesting that corticofugal fibre integrity is a critical determinant for assessing residual motor impairment as well as being an important indicator for predicting the quality of motor recovery that may follow stroke or related neuronal damage (Chollet et al., 1991Go; Fries et al., 1993Go; Binkofski et al., 1996Go; Furlan et al., 1996Go; Dettmers et al., 1997Go; Hallett et al., 1998Go; Seitz et al., 1998Go; Werring et al., 1998Go; Byrnes et al., 1999Go; De Vries et al., 1999Go; Lee et al., 2000Go; Pineiro et al., 2000Go; Binkofski et al., 2001Go). Substantial corticofugal fibre disruption appears to correlate with poor functional outcome and, conversely, limited corticofugal fibre damage correlates with more favourable levels of motor recovery. Further supporting this view is a growing body of evidence indicating that alterations in regional cerebral blood flow patterns in recovered stroke patients may, in part, be related to corticofugal sparing (Chollet et al., 1991Go; Weiller et al., 1992Go, 1993, 1995; Dettmers et al., 1997Go; Honda, 1997Go; Liepert et al., 1998Go; Silvestrini et al., 1998Go; Fandino et al., 1999Go; Green et al., 1999Go; Nelles et al., 1999Go; Caramia et al., 2000Go; Marshall et al., 2000Go). For example, comparisons of regional cerebral blood flow patterns between stroke patients and controls show that patients with damage to the posterior limb of the internal capsule (ICp) or the lateral region of the primary motor cortex exhibit enhanced metabolic activation in brain regions such as the ipsilateral supplementary and cingulate motor cortices when performing finger movements with the recovered limb. It has been suggested that the elevated activity in these spared motor areas may reflect a compensatory response for the partial loss of corticofugal projections from primary motor cortex (M1), thus being beneficial to the overall process of motor recovery. It therefore appears that the brain has the innate capability to shift functional propriety of hand movements, once heavily mediated by M1, to other cortical areas which are excluded from the insult and are endowed with intact corticofugal projections ending in key subcortical targets that influence arm movements, such as the spinal cord and brainstem nuclei.

Historically, one of the most widely held concepts in neurological thought suggested that the M1 of the precentral gyrus was the sole contributor to the pyramidal tract in primates (Holmes and May, 1909Go; Lassek, 1952Go, 1954; Bucy, 1957Go). However, modern studies have revised this idea by showing that the pyramidal tract is formed by many corticofugal systems and that the corticospinal component of the pyramidal tract originates from widespread portions of the cerebral cortex (Catman-Berrevoets and Kuypers, 1976Go; Kuypers, 1981Go; Toyoshima and Sakai, 1982Go; Davidoff, 1990Go; Nudo and Masterson, 1990Go). Even more discriminating efforts have demonstrated that there are several arm representations in the human and non-human primate motor cortices which mediate motor control and preferentially innervate not only spinal levels forming the brachial plexus (Biber et al., 1978Go; Murray and Coulter, 1981Go; Lawrence et al., 1985Go; Hutchins et al., 1988Go; Dum and Strick, 1991Go, 1996; Huntley and Jones, 1991Go; He et al., 1993Go, 1995; Paus et al., 1993Go; Galea and Darian-Smith, 1994Go; Luppino et al., 1994Go; Rouiller et al., 1994Go, 1996; Picard and Strick, 1996Go; Morecraft et al., 1997Go; Luppino and Rizzolatti, 2000Go; Wu et al., 2000Go; Wang et al., 2001), but also additional subcortical sites influencing arm movements such as the red nucleus, brainstem reticular formation and pontine grey matter (Brodal, 1978Go; Künzle, 1978Go; Catman-Berrevoets et al., 1979Go; Hartmann-von Monakow et al., 1979Go; Kuypers, 1981Go; Humphrey et al., 1984Go; Leichnetz, 1986Go; Burman et al., 2000Go; Schmahmann, 2000Go) (Fig. 1). The redundancy of arm representation at the cortical level, and the fact that the corticofugal projection contains not only corticospinal axons, but also other important projections, which include the corticopontine and corticoreticular projection, has contributed further to the idea that significant gains in motor recovery may occur, in part, through reorganization of the corticofugal projection from intact cortical motor areas located ipsilateral to a subtotal brain lesion. From a clinical and rehabilitative perspective, determining corticofugal integrity requires a detailed understanding of the origin of each projection at the cortical level. Equally as important, however, is a precise understanding of the organization of the various corticofugal pathways as they descend through each major subdivision of the CNS since lesions are rarely confined to the cortical grey matter. Indeed, axons forming the various corticofugal pathways course through a wide variety of anatomical and vascular territories, and those ending in the spinal cord represent the longest efferent projection within the CNS (Kuypers, 1981Go; Ghika et al., 1990Go; Hupperts et al., 1994Go; Tatu et al., 1996Go, 1998). Clearly, it would be advantageous to know the subcortical organization of the various corticofugal projection systems in order to assist in interpreting emerging clinical sequelae which present in the motor domain following localized brain trauma. Such information may also be useful for guiding therapeutic options based on lesion location.



View larger version (39K):
[in this window]
[in a new window]
 
Fig. 1 (A) Schematic diagrams of the medial (top) and lateral (bottom) surfaces of the cerebral cortex of the rhesus monkey depicting the basic organization of the frontal lobe and adjacent cingulate cortex. (B) Enlarged and slightly modified depiction of A illustrating the basic somatotopical organization of the motor cortices and the locations of the six different arm representations whose corticofugal trajectories were studied in the corona radiata and internal capsule. A = arm; as = arcuate sulcus; cf = calcarine fissure; cgs = cingulate sulcus; cs = central sulcus; F = face; FEF = frontal eye fields; ios = inferior occipital sulcus = ips = intraparietal sulcus; L = leg; lf = lateral fissure; LPMCd = dorsal lateral premotor cortex; LPMCv = ventral lateral premotor cortex; ls = lunate sulcus; M1 = primary motor cortex; M2 = supplementary motor cortex; M3 = rostral cingulate motor cortex; M4 = caudal cingulate motor cortex; pre-SMA = pre-supplementary motor cortex; rs = rhinal sulcus; SEF = supplementary eye field; sts = superior temporal sulcus; poms = medial parieto-occipital sulcus; sts = superior temporal sulcus; UL = upper lip. Numerical designations (9, 10, 12, 14, 23a, 23b, 24a, 24b, 25, 45, 46) identify basic cytoarchitectonic areas.

 
In light of the current emphasis on corticofugal fibre integrity and its potential role in functional restitution of arm movement following localized brain trauma, we investigated the organization of the corticofugal projection from multiple cortical arm representations in the territory of the corona radiata (CR) and internal capsule (IC). In this report, we focused on the projection emanating from six different arm representations in the monkey motor cortex. Our results demonstrate that arm representation in the CR and IC is highly organized and widely dispersed. We also found that each pathway continually altered, or shifted, its location as it progressed from superior to inferior levels of the CR and IC. Our findings suggest that small lesions occupying localized regions of the CR and IC are likely to disrupt corticofugal innervation from predictable cortical origins. On the other hand, our observations suggest that the widespread arrangement of arm representation occurring throughout the CR and IC increases the likelihood of corticofugal sparing which, subsequently, may contribute to the reversal of functional poverty and eventual recovery of hand and arm movements that may follow localized supratentorial brain damage.


    Material and methods
 Top
 Summary
 Introduction
 Material and methods
 Results
 Discussion
 References
 
The supramesencephalic organization of corticofugal fibres arising from six different cortical arm representations was studied in seven rhesus monkeys (Macaca mulatta) (Table 1). The cortical arm representations studied in this report included the primary motor cortex (M1 or area 4), ventral lateral premotor cortex (LPMCv or area 6V), dorsal lateral premotor cortex (LPMCd or area 6D), supplementary motor cortex (M2 or area 6m), rostral cingulate motor cortex (M3 or area 24c) and caudal cingulate motor cortex (M4 or area 23c) (Fig. ). Each monkey was injected with an anterograde tracer into one or more arm representation(s) of the frontal and cingulate cortices (Table 1 and Fig. 2). In five monkeys, the cortex was processed for fibre visualization in the coronal plane (Figs –5). In two monkeys (cases SDM 15 and SDM 23), five different anterograde tracers were injected into five different arm representations in the same hemisphere (Figs 6 and 7). In these cases, the cortex was processed for fibre visualization in the horizontal plane in reference to the intercommissural line (Talairach and Tournoux, 1988Go) (Figs and 7). In all cases, the brainstem and spinal cord were sectioned in the standard horizontal plane (Fig. 8) with the exception of IM 121, SDM 6 and SDM 7, in which the spinal cord was sectioned longitudinally.


View this table:
[in this window]
[in a new window]
 
Table 1 Description of the injection parameters and arm representations involved in each monkey
 


View larger version (113K):
[in this window]
[in a new window]
 
Fig. 2 Plate of low-power photomicrographs illustrating representative examples of injection sites in different motor fields following tissue processing for immunohistochemical visualization. (A) Injections of PHA-L into M1 in case IM 121 visualized using ABC/DAB nickel-enhanced immunostaining (black reaction product). White arrows identify the Betz cell layer of layer V. (B) Injections of FR in M2 in case SDM 23 visualized using ABC and Vector SG proxidase substrate immunostaining (blue reaction product). This injection site is also depicted graphically in Fig. 7A. Each asterisk identifies a needle tract. (C) An injection of BDA in M3 of case SDM 6 visualized using standard alkaline phosphatase histochemistry. The white arrows identify pyramidal tract fibres arching over the head of the caudate. (D) An injection of BDA into M3 in case SDM 23 visualized using ABC/DAB immunostaining (brown reaction product). The white arrows indicate BDA-filled needle tracts. An adjacent level of this injection site is also depicted graphically in Fig. 7C. (E) An injection of BDA into M4 in case SDM 10 visualized using ABC/DAB immunostaining (brown reaction product). (F) An injection of LYD in M4 in case SDM 23 visualized using ABC and Vector SG peroxidase substrate immunostaining (blue reaction product). The white arrow identifies the LYD-filled needle tract. An adjacent level of this injection site is also depicted graphically in Fig. 7B. ca = caudate nucleus; cb = cingulum bundle; cc = corpus callosum; cgs = cingulate sulcus; cs = central sulcus; v = ventricle. Scale bar in F corresponds to all photomicrographic plates.

 


View larger version (63K):
[in this window]
[in a new window]
 
Fig. 5 Illustrated in rows, from the top to the bottom of the diagram, is the descending projection from the arm representation of M3 (SDM 6), M2 (SDM 7), M4 (SDM 10) and M1 (IM 121) in coronal section. Columns A–D are representative rostral to caudal coronal sections from each individual monkey case. Columns A and B illustrate coronal levels through the anterior limb of the internal capsule, with B coinciding with its approximate posterior boundary. Columns C and D illustrate rostral to caudal levels of the posterior limb, respectively. Medial motor area fibres arched over the caudate nucleus (M3, M2 and M4) whereas the lateral motor area fibres from M1 arched over putamen. Note the involvement of descending fibres from M3 and M2 at more anterior levels of the corona radiata and internal capsule (A and B) with progressive involvement of M4 and M1 fibres at more caudal levels (C and D). ac = anterior commissure; as = arcuate spur; ag = arcuate genu; ca = caudate nucleus; cgs = cingulate sulcus; cl = claustrum; cs = central sulcus; gp = globus pallidus; hy = hypothalamus; ilas = inferior limb of the arcuate sulcus; in = insula; ips = intraparietal sulcus; oc = optic chiasm; ot = optic tract; pu = putamen; slas = superior limb of the arcuate sulcus; th = thalamus; v = ventricle.

 


View larger version (33K):
[in this window]
[in a new window]
 
Fig. 6 Summary diagrams of the lateral (top) and medial (bottom) surfaces of the cerebral cortex illustrating the experimental design in monkey SDM 23. Cortical injections (coloured irregular spheres) were made into five different arm representations in the left hemisphere and the course of the corticofugal projection was mapped in the corona radiata and internal capsule using immunohistochemically processed tissue sections. The top enlargement depicts the location of the PHA-L injection site in M1 (blue) and the FD injection site in LPMCd (green) in relation to physiological mapping. The bottom enlargement demonstrates the location of the FR injection site in M2 (red) in relation to physiological mapping. Also depicted is the anatomical location of the BDA injection in M3 (light blue) and the LYD injection in M4 (yellow) in the cingulate sulcus. The M3 and M4 injection sites are not visible on surface view (see Fig. 7B and C). Each black dot represents a stimulation point labelled with the corresponding threshold level (µA) and body part where the movement was observed. The white dots within each coloured injection site represent the location of each Hamilton syringe penetration. The horizontal lines indicate the level of each representative tissue section shown graphically in Fig. 7 (see A–K) and are in reference to the anterior (CA) and posterior (CP) commissural plane (see horizontal section J). Generally, section F is representative of superior thalamic levels, G and H of middle thalamic levels and I and J of inferior thalamic levels. Section K represents the junctional region between the internal capsule and cerebral peduncle of the midbrain. ac = anterior commissure; apos = anterior parieto-occipital sulcus; cc = corpus callosum; D = digit; El = elbow; ilas = inferior limb of the arcuate sulcus; ots = occipitotemporal sulcus; pc = posterior commissure; pos = parieto-occipital sulcus; ps = principle sulcus; ros = rostral sulcus; scs = supracalcarine sulcus; Sh = shoulder; slas = superior limb of the arcuate sulcus; spd = superior pre-central dimple; spd = supraprinciple dimple; Th = thumb; To = tongue; Tr = trunk; Wr = wrist; for other conventions, see Figs 1 and 5.

 





View larger version (169K):
[in this window]
[in a new window]
 
Fig. 7 Representative horizontal serial sections through the cerebral cortex of monkey SDM 23 illustrating the individual cortical injection sites and corresponding trajectories of the five pathways studied through the corona radiata and internal capsule. Sections are shown from superior (A) to inferior (K). The location of each level is depicted in the top left of each plate as well as in relation to all other levels as shown in Fig. 6. Each injection site is colour coded and specifically identified to its right by tracer and motor area (A–C). Each descending fibre bundle is identified by a corresponding, colour-coded outline. Only the respective pyramidal tract pathways are illustrated. ca = caudate nucleus; cgs = cingulate sulcus; cl = claustrum; cs = central sulcus; fx = fornix; hb = habenula; hp = hippocampus; in = insula; ios = inferior occipital sulcus; lgn = lateral geniculate nucleus; lf = lateral fissure; los; lateral orbital sulcus; ls = lunate sulcus; mos = medial orbital sulcus; na; nucleus accumbens; pu = putamen; rn = red nucleus; sc = superior colliculus; scpd = superior precentral dimple; sn = substantia nigra; spcc = splenium of the corpus callosum; th = thalamus; v = lateral ventricle; for other conventions, see Figs 1 and 6.

 


View larger version (34K):
[in this window]
[in a new window]
 
Fig. 8 A series of line drawings through the spinal cord of adjacent serial sections through spinal level C7 in case SDM 23 illustrating anterograde labelling following injections of neuronal tract tracers into the cortical arm areas of LPMCd (FD), M2 (FR), M3 (BDA) and M4 (LYD) (see Figs 6 and 7). Labelled axons in the lateral corticospinal tract (LCST) are illustrated by the large black dots, and terminal boutons in the spinal grey are illustrated by the small black dots. Heavy labelling also occurred in other key subcortical targets influencing arm movements such as the brainstem reticular formation and pontine grey matter. ACST = anterior corticospinal tract; CT = cuneate tract; DIS = dorsal intermediate septum; DMS = dorsal median septum; GT = gracile tract; VMF = ventral median fissure. Roman numerals denote Rexed’s laminae.

 
This experimental design was developed to generate an understanding of corticofugal organization in multiple dimensions (e.g. coronal and horizontal preparations) and to facilitate direct comparisons of all descending fibre trajectories in the same experimental case (cases SDM 15 and SDM 23). Horizontal reconstruction was also conducted to establish an experimental database in the non-human primate that can be correlated to Talairach and Tournoux’s stereotaxic coordinate system and eventually used in conjunction with neuroimaging to interpret corticofugal fibre integrity in patients and develop other practical applications within the fields of neurosurgical and neurological practice.

Surgery and intracortical microstimulation
Experimental procedures conducted in these studies followed the United States Department of Agriculture and Society for Neurosciences guidelines and were approved by the Institutional Animal Care and Use Committee at The University of South Dakota. Surgical exposure of the cerebral cortex was accomplished as described previously (Morecraft and Van Hoesen, 1992Go, 1993, 1998; Morecraft et al., 1992Go, 1993). Briefly, each monkey was immobilized with ketamine hydrochloride, anaesthetized intravenously with pentobarbital (25 mg/kg followed by supplemental dosages to maintain a surgical level of anaesthesia) and transported to the surgical suite. A scalp incision and craniotomy were made over the lateral surface of the cerebral hemisphere and the monkey was administered mannitol (25%). A dural flap was made over the cortex of interest and selected veins were cauterized to enhance the exposure. Motor areas on the lateral wall (M1, LPMCd and LPMCv) and medially (M2) were localized using intracortical mapping (Morecraft et al., 1996Go, 1997, 2001). A monopolar tungsten microelectrode (impedance 1.0–2.0 {Omega}M) was attached to a Grass S88 stimulator and a constant current stimulus isolation unit. The system was grounded and the electrode was lowered to a depth of 100 µm and advanced at 500 µm intervals. Cortex was mapped with a train duration of 50–100 ms and a pulse duration of 0.2 ms delivered at 330 Hz. Stimulation points were spaced 1–2.5 mm apart to reduce tissue damage and maximize uptake and transport of each neuronal tracer. The minimal level of current needed to evoke movement was considered the movement threshold. During all experiments, applied current levels were monitored continuously by measuring the voltage with the use of a digital storage oscilloscope (LG Precision Co., Ltd) and determining the system resistance. The current values ranged between 3 and 80 µA. All movements at threshold were discrete and confined to a specific body part (e.g. upper lip, tongue, wrist, finger). The arm representation in M1, LPMCv, LPMCd and M2 was localized and the approximate boundaries between the adjacent face and leg representations were determined. Anterograde neuronal tract tracers were injected within the central portion of a defined arm representation to avoid involvement of the adjacent face and leg areas. Overlap between finger, wrist, elbow and shoulder movements was not a concern since the primary aim of this study was to map the corticofugal projection from each respective arm area. In the injection site for case SDM 15-LPMCv, the injection involved the face and arm transition area where both arm and face movements were observed following stimulation. Since the purpose of using microstimulation in these surgeries was to define sites for neuronal tracer injection, physiological mapping was minimized to reduce tissue damage in the area of tracer injection. Reference lesions (anodal current for 10–20 s) were placed in the cortex to assist in reconstructing the physiological data in relation to the histologically assessed injection sites. The location of each stimulation point and injection site was recorded on a HR-S9400U JVC videocassette recorder (JVC, Aurora, Ill., USA) that was connected to an MTI DC-330 3 chip colour camera (Dage-MTI, Inc., Michigan City, Ind., USA) mounted on the photo/video port of a Stortz M-703F surgical microscope (Leeds Precision, Inc, Minneapolis, Minn., USA). The rostral and caudal cingulate motor areas were exposed on the medial wall through an interhemispheric approach and located using anatomical landmarks and stereotaxic coordinates as guides. The arm area of M3 was localized in the lower bank of the cingulate sulcus, anterior to a coronal plane established at the genu of the arcuate sulcus (Biber et al., 1978Go; Hutchins et al., 1988Go; Shima et al., 1991Go; Morecraft and Van Hoesen, 1992Go; Galea and Darian-Smith, 1994Go; Luppino et al., 1994Go; He et al., 1995Go; Nimchinsky et al., 1996Go; Morecraft et al., 1997Go; Shima and Tanji, 1998Go). Specifically, this area is positioned between coronal levels +3.15 mm anterior to bregma and +0.90 mm anterior to bregma (according to the stereotaxic atlas of Paxinos et al., 2000Go). The arm representation of the caudal cingulate motor area (M4) was localized in the lower bank of the cingulate sulcus posterior to a coronal plane established at the genu of the arcuate sulcus (Biber et al., 1978Go; Muakkassa and Strick, 1979Go; Hutchins et al., 1988Go; Shima et al., 1991Go; Morecraft and Van Hoesen, 1992Go; Morecraft et al., 1996Go, 1997; Nimchinsky et al., 1996Go; Tokuno et al., 1997Go; Shima and Tanji, 1998Go). Specifically, this area is positioned between coronal levels –3.15 mm posterior to bregma and –6.75 mm posterior to bregma (according to the stereotaxic atlas of Paxinos et al., 2000Go).

Following localization, the respective arm representations were injected with an anterograde tracer (Table 1) using a Hamilton microsyringe held in a specially designed microdrive that was attached to a stabilized electrode micromanipulator (Kopf Instruments, Model 1460, Tujunga, Calif., USA). The anterograde tracers used were 10% biotinylated dextran amine (BDA), 10% lucifer yellow dextran (LYD), 2.5% Phaseolus vulgaris leucoagglutinin (PHA-L), 10% fluorescein dextran (FD) and 10% fluoro ruby (FR). The FD injectate was composed of an equal mixture of 3000 and 10 000 molecular weight volumes, as was the FR injectate. Graded pressure injections (0.2–0.4 µl per penetration) of anterograde neuronal tracers were made into the various cortical arm representations using the Hamilton microsyringe inserted 2–3 mm below the cortical surface under microscopic guidance. The dural flap was replaced in its original position and closed using 5-0 silk. The bone flap was restored to its original position and anchored securely. Finally, the temporalis muscle was repositioned and sutured to its origin and the skin was closed using 3-0 silk.

Tissue processing
Following a survival period of 21–30 days, each monkey was deeply anaesthetized with nembutal and perfused with 0.9% saline followed by 4% paraformaldehyde in 0.1 M phosphate buffer, then sucrose in 0.1 M phosphate buffer as described previously (Morecraft et al., 2000Goa, 2001). The brain was removed, placed in 30% sucrose in 0.1 M phosphate buffer and stored for 2–4 days at 4°C. Post-mortem photographs and electronic images were taken of the dorsal, ventral, lateral and medial surfaces of the cerebral cortex, brainstem and spinal cord.

The cortex was frozen with dry ice and cut coronally or horizontally (Table 1) on an American Optical sliding microtome (AO 860) at a thickness of 50 µm in cycles of 10. The first section of each cortical, brainstem and spinal cord series was mounted on subbed slides, dried overnight and stained for Nissl substance using thionin (Morecraft and Van Hoesen, 1992Go, 1993, 1998; Morecraft et al., 1992Go, 1993). In case IM 121, the second section in each series was processed for PHA-L using the avidin–biotin (ABC) methodology of Rockland and Virga (Rockland and Virga, 1989Go) (Figs A and 3A). In the other single tracer cases (SDM 6, 7 and 10), the second section in each series was processed for BDA using the ABC labelling procedure in combination with the 3,3'-diaminobenzidine tetrahydrochloride (DAB) staining method (Morecraft et al., 2001Go) (Figs E and D). Therefore, the BDA injection site and labelled cells, axons and axon profiles (boutons) stained by nickel-enhanced DAB appeared blue–black upon visualization (Fig. B). Sections processed without the aid of nickel ammonium sulphate appeared brown (Fig. D and E). Finally, in one series of sections in case SDM 6, BDA was also visualized using the alkaline phosphatase method (Morecraft et al., 1996Go).



View larger version (124K):
[in this window]
[in a new window]
 
Fig. 3 Plate of photomicrographs illustrating examples of fibre bundles studied in the coronal radiata and internal capsule following immunohistochemical visualization. (A) A coronal section illustrating PHA-L-labelled fibres (black) in the coronal radiata following an injection of PHA-L in M1 in case IM 121. The arrows identify fibres coursing inferiorly, toward the superior portion of the internal capsule. The arrowheads identify fibres coursing medially, toward the corpus callosum. (B) FD-labelled fibres (blue) in the posterior limb of the internal capsule following an injection of FD into LPMCd in case SDM 23. This horizontal section corresponds to inferior capsular levels marked by the presence of the anterior commissure (ac). Note the specificity in topography of the labelled field. The inset (bottom right) is an enlargement of fibres from the region marked by the asterisk. (C) Horizontal section depicting FR-labelled fibres (blue) in the internal capsule following an injection in the arm representation of M2 in case SDM 23. Note the close proximity of the fibres to the medially located thalamus (th). (D) Coronal section illustrating BDA-labelled fibres (brown) emanating from the injection site and forming a discrete bundle in the corona radiata following an injection of BDA in the arm area of M2 in case SDM 7. (E) Horizontal section showing densely packed BDA-labelled fibres (brown) sweeping around the head of the caudate within the anterior limb of the internal capsule following an injection of BDA into the arm area of M3 in case SDM 23. (F) LYD-labelled fibres (blue) at mid-levels of the internal capsule following an injection of LYD into the arm area of M4 in case SDM 23. Note the close proximity of the fibres to the reticular nucleus of the thalamus in this horizontal section. ac = anterior commissure; cr = corona radiata; gp = globus pallidus; ICp = posterior limb of the internal capsule; rtn = reticular nucleus of the thalamus; th = thalamus.

 
Monkeys SDM 15, SDM 19 and SDM 23 received injections of multiple anterograde tracers (Table 1). In these experiments, BDA was processed alone in a complete series of tissue sections as indicated above, mounted on glass slides and coverslipped. In an additional set of unprocessed serial tissue sections through the cortex, brainstem and spinal cord, double labelling immunohistochemistry was performed (Morecraft et al., 2001Go). To accomplish this, BDA was reacted first according to the above protocol, with the exception that BDA was always stained brown using DAB in the absence of nickel ammonium sulphate (Figs D and E). After reacting for visualizing BDA, the same tissue sections were then incubated in avidin–biotin blocking reagent (Vector SP-2001), rinsed in 15% normal goat serum in Tris buffer for 2 h then incubated overnight, with the appropriate biotinylated antibody directed against a second neuronal tract tracer (e.g. biotinylated anti-LYD). The tissue was then incubated in a solution of avidin–biotinylated peroxidase complex, rinsed in buffer and colorized blue using the Vector SG peroxidase substrate kit (Vector SK-4700) (Figs B and F, and B, C and F). Thus, BDA was colorized brown and the second tracer (e.g. LYD) was stained blue in the same tissue sections. All sections were rinsed, dehydrated, mounted and coverslipped on glass slides. This process was conducted on another set of serial sections staining BDA brown with DAB, then another anterograde tracer blue (e.g. PHA-L), again using the appropriate antibody and blue reaction process to visualize the second tracer (e.g. biotinylated anti-PHA-L). FR and FD were also processed immunohistochemically in the same manner (e.g. BDA brown then FR blue). In these series, FR and FD labelling was verified by comparing the locations of immunochemically localized reaction products with fluorescent labelling in the non-reacted fluorescent series. The specific biotinylated antibodies used to localize each anterograde tracer were all purchased from Vector Laboratories (Burlingame, Calif., USA). Remaining cortical tissue sections were stained for myelin using the gold chloride method (Schmued, 1990Go) and cytochrome oxidase for evaluating histochemical boundaries of the frontal motor areas (Matelli et al., 1985Go).

Data analysis and reconstruction
Nissl, myelin and immunohistochemically labelled tissue sections were examined under bright-field illumination on an Olympus BX60 or BX51 microscope (Leeds Precision Instruments Inc., Minneapolis, Minn., USA). Data from each tissue section were collected with the use of a conventional Hewlett Packard X-Y plotter (HP-7045B) that was attached to the x–y axes of the microscope stage or the use of a Neurolucida system (Microbrightfield Inc., Colchester, Vt., USA) that was attached to a computer-controlled MAC 2000 motorized microscope stage (Ludl Inc., Hawthorne, USA). The initial phases of data collection entailed plotting the outline of the tissue section, anatomical landmarks such as blood vessels and ventricles, the white and grey matter interfaces, the periphery of the injection sites and transported tracing substances at intervals of 250–500 µm. Nissl- and myelin-stained sections were used for determining cytoarchitecture and grey and white matter boundaries. Publication quality images of injection sites and labelled fibres were captured using an Olympus PM20 35 mm camera (Leeds Precision Instruments Inc., Minneapolis, Minn., USA) mounted on the video port of the Olympus BX60 or BX51 microscope (Figs and ). Photographic montages of the injection sites and labelled fibres (Figs and ) were developed by digitizing original 35 mm Kodachrome slides and using Adobe PhotoShop 5.0 (Adobe Systems Inc., San Jose, Calif., USA) and a Hewlett Packard Desk Jet 1220c colour printer (Hewlett Packard, Palo Alto, Calif., USA) or Kodak Digital Science 8650 PS colour printer (Eastman Kodak Co., Rochester, NY, USA).

Cortical and subcortical reconstructions were accomplished using original metrically calibrated surface images of the cerebral hemisphere obtained prior to serial sectioning. Surface renderings illustrating the injection sites were constructed from individual plotted tissue sections and transferred onto lateral and medial views of the surface drawings (Figs 4 and ). Section spacing, sulcal patterns and subcortical structures were used for precise alignment. The overall process enabled the portrayal of the location of each injection site as well as the topographical distribution of transported label on the cortical surface. Tissue section reconstruction for line illustrations (Figs –) was accomplished using Adobe Illustrator 8.0 (Adobe Systems Inc., San Jose, Calif., USA) by electronically transferring plots from the Neurolucida system to the Illustrator program or importing scanned images of original Hewlett-Packard X-Y plottings using a ScanMaker 9600XL (Microtek Lab, Inc., Redondo Beach, Calif., USA). Only fibre bundles containing fibres ending in the spinal cord are described in this report and illustrated in the figures for clarity of presentation (Figs , and ). Thus, it is important to note that the fibres ending in the spinal cord that are identified in the present report are intermingled with other corticofugal projections such as the corticomesencephalic, corticopontine and corticomedullary systems.



View larger version (29K):
[in this window]
[in a new window]
 
Fig. 4 Line drawing illustrating the general location of 12 different injection sites used to study the organization of the corticofugal fibres in the corona radiata and internal capsule. The location of the five injection sites evaluated in case SDM 23 are depicted separately in Fig. 6. as = arcuate sulcus; cf = calcarine sulcus; cgs = cingulate sulcus; cs = central sulcus; IM = Iowa monkey; ips = intraparietal sulcus; lf = lateral fissure; LPMCd = dorsal lateral premotor cortex; LPMCv = ventral lateral premotor cortex; ls = lunate sulcus; M1 = primary motor cortex; M2 = supplementary motor cortex; M3 = rostral cingulate motor cortex; M4 = caudal cingulate motor cortex; poms = medial parieto-occipital sulcus; ros = rostral sulcus; SDM = South Dakota monkey; sts = superior temporal sulcus.

 
Injection site analysis
In all cases presented, the injection sites were evaluated for somatotopic affiliation on the basis of multiple criteria. (i) The location and peripheral extent of the injection sites made in M1, M2, LPMCd and LPMCv were reconstructed in direct relation to the electrophysiological stimulation map (Fig. , see enlargements). In the cingulate region, the location of the arm areas was determined anatomically using the anterior commissure, spur of the arcuate sulcus and previously identified stereotaxic reference points. (ii) Matching Nissl and cytochrome oxidase sections were used to evaluate the histological boundaries of each motor area (Barbas and Pandya, 1987Go; Preuss and Goldman-Rakic, 1991Go; Matelli et al., 1985Go, 1991). (iii) Corticocortical interconnections amongst the various injection sites were evaluated for topographical reciprocity. For example, cortical arm representations are interconnected selectively, as are cortical face representations (Pandya and Vignolo, 1971Go; Muakkassa and Strick, 1979Go; Morecraft and Van Hoesen, 1988Go, 1992; Luppino et al., 1994Go; Nimchinsky et al., 1996Go; Tokuno et al., 1997Go; Wang et al., 2001). (iv) Descending projections from each injection site to the spinal cord were analysed. Injection site involvement of cortical arm areas was verified by the presence of direct terminal projections to the brachial spinal cord (Fig. ) (Biber et al., 1978Go; Hutchins et al., 1988Go; Dum and Strick, 1991Go; Galea and Darian-Smith, 1994Go; Rouiller et al., 1994Go; Luppino et al., 1994Go; Morecraft et al., 1997Go). Direct continuity was established between fibre bundles studied in the CR and IC with fibre bundles in the cerebral peduncle, pontine pyramidal tract, medullary pyramids and corticospinal tract to support the assumption that the localized bundles identified in the CR and capsular region specifically contributed to the corticospinal projection.

Definition of anatomical terminology
The peri- and paraventricular white matter was subdivided according to the nomenclature of Alexander and colleagues (Naeser et al., 1982Go; Alexander, 1989Go). Superior and inferior limits of the IC were defined according to the maps of Martin and Bowden (1996Go, 2000). The general location of the anterior limb, genu and posterior limb of the IC were defined in consultation with the atlas of Roberts and colleagues and Carpenter and Sutin (Carpenter and Sutin, 1983Go; Roberts et al., 1987Go) (Fig. 9). The anterior limb (ICa) occupied coronal levels anterior to and including the anterior commissure. The genu (ICg) was situated in coronal levels immediately caudal to the anterior limb. The posterior limb (ICp) was located immediately caudal to the genu and bordered laterally by the striatum and medially by the thalamus. The ICp was subdivided further into quarters (Ross, 1980Go).



View larger version (30K):
[in this window]
[in a new window]
 
Fig. 9 Summary diagram illustrating the basic organization of the corticofugal projection at superior and inferior levels of the internal capsule in case SDM 23. The individual corticofugal pathways in the IC were found to be widespread, partially overlapping and topographically organized. Descending fibres from the rostral cingulate (M3), supplementary (M2), dorsal lateral pre- (LPMCd), caudal cingulate (M4) and primary (M1) motor cortices occupied rostral to caudal positions of the internal capsule, respectively, at both superior (left) and inferior (right) levels. As each fibre system progressed inferiorly within the IC, all fibres shifted posteriorly to lie within the posterior limb (see also Fig. 6 sections I and J). M2 and LPMCd fibres were found to overlap extensively, as were fibres from M4 and M1. On a comparative basis, data from SDM 15 indicate that the projection from LPMCv would overlap with the projection from LPMCd with a slight anterior displacement. These data suggest that superior capsular lesions are likely to correlate with more favourable levels of functional recovery due to the widespread nature of arm representation (left). In contrast, the extensive overlap and comparatively condensed organization of arm representation at inferior levels (right) suggest that capsular lesions seated inferiorly are likely to correlate with poorer levels of recovery of upper limb movement. Since the origin of the corticospinal projection is densest in the lateral motor areas, our findings also suggest that motor deficit severity is likely to increase as a lesion occupies progressively posterior locations of the internal capsule. AC = anterior commissure; GP = globus pallidus; ICa = anterior limb of the internal capsule; ICg = genu of the internal capsule; ICp = posterior limb of the internal capsule; LPMCd = dorsal lateral premotor cortex; LPMCv = ventral lateral premotor cortex; M1 = primary motor cortex; M2 = supplementary motor cortex; M3 = rostral cingulate motor cortex; M4 = caudal cingulate motor cortex; PC = posterior commissure; Pu = putamen; Th = thalamus.

 

    Results
 Top
 Summary
 Introduction
 Material and methods
 Results
 Discussion
 References
 
All injection sites presented in this report involved primarily the arm representation of the intended motor area. This was verified by each injection site: confinement to the physiologically defined arm area (e.g. Fig. ); cytoarchitectural and histochemical correlate; gross anatomical location; projections to other cortical arm areas; and the presence of prominent labelling in brachial levels of the spinal cord (e.g. Fig. ). In all cases, heavy labelling was found in the brainstem (e.g. reticular formation and pontine grey), but no labelling occurred at lumbosacral levels of the spinal cord. In cases with multiple injections of neuronal tract tracer (Table 1; see SDM 15, SDM 19 and SDM 23), the injection sites were all found to be interconnected reciprocally. In case SDM 23, the injection site in LPMCd also labelled a few terminals in the facial nucleus, indicating minimal involvement of corticobulbar projection neurones. However, heavy labelling occurred at the cervical and brachial spinal levels (Fig. ). Similarly, the BDA injection site in SDM 10 (M4) and the FD injection site in SDM 15 (LPMCv) gave rise to light labelling in the facial nucleus but comparatively much heavier labelling at cervical and brachial levels of the spinal cord. Finally, the FD injection involved primarily LPMCv, but a small portion spread dorsally to involve the ventral rim of area LPMCd based upon cytoarchitectural and histochemical analysis. The results of cases SDM 15 and SDM 23 were similar, with the exception of the course of the projection from the lateral premotor region. The specific differences are reported below.

Trajectory of M1 arm fibres through the corona radiata and internal capsule
All injection sites involving the hand/arm representation of M1 gave rise to an abundant number of labelled fibres which immediately streamed inferiorly (Figs A and A). As the concentrated aggregates descended from the arm area, they turned gradually to assume a medial and slightly rostral course (Fig. B). The fibres continued on their descent within the CR and curved caudally to arch over the putamen and enter the junctional region between the CR and IC [Figs C (M1), and D and E], then the ICp proper (Fig. E and F). At superior thalamic levels, the fibres descended vertically, occupying the second and third quarters of the ICp, being placed slightly lateral from the centre [Figs D and E (M1), and E and F]. As the fibres continued their descent to mid-thalamic levels, they continued to occupy the same general location, with the majority of fibres residing in the third quarter of the ICp (Fig. G and H). At inferior thalamic levels, the fibres moved slightly anteriorly to occupy both the third and second quarters of the ICp equivalently (Fig. I and J). At these levels, the ICp appeared to elongate, possibly due to the posterior extension of the thalamus and incorporation of inferior parietal, temporal and occipital fibres which join the ICp at levels just above, and including lateral geniculate nucleus (Schmahmann and Pandya, 1992Go). In the medial–lateral dimension, the fibres occupied the lateral and central region of the capsule (Fig. G– I). Of all fibre systems studied, M1 fibres consistently occupied the most posterior position within the IC. However, extensive overlap occurred with the descending fascicle from M4. At the superior border of the cerebral peduncle, M1 fibres occupied the most caudal and lateral position, again being intermingled extensively with fibres from M4 (Fig. K).

Trajectory of LPMC arm fibres through the corona radiata and internal capsule
Labelled fibres from the arm area of LPMCd destined for the spinal cord emerged inferiorly from the injection site (SDM 23). On their descent toward the IC, they coursed medially then caudally in the subcortical white matter of the CR (Fig. B–D). Throughout this entire segment, LPMCd fibres were positioned immediately anterior to fibres emanating from the arm representation of M1. As the LPMCd labelled axons continued medially and inferiorly, they arched over the putamen and abruptly deviated caudally in the subcortical white matter to reach and join the middle of the superior portion of the IC (Fig. D and E). At this level, the fibres were positioned caudal to M2 fibres and anterior to M1 and M4 fibres (Fig. E and F). Labelled axons continued to descend vertically, occupying the first and second quarters of the posterior limb, positioned slightly lateral from the centre (Fig. F). At mid-thalamic levels, the bundle continued to progress inferiorly in the same general position of the posterior limb and shifted slightly, from a medial location to a more central location (Fig. G and H). At inferior thalamic levels, fibres from LPMCd occupied primarily the second quarter of the ICp (Figs 3B, I and J). At this level, LPMCd fibres intermingled extensively with fibres from M2 and minimally with the anterior shell of the fibre bundle from M4. At the superior edge of the cerebral peduncle, LPMCd fibres were located centrally where they overlapped completely with fibres from M2 (Fig. K).

The projection from the dorsal portion of LPMCv (SDM 15, see Fig. ) followed a similar course, with the exception of three major findings. First, in the CR, LPMCv fibres were slightly inferior to (i.e. below) fibres from LPMCd, with apparent overlap in these pathways. Secondly, at superior capsular levels, the LPMCv projection involved the ICg. Thirdly, when the LPMCv fibres entered the ICp inferiorly, they appeared to be slightly rostral to the LPMCd location found in case SDM 23, but potentially with extensive overlap.

Trajectory of M2 arm fibres through the corona radiata and internal capsule
Labelled fibres from the arm area of M2 emanated laterally from the injection site and curved abruptly to assume an inferior and posterior course (Figs B, 3D and A–C). Progressing caudally, the fibres gradually descended in the CR, passing over the caudal region of the head of the caudate [Figs D, A (M2) and D] to enter the posterior portion of the ICa [Figs B (M2) and E]. At this location, the fibres formed a bundle lying adjacent to the lateral surface of the caudate. As M2 labelled axons descended in the superior portion of the IC, they shifted posteriorly to enter the ICg and the first quarter of the ICp (Fig. F). In the medial–lateral dimension, the heaviest labelling occurred medially [Figs C, and B and C (M2)]. Progressing through middle and inferior levels of the IC, the M2 fibres entered the first quarter of the ICp then eventually the first and second quarters where they merged with fibres from LPMCd (Fig. G–J). Thus, lateral area 6 fibres (LPMCd) and medial area 6 fibres (M2) followed a similar descending pattern once they left the CR and entered the IC. Throughout the capsular region, M2 fibres consistently occupied a position posterior to M3 fibres and overlapped slightly with the posterior shell of the M3 fascicle. At the superior level of the cerebral peduncle, M2 fibres occupied an intermediate position, maintaining extensive overlap with fibres from LPMCd (Fig. K).

Trajectory of M3 arm fibres through the corona radiata and internal capsule
An intense aggregate of labelled fibres left the arm area of M3 laterally, and passed over the cingulum bundle (Fig. C and D). Labelled axons then immediately turned posteriorly to form a densely packed and highly organized bundle oriented in a relatively horizontal plane (Figs D, and C and D). Progressing caudally, the fibres gradually descended in the CR passing over, and partly within, the anteriolateral periventricular region (region 10 of Alexander, 1989Go) then the anterior third of the superior paraventricular white matter (region 11 according to Alexander, 1989Go) [Figs A (M3) and D]. The fibres continued caudally and laterally to pass over the head of the caudate to enter the middle portion of the ICa [Figs C, A (M3), and D and E]. At this location, the fibres were immediately adjacent to, and partially lying against, the head of the caudate nucleus (Fig. E), thus placing the primary distribution of labelled axons medially to the centre of the IC [Figs B (M3), and E and F]. As the M3 axons continued to descend in the superior portion of the IC, they shifted posteriorly to enter the posterior portion of the ICa and ICg (Fig. F). Progressing to mid-thalamic levels, the fibres occupied the ICg as well as the first quarter of the ICp (Fig. G and H). At inferior thalamic levels, labelled fibres from M3 occupied the first quarter of the ICp (Fig. I and J). Throughout the entire capsular course, M3 fibres occupied the most anterior position of all arm representations investigated (Fig. E–J). Similarly, when they reached the most superior level of the cerebral peduncle of the midbrain, they maintained the most anterior and medial position of all arm representations studied (Fig. K).

Trajectory of M4 arm fibres through the corona radiata and internal capsule
Labelled fibres from the arm area of M4 destined for the spinal cord emerged laterally from the injection site and passed over the cingulum bundle [Fig. B (M4)]. Much like the initial trajectory found for M3 fibres, the M4 fibres immediately turned posteriorly in the subcortical white matter to form a densely packed aggregate oriented in a relatively horizontal plane (Fig. C and D). Progressing caudally, the fibres gradually descended in the CR, passing over and partly within the middle third of the superior periventricular region [Fig. B (M4)] (region 12 of Alexander, 1989Go). The fibres continued laterally to pass over the anterior region of the body of the caudate [Figs C (M4) and D] to enter the second quarter of the ICp, immediately caudal to fibres from LPMCd (Fig. E and F). In the medial lateral dimension, the fibres were placed centromedially in the IC (Fig. F). As the fibres continued their descent to mid- and inferior thalamic levels, they progressed caudally to occupy the second and third quarters of the ICp where they overlapped extensively with M1 fibres (Fig. G and H). At the superior border of the cerebral peduncle, M4 fibres occupied the most caudal and lateral position where they overlapped extensively with M1 fibres (Fig. K).

Summary of results
At the level of the CR, lateral motor area fibres (from M1, LPMCd and LPMCv) arched over the putamen as classically described (Fig. ). LPMCd and LPMCv fibres were positioned anterior to M1 fibres. Medial motor area fibres (M3, M2 and M4) arched over the caudate nucleus (Figs E and ). M3, M2 and M4 fibre bundles maintained anterior to posterior orientation, respectively. As the various fibre bundles entered the superior aspect of the IC, the M3, M2, LPMCv, LPMCd, M4 and M1 motor cortices occupied anterior to posterior positions, respectively (Fig. , left). Minimal, but notable overlap occurred between adjacent pathways. Inferiorly, this general orientation was retained; however, overlap between adjacent bundles gradually increased, particularly amongst area 6 fibre systems (M2, LPMCv and LPMCd) as well as M4 and M1 fibres. A prominent feature and potential principle of frontal and cingulate supratentorial corticofugal fibre organization was that all pathways shifted progressively, to a more posterior position, as they descended through the IC (Fig. , right). Eventually, at horizontal levels including the anterior commissure, all pathways occupied the posterior limb, again maintaining their general anterior to posterior topographical relationship (Figs I and , right).


    Discussion
 Top
 Summary
 Introduction
 Material and methods
 Results
 Discussion
 References
 
Arm representation in the CR and IC has been a subject of great interest in the fields of neurology and neurosurgery for more than a century (Charcot, 1883Go; Bennett and Campbell, 1885Go; Beevor and Horsley, 1890Go; Dejerine and Dejerine-Klumpke, 1901Go; Lassek, 1954Go; Englander et al., 1975Go; Hanaway and Young, 1977Go; Davidoff, 1990Go; Adams et al., 1997Go). The basis for this interest lies, in part, in its practical application for assessing the functional status of patients following subtotal brain injury and predicting their potential for motor recovery. Equally important is a detailed understanding of corticofugal fibre organization for guiding intraoperative stereotaxic navigation and designing surgical approaches to manage movement disorders. As a result, the published observations are abundant and range considerably from widespread arm representation involving the anterior limb, genu and posterior limb, to localized positions confined to various subsectors of the posterior limb. Our results support and unify many of the seemingly contradictory observations when considering two major points. First is recognizing the fact that corticofugal fibres innervating the brainstem and brachial spinal cord arise from multiple regions of the frontal and cingulate cortices. Secondly is acknowledging the fact that the trajectories of all corticofugal pathways examined in the present study are not fixed as they descend through their supratentorial course. Thus, corticofugal redundancy in a pure structural sense, as well as location along the neuraxis, are important factors when comparing the present findings with conclusions drawn in previous reports. Specifically, we demonstrate that the multiple corticofugal pathways contributing to the pyramidal tract travel in discretely organized and partially overlapping bundles throughout localized, but widespread regions of the CR and IC (Figs and ). These observations may be useful for interpreting the effects of intraoperative subcortical stimulation and designing surgical approaches for treating medically intractable movement disorders. Our findings also provide a detailed, comparative and sequential understanding of frontal and cingulate corticofugal fibre organization in the non-human primate brain that may assist in evaluating corticofugal fibre integrity in patients with localized supratentorial damage.

Localization of M1 arm representation
As anticipated, our results support previous anatomical, physiological and clinical studies suggesting that the arm area of the precentral motor region gives rise to axons which pass through the ICp (Charcot, 1883Go; Bennett and Campbell, 1885Go; Beevor and Horsley, 1890Go; Probst, 1898Go; Mellus, 1899Go; Dejerine and Dejerine Klumke, 1901Go; Levin, 1936Go; Barnard and Woolsey, 1956Go; Guiot et al., 1959Go; Gillingham, 1962Go; Bertrand et al., 1963Go, 1965; Liu and Chambers, 1964Go; Bertrand, 1966Go; Smith, 1967Go; Englander et al., 1975Go; Hardy et al., 1979Go; Ross, 1980Go; Tredici et al., 1982Go; Dawnay and Glees, 1986Go; Alexander, 1989Go; Danek et al., 1990Go; Fries et al., 1993Go; Seitz et al., 1994Go; Axer and Keyserlingk, 2000Go; Duffau et al., 2000Go; Pineiro et al., 2000Go; Yoshimura and Kurashige, 2000Go). Specifically, we found M1 fibres to be concentrated in the second and third quarter of the ICp (Figs and ). This confirmed the general anatomical location of the monkey M1 projection previously reported, which suggests passage in the middle third of the posterior limb (Mellus, 1899Go; Fries et al., 1993Go). However, our report identifies the discrete and progressive topography of the M1 pathway in relation to multiple corticofugal fibre pathways and major anatomical landmarks located above the midbrain. In agreement with Fries, we did not find M1 fibres to overlap with M2 fibres. However, we found extensive overlap between M1 fibres and fibres from the caudal cingulate motor region (M4) which was not studied by Fries and colleagues. The general anatomical characteristic of segregated, but partially overlapping, organization has also been found to occur between specific corticopontine projections travelling through the IC from the parietal, temporal and occipital association areas (Schmahmann and Pandya, 1992Go; Schmahmann, 2000Go). Importantly, our findings parallel closely those demonstrated by whole brain dissection in the human brain (Ross, 1980Go) which suggest that the projection from the precentral gyrus enters the posterior limb and shifts from the second quarter of the posterior limb superiorly, to the third quarter inferiorly. Since our finding is similar to the organization of the M1 projection in the human IC (Ross, 1980Go), it is possible to suggest cautiously that structural homologies are likely to occur for other pathways contributing to the corticofugal projection.

Localization of LPMC arm representation
Our study demonstrates that LPMCd fibres pass primarily through the ICp, occupying the first and second quarters superiorly, then shifting to the second quarter inferiorly (Figs and ). It has been shown previously that the capsular projection originating from cortex located more ventrally in lateral area 6 (corresponding to the ventral lateral premotor region, e.g. LPMCv) initially travels through the genu and anterior quarter of the posterior limb (Fries et al., 1993Go). Our results from case SDM 15, which also had an injection located primarily in LPMCv but in a location slightly dorsal to the injection depicted in the study of Fries, are in general agreement with these findings. On a comparative basis, these patterns indicate that LPMCv gives rise to fibres that may course slightly rostral to LPMCd fibres, with potential overlap. Electrophysiological evidence indicates that face movements are heavily represented in the ventral portion of the lateral premotor cortex and arm movements in the dorsal portion of this cortex, located around the genu of the arcuate sulcus (Gentilucci et al., 1988Go; Rizzolatti et al., 1988Go; Godschalk et al., 1995Go; Preuss et al., 1996Go). Although there is some overlap in movement representation in areas F4 and F5, it follows then that a general somatotopy can be suggested. Fibre bundles travelling in the IC which originate from the face region of the lateral premotor cortex may lie anterior to fibres from the arm area of the lateral premotor cortex, i.e. much like the classic interpretation for the somatotopic organization of capsular fibres from M1, where fibres from the face area of M1 are thought to be positioned rostral to fibres originating from the arm representation of M1 (Bertrand, 1966Go).

Localization of M2 arm representation
We found the projection from the arm area of M2 to enter the posterior part of the ICa then shift gradually, to occupy the ICg then the first and second quarters of the ICp (Figs and ). These observations parallel those of Fries and colleagues who found the M2 projection passing through the anterior limb then the genu at superior capsular levels (Fries et al., 1993Go). Our findings differ slightly in that anterior limb labelling extended more anteriorly in the experimental case of Fries. This anterior extension may be attributable to injection site involvement of the M2 face area and pre-supplementary area, which lie in the superior frontal lobule, but anterior to the M2 arm representation (Fig. A and B). Our observations extend these findings by illustrating the position of M2 fibres en route to the internal capsule (e.g. within the CR) as well as through middle and inferior capsular levels where the M2 projection