THE JOURNAL OF COMPARATIVE NEUROLOGY 504:149 –167 (2007)

Localization of Arm Representation in the Cerebral Peduncle of the Non-Human Primate ROBERT J. MORECRAFT,* DAVID W. MCNEAL, KIMBERLY S. STILWELL-MORECRAFT, ZELJKO DVANAJSCAK, JIZHI GE, AND PRESTON SCHNEIDER Division of Basic Biomedical Sciences, Laboratory of Neurological Sciences, The University of South Dakota, Sanford School of Medicine, Vermillion, South Dakota 57069

ABSTRACT Motor deficit severity and the potential for recovery in patients with brain injury depend on the integrity of descending corticofugal projections. Clinical assessment of these conditions following subtotal brain trauma requires a comprehensive understanding of the anatomical structures involved in the lesion as well as those structures that are spared. To assist in this endeavor, we investigated motor fiber organization in the crus cerebri of the cerebral peduncle (ccCP) in the rhesus monkey. Fibers originating from the arm representations of the primary (M1), supplementary (M2), rostral cingulate (M3), caudal cingulate (M4), dorsolateral pre- (LPMCd) and ventrolateral pre- (LPMCv) motor cortices were studied. The projections from the frontal and cingulate motor cortices formed descending longitudinal bundles that occupied the medial three-fifths of the ccCP at superior and middle levels. Although considerable overlap characterized these corticofugal projections, a general topography was discernable. Fibers from M1 and M4 occupied the central subsector of the ccCP, and fibers from M3 resided medially. The main distribution of LPMCd, LPMCv, and M2 fibers occupied the centromedial region and overlapped extensively. Progressing inferiorly, all fiber bundles in the central and centromedial sectors gradually extended medially, and overlap increased. A common location of fiber passage occurred at the midbrain-pontine isthmus where all of the fiber bundles overlapped. Our findings indicate that the widespread distribution of corticofugal motor projections may account for the favorable levels of motor recovery that accompany subtotal midbrain injury. At superior and mid-levels of the ccCP anteromedial lesions may disrupt projections from M3, whereas anterolateral lesions may disrupt projections from M1 and M4. Fibers from M2, LPMCv, and LPMCd may be compromised to some degree in both situations. The compact and commixed nature of motor fiber organization at inferior levels and the midbrain-pontine isthmus suggests a vulnerable region of passage for comprehensive disruption of frontal and cingulate corticofugal projection fibers. J. Comp. Neurol. 504:149 –167, 2007. © 2007 Wiley-Liss, Inc. Indexing terms: corticofugal; corticospinal; tractography; stroke; brainstem

The crus cerebri occupies the ventral region of the midbrain cerebral peduncle and is of central importance for maintaining a critical corticofugal projection system from the primate cerebral cortex. Studies reporting the effects of peduncular lesions on residual motor capacity indicate that damage to the cerebral peduncle produces variable degrees of motor dysfunction, ranging from the highly disabled condition of “locked-in syndrome” to remarkable levels of motor recovery (Walker, 1949, 1952, 1955; Meyers, 1951, 1956a,b; Hamby, 1953; Broager, 1955; Bucy, 1957; Bucy and Keplinger, 1961; Bucy et al., 1964, 1966; Walker and Richter, 1966; Karp and Hurtig, 1974; Bauer © 2007 WILEY-LISS, INC.

Grant sponsor: National Institutes of Health; Grant numbers: NS 046367 and NS 33003; Grant sponsor: The South Dakota Spinal Cord Injury and Traumatic Brain Injury Research Council (to R.J.M.). *Correspondence to: Robert J. Morecraft, Ph.D., Laboratory of Neurological Sciences, Division of Basic Biomedical Sciences, The University of South Dakota, Sanford School of Medicine, Vermillion, SD 57069. E-mail: [email protected]. Received 16 January 2007; Revised 8 May 2007; Accepted 26 May 2007 DOI 10.1002/cne.21438 Published online in Wiley InterScience (www.interscience.wiley.com).

The Journal of Comparative Neurology. DOI 10.1002/cne

150

R.J. MORECRAFT ET AL.

et al., 1979; Meienberg et al., 1979; Ho, 1982; Pendl and Vorkapic, 1988; Chia, 1991; Gaymard et al., 1991; Bogousslavsky et al., 1994; Duncan and Weindling, 1995; Park et al., 1997; Dabby et al., 2004; Kim and Kim, 2005). Despite these observations, the underlying structural substrates responsible for this wide range of clinical outcomes remain obscure. Based on mounting evidence suggesting that motor recovery depends on corticofugal integrity (Fries et al., 1993; Weiller et al., 1993; Binkofski et al., 1996, 2001; Pineiro et al., 2000; Shelton and Reding, 2001; Newton et al., 2006; Ward et al., 2006), it would be reasonable to speculate that the degree of motor deficit in patients with midbrain injury may also correlate with the extent of axonal involvement of cortical motor projections in the province of the crus cerebri of the cerebral peduncle (ccCP). However, the precise organization of corticofugal representation from the various motor cortices in the midbrain is unknown. The need to understand this organization is further underscored by evidence from patients with localized midbrain infarction, or selective surgical interruption of the ccCP, suggesting that “pyramidal fibers” may be more diffusely represented in this critical brainstem region than classically interpreted (Bucy et al., 1964; Kim and Kim, 2005). Our current knowledge of corticofugal organization of motor pathways in the crus cerebri of the midbrain cerebral peduncle is insufficient considering that what is known provides little insight for interpreting the wide variety of clinical manifestations accompanying midbrain insult. From an anatomical standpoint, motor tract organization has been variously depicted in the literature, and its representation continues to be based on classical interpretations (Barnes, 1901; Mettler, 1942; Meyers, 1951;

Walker, 1955; Bucy et al., 1964; Brodal, 1981; QuinonesHinojosa et al., 2005). Furthermore, what has been examined is largely limited to the corticofugal fibers originating from the lateral surface of the cerebral hemisphere, with most authors emphasizing fiber representation from the primary motor cortex (M1). In review of this material, authorities on this subject assert that a portion of the corticofugal fibers from the arm area of M1 reside within the central region of the human cerebral peduncle, with the main difference in opinion regarding the degree of representation medial and lateral to this region. The acclamation of a centrally located arm representation would largely be in accord with monkey studies examining the location of fibers in the upper crus cerebri originating from the arm region of M1 (Levin, 1936; Barnard and Woolsey, 1956; Schmahmann et al., 2004; Schmahmann and Pandya, 2006). As for the other cortical motor areas, evidence exists to indicate that corticofugal fibers from the arm region of the dorsal part of lateral area 6 (LPMCd) are located medial to a centrally located M1 tract (Levin, 1936; Barnard and Woolsey, 1956; Kuypers, 1981). However, the degree of overlap shared between these pathways is unclear, and their descending course through the midbrain has yet to be investigated. What is completely deficient from our knowledge is the availability of information on the location of peduncular fibers from the arm regions of the ventral lateral premotor cortex (LPMCv), supplementary motor cortex (M2), rostral cingulate motor cortex (M3), and caudal cingulate motor cortex (M4). Notably, all of these cortical motor areas give rise to substantial corticofugal projections including a corticospinal projection (Catsman-Berrevoets and Kuypers, 1976;

Abbreviations AL AM as bc BDA c C ca cc ccCP cp cf cgs cl CL CM Cs d FD FR hy ic ilas III in ios ips IV l L lf lgn LPMCd

anterolateral vascular territory anteromedial vascular territory aqueduct of Sylvius brachium conjunctivum biotinylated dextran amine caudal central subsector of the crus cerebri caudate nucleus corpus callosum crus cerebri of the cerebral peduncle cerebral peduncle calcarine fissure cingulate sulcus claustrum centrolateral subsector of the crus cerebri centromedial subsector of the crus cerebri central sulcus dorsal fluorescein dextran Fluoro Ruby hypothalamus inferior colliculus inferior limb of the arcuate sulcus oculomotor complex/nerve insula inferior occipital sulcus intraparietal sulcus trochlear nucleus lateral lateral subsector of the crus cerebri lateral fissure lateral geniculate nucleus dorsal lateral premotor cortex

LPMCv ls LYD m M M1 M2 M3 M4 mcp mgn ml oc ot ots pag PHA-L pn pos ps r rn ros rs S1 sc slas sn stn sts v 3v 4v

ventral lateral premotor cortex lunate sulcus lucifer yellow dextran medial medial subsector of the crus cerebri primary motor cortex supplementary motor cortex rostral cingulate motor cortex caudal cingulate motor cortex middle cerebellar peduncle medial geniculate nucleus medial lemniscus optic chiasm optic tract occipital temporal sulcus periaqueductal gray Phaseolus vulgaris-leucoagglutinin pontine nuclei medial parieto-occipital sulcus principle sulcus rostral red nucleus rostral sulcus rhinal sulcus primary somatosensory cortex superior colliculus superior limb of the arcuate sulcus substantia nigra subthalamic nucleus superior temporal sulcus ventral third ventricle fourth ventricle

The Journal of Comparative Neurology. DOI 10.1002/cne

ARM REPRESENTATION IN THE CEREBRAL PEDUNCLE Biber et al., 1978; Murray and Coulter, 1981; Nudo and Masterson, 1990; Dum and Strick, 1991, 1996; Galea and Darian Smith, 1994; He et al., 1995; Morecraft et al., 1997, 2002). It is also unknown whether the various tracts shift in their location as they progress from superior to inferior levels of the midbrain, which has been shown to be an anatomical characteristic of corticofugal fiber representation in the corona radiata and internal capsule in the monkey (Morecraft et al., 2002) and human brain (Ross, 1980; Newton et al., 2006). Indeed, the clinical literature would suggest that the positions of these tracts may not be stagnant as currently interpreted, and the extent of overlap may increase as fibers progress toward inferior levels of the brainstem (Kim et al., 1995). In a previous study, we investigated the trajectory of corticofugal fibers originating from the arm representation of M1, M2, M3, M4, LPMCd, and LPMCv in the corona radiata and internal capsule of the rhesus monkey and found a highly organized pattern of projections that progressively shifted in location and gradually increased in their juxtaposition from superior levels of the corona radiata to inferior levels of the internal capsule (Morecraft et al., 2002). Because recent studies on the human brain have shown significant structural and functional homologies with these trajectories (Lang and Schieber, 2003; Fridman et al., 2004; Wenzelburger et al., 2005; Newton et al., 2006; Raghavan et al., 2006), we extended this effort by defining the organization of frontal and cingulate arm representation in the crus cerebri of the cerebral peduncle in the non-human primate. In contrast to the rostrocaudal associations affiliated with capsular organization, we found a medial to lateral arrangement characterizing peduncular organization. Much like capsular organization, we found widespread fiber representation superiorly with, progressively, an increased degree of overlap that was extensive at the most inferior levels of the crus cerebri. Collectively, these observations provide a detailed template for interpreting the effects of localized lesions involving the human cerebral peduncle and offer essential information for predicting motor recovery following subtotal midbrain injury. These observations are also relevant for interpreting neuroimaging and tractography observations directed toward evaluating the organization of frontal and cingulate corticofugal arm representation in the crus cerebri of the human cerebral peduncle.

MATERIALS AND METHODS The organization of corticofugal fibers arising from six different cortical arm representations was studied in the midbrain crus cerebri in eight rhesus monkeys (Macaca mulatta; Fig. 1, Table 1). Three of these monkeys (SDM 15, SDM 19, and SDM 23) were used in our previous report investigating the organization of arm representation in the coronal radiata and internal capsule (Morecraft et al., 2002). The cortical arm representations studied in the present 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 areas 24c and 24d), and caudal cingulate motor cortex (M4 or areas 23c and 23d) (Woolsey et al., 1952; Muakkassa and Strick, 1979; Barbas and Pandya, 1987; Morecraft and Van Hoesen, 1992; Morecraft et al., 2002 (see their Fig. 1), 2004, 2007). M3 and M4

151

are synonymous with the terms rostral cingulate motor area (CMAr) and caudal cingulate motor area (CMAc), respectively (Morecraft and Tanji, 2007). Each monkey was injected with anterograde tract tracers into multiple arm representations of the frontal and cingulate motor cortices, and fiber organization was evaluated in the crus cerebri (Table 1). Stimulation in M1 and M2 elicited movements that followed a basic somatotopic arrangement, and only arm and trunk movements were encountered following stimulation of the LPMCd. Isolated as well as combined arm and face movements occurred following stimulation of the LPMCv with arm movements predominating dorsally and face movements ventrally. Anatomical localization of the arm areas of M3 and M4 was performed as previously described (Morecraft et al., 2001, 2002).

Surgery, intracortical microstimulation, and tissue processing All experimental and surgical procedures employed in this study followed United States Department of Agriculture, National Institutes of Health, and Society for Neurosciences guidelines for the ethical treatment of experimental animals and were approved by the Institutional Animal Care and Use Committee at The University of South Dakota. For SDM 15, 19, 23, 39, and 43, animal preparation, surgical procedures, intracortical microstimulation, and motor cortex localization were accomplished as previously described (Morecraft et al., 2001, 2002). For SDM 54, 57, and 61, each animal was injected with atropine sulfate (0.06 mg/kg), immobilized with ketamine hydrochloride (10 mg/kg), intubated, and then anesthetized with isofluorane (1–1.5% with surgical grade air/oxygen mixture) administered with the assistance of a mechanical respirator. Following aseptic cortical exposure, the animal was transferred to intravenous ketamine anesthesia for electrophysiological mapping of the frontal motor cortices. Intracortical microstimulation was used to localize the arm area of the M1, LPMCd, LPMCv, and M2. To accomplish this, a tungsten electrode (impedance 0.5–1.5 M⍀) was inserted 200 ␮m below the pial surface and then advanced at 500-␮m intervals. Movements were evoked by using a train duration of 50 ms and pulse duration of 0.2 ms delivered at 330 Hz. Current intensity ranged between 7 and 90 ␮A. Threshold currents were determined, and the evoked movements were recorded when noted by two observers. To assist with injection site placement and reconstruction, digital images were taken of the surgically exposed cortex by using a Cannon EOS 20D 8 mega pixel digital camera and attached Cannon EF17-40 mm F4L USM lens or EF17-100 mm lens. The surgical exposure was made large enough so that the photomicrographs contained important anatomical landmarks such as the central sulcus, arcuate sulcus, posterior tip of the principle sulcus, and superior and inferior precentral sulci. A high-resolution color print was made by using a Hewlett Packard 1220C printer, and the precise location of each stimulation point was recorded directly on the print of the cortical surface in the operating room. The borders between the various somatotopical regions were also marked on the photograph. Once the respective arm representations were localized, they were injected with an anterograde tracer 2.5–3.5 mm below the cortical surface by using a Hamilton microsy-

The Journal of Comparative Neurology. DOI 10.1002/cne

152 ringe (Table 1). The microsyringe was held in a specially designed microdrive attached to a stabilized electrode micromanipulator (model 1460, Kopf Instruments, Tujunga, CA). The penetration site for each injection was also recorded on the digital photograph of the cortical surface to assist in data reconstruction. The anterograde tracers used were 10% biotinylated dextran amine (BDA) in 0.9% saline, 10% lucifer yellow dextran (LYD) in 0.9% saline, 2.5% Phaseolus vulgaris-leucoagglutinin (PHA-L) in 0.1 M phosphate buffer, 10% fluorescein dextran (FD) in distilled water, and 10% Fluoro Ruby (FR) in distilled water (Molecular Probes, Eugene, OR; Table 1). The FD solution was composed of an equal mixture of 3,000 and 10,000 MW volumes, as was the FR solution. After the injections were made, the dura was closed by using 5-0 silk, and then the bone flap was replaced and anchored securely. The temporalis muscle was returned to its origin and sutured in place, and the skin was closed by using 3-0 silk or sterile staples. Each animal was carefully monitored postsurgically throughout the survival period. Buprenorphine (0.01 mg/kg) was administered postoperatively for 48 –72 hours. Twenty four hours prior to the surgery, and for 9 days post surgery, each animal was administered either bicillin L-A or amoxicillin as a preparatory and postoperative antibiotic. Following a survival period of 26 –33 days, each monkey was deeply anesthetized with an overdose of pentobarbital (50 mg/kg or more) and perfused transcardially with 0.9% saline. Saline infusion was followed by 2 liters of 4% paraformaldehyde in 0.1 M phosphate buffer at pH 7.4 (PB) and then 1 liter each of 10% and 30% sucrose in 0.1 M PB for cryoprotection. The brain was removed, placed in 30% sucrose in 0.1 M PB, and stored for 2–5 days at 4°C. Postmortem digital electronic images were taken of the dorsal, ventral, lateral, and medial surfaces of the brain and brainstem for data reconstruction. The tissue was then processed for histochemical and immunohistochemical visualization (Morecraft et al., 2001). The cortices of SDM 15 and 23 were frozen sectioned in the horizontal plane and all other cases in the coronal plane on a sliding microtome (American Optical 860, Buffalo, NY) at a thickness of 50 ␮m in cycles of 10, forming 10 complete series of evenly spaced tissue sections, respectively. Each brainstem was blocked from the central nervous system, frozen with dry ice, and cut horizontally on the sliding microtome at a thickness of 50 ␮m in cycles of 8 or 10, forming 8 or 10 complete series of evenly spaced tissue sections, respectively. For both the cortical and midbrain blocks, one series of tissue sections was mounted on gelatin-coated slides, dried overnight, and stained for Nissl substance by using thionin (Morecraft and Van Hoesen, 1992; Morecraft et al., 2004). A second series was used for fluorescent tracer analysis of FD and FR and was immediately mounted on gelatincoated slides, dried overnight at 4°C, and then coverslipped by using D.P.X. (Aldrich Chemical Company, Milwaukee, WI). Subsequent series of tissue sections were then processed by using single and double label immunohistochemical procedures for visualization of the neural tracers (Figs. 2, 3) (Morecraft et al., 2001, 2002). In all monkeys, one series of tissue sections was used to process BDA alone (single labeling procedure) by using the avidinbiotin (ABC) peroxidase labeling procedure. Briefly, the tissue sections were rinsed in 0.05 M Tris-buffered saline

R.J. MORECRAFT ET AL. (TBS; pH 7.4) and then incubated overnight in TBS with 5% normal goat serum (NGS) and 1.25% Triton X-100. Next, the sections were rinsed in TBS and then incubated in the ABC complex for 4 hours at room temperature. The sections were then rinsed with TBS and incubated in a 0.05% solution of 3,3⬘-diaminobenzidine tetrahydrochloride (DAB) without nickel enhancement for approximately 10 minutes. Subsequently, 30% hydrogen peroxide (H202) was added to the DAB solution, achieving a final H202 concentration of 0.012%. The tissue was incubated in the DAB/H202 solution for another 8 –10 minutes, yielding an insoluble brown reaction product, and immediately placed in TBS to stop the reaction. Following immunohistochemical processing, the BDAstained tissue sections were then rinsed in TBS, mounted on subbed slides, dried over several days, dehydrated in graded alcohol solutions, cleared in citrisolv and coverslipped by using Permount. Finally, additional separate series of tissue sections were used for double labeling immunohistochemistry (i.e., BDA ⫹ FD; BDA ⫹ FR; BDA ⫹ LYD; and BDA ⫹ PHA-L). To accomplish this, BDA was reacted first in a full series of tissue sections according to the above protocol, staining BDA brown. The same tissue sections were rinsed in TBS and then incubated in TBS with 5% NGS and 1.25% Triton X-100 overnight. The tissue sections were then transferred and incubated in 5% NGS in TBS with the appropriate biotinylated antibody directed against a second neuronal tract tracer (e.g., biotinylated anti-FD; Vector, Burlingame, CA) for approximately 40 hours. The antibodies to PHA-L, FR, FD, and LYD were used at dilutions of 1:500, 1:500, 1:500, and 1:300 respectively. The tissue was rinsed in TBS, incubated in a solution of avidin-biotin peroxidase complex for 4 hours at room temperature, rinsed again in TBS, and incubated with the Vector SG peroxidase substrate kit (Vector SK-4700) for approximately 5–10 minutes, yielding a blue reaction product for the second tracer. The sections were rinsed, mounted on glass slides, dried overnight at room temperature, dehydrated, cleared in citrisolv and coverslipped by using Permount. Thus, BDA was stained brown, and the second tracer (e.g. FD) was stained blue in the same tissue sections (Fig. 2B).

Definition of anatomical terminology The midbrain is divided into a right and left cerebral peduncle and a dorsal tectum (Standring, 2005). In the transverse dimension, the dorsal part of the cerebral peduncle is formed by the tegmentum and the ventral part by the crus cerebri, which are separated by the substantia nigra. The tegmenti are continuous across the midline, but the crura are spatially separate, forming part of the interpeduncular fossa. Each crus cerebri is roughly semilunar in shape and contains corticofugal fibers that descend to reach the midbrain, pons, medulla, and spinal cord. The superior limit of the cerebral peduncle is defined at the superior level of the optic tract and the inferior limit of the peduncle by the pontomesencephalic sulcus (Duvernoy, 1999). For the purpose of this report, the crus cerebri of the cerebral peduncle was divided into five anatomical subsectors in the horizontal plane including a medial subsector (M), centromedial subsector (CM), central subsector (C), centrolateral subsector (CL), and lateral subsector (L) (Figs. 3C, 4A,E, 5A,D).

The Journal of Comparative Neurology. DOI 10.1002/cne

ARM REPRESENTATION IN THE CEREBRAL PEDUNCLE

Data analysis and reconstruction Nissl and immunohistochemically labeled tissue sections through the cortex and midbrain were examined under brightfield illumination on an Olympus BX60 or BX51 microscope (Leeds Precision Instruments, Minneapolis, MN). Immunoreactive fibers in the ccCP were plotted by charting individual tissue sections using brightfield illumination on an Olympus microscope (BX-51 or BX-60) attached to a computer-controlled MAC 5000 motorized microscope stage (Ludl Electronic Products, Hawthorne, NY) joined to a Neurolucida neuroanatomical data collection and analysis system (Microbrightfield, Colchester, VT). Each injection site was plotted by outlining the perimeter of dense precipitate that occurred from the center of the injection site out to the edge or fringe. The fringe, in turn, corresponded to the region where the dense precipitate diminished, fibers emanating from the injection site and neuron profiles were clearly distinguishable, and the staining intensity in the neuropil reflected average background levels. The fluorescent stained tissue sections through the cortex and brainstem were visualized under epifluorescence and used to verify the immunohistochemical observations of FD and FR. Nissl-stained sections through the cortex and midbrain were used to identify the cytoarchitectonic organization, which was then superimposed onto the cortical and midbrain plottings. Publication-quality images of injection sites and labeled fibers were captured by using a Spotflex 64 mega pixel camera, (Diagnostic Instruments, Sterling Heights, MI, version 4.6), mounted on an Olympus BX51 microscope. Photographic montages of the injection sites and labeled fibers were created by using Adobe PhotoShop 7.0 (Adobe Systems, San Jose, CA; Figs. 2, 3). Brightness and contrast were adjusted in the images. In addition, in the low-magnification photomicrographs, the dodge tool was used to reduce dark spots around the perimeter and the burn tool to reduce the hotspots toward the center. Midbrain reconstructions (Figs. 4, 5) were accomplished by using Neurolucida data files containing plotted fibers and anatomical landmarks, as well as matching Nissl-stained tissue sections. Cortical reconstructions (Fig. 1) were developed as previously described (Morecraft and Van Hoesen 1992). Publication-quality line illustrations were created by using Adobe Illustrator 10.0 (Figs. 1, 4 – 6).

RESULTS Histological analysis revealed that all frontal lobe injection sites were confined to the intended hand/arm representation of each motor area by comparing the locations of the needle tract and dense precipitate surrounding the tract with the locations of the stimulation points and anatomical landmarks on the cortical reconstruction (Figs. 1, 2). In terms of cingulate cases, analysis of the injection sites in cases SDM 15 and SDM 23 has been previously reported (Morecraft et al., 2002). The M3 injection site in cases SDM 39 and SDM 43 involved both face and arm areas (Table 1), as evidenced by axon terminal labeling in both the facial nucleus and spinal cord. Finally, the M4 injection site in SDM 43 involved the head and arm areas of M4 located in the lower bank of the cingulate sulcus, which was supported by the presence of hypoglossal terminals and corticospinal terminals at levels C1–T1. This injection also involved a small portion of cortex lo-

153

cated on the medial wall of the cingulate gyrus (area 23b) (Fig. 2F). All injection sites reported in this study gave rise to an abundant number of labeled fibers at all levels of the crus cerebri of the cerebral peduncle (Fig. 3). These fiber bundles were traced inferiorly through the pons, medulla, and upper spinal cord and were found to issue axonal branches with terminal-like particles at each level. Below we report our observations of fiber distribution in the ccCP from the lateral motor areas (M1, LPMCd, and LPMCv) followed by the medial motor areas (M2, M3, and M4). Tissue sections from SDM 39, which were used to study the M3 projection, were available only at superior midbrain levels and the upper levels (C1–T1) of the spinal cord.

Trajectory of M1 arm fibers through the crus cerebri The descending projection from M1 was studied in six experimental cases (Figs. 1, 2A, 3A,E, Table 1). Fibers from the arm area of M1, which occupied the caudal third of the posterior limb of the internal capsule, abruptly shifted to occupy primarily the central region of the crus cerebri at superior levels with minimal centromedial involvement in some cases. In this region, labeled fibers were densely localized (Figs. 3A, 4A,B, 5A,B). As the fibers continued their descent, the main distribution maintained a central position with increased involvement of the centromedial subsector (Figs. 4C–E, 5C,D). Few fibers occupied the medial part of the centromedial subsector. At inferior levels of the crus cerebri, the main distribution of labeled fibers migrated medially to occupy primarily the medial part of the central subsector, the centromedial subsector, and the lateral portion of the medial subsector (Figs. 4F,G, 5E,F). Upon reaching the midbrain-pontine isthmus, labeling was diffuse and involved the medial half of the crus cerebri (Figs. 3E, 4H, 5G). At this location, fibers were found to intermingle with all the other pathways investigated. This commixed pattern of fiber organization continued as the crus cerebri began to fractionate into white matter islets surrounded by clusters of gray matter in the superior pontine tegmentum.

Trajectory of LPMCd and LPMCv arm fibers through the crus cerebri The projection from the lateral premotor cortex was studied through the crus cerebri from seven different injection sites (Table 1). For direct comparison of the projection from the LPMCd with the projection from the LPMCv, injections of different compounds were made in both motor areas in the same cerebral hemisphere in cases SDM 57 and 61 (Figs. 1, 2B,C, 3B, 5). Our experiments demonstrated that at superior levels of the midbrain, the corticofugal projection from the arm area of the LPMCd and LPMCv overlapped extensively and occupied primarily the centromedial region of the crus cerebri with limited involvement of the medial segment of the central subsector and lateral portion of the medial subsector (Figs. 3A,B, 4A,B, 5A,B). Of these peduncular regions, the densest distribution of fibers occupied the centromedial subsector (Fig. 3B). At mid-levels of the midbrain, the descending bundles from both the LPMCd and LPMCv occupied the centromedial subsector with, again, limited presence in the central subsector (Figs. 4C–E, 5C,D). However, fiber presence increased in the lateral region of the medial sector (Fig. 4E). In terms of density, the fibers were dis-

The Journal of Comparative Neurology. DOI 10.1002/cne

154

R.J. MORECRAFT ET AL.

Fig. 1. Line drawings of the lateral and medial cortical surfaces of cases SDM 23 and 54 and the lateral cortical surface of cases SDM 57 and SDM 61 depicting the locations of anterograde tract tracer injections. Yellow injection sites represent the location of the neural tract tracer lucifer yellow dextran (LYD), light blue represents biotinylated dextran amine (BDA), green represents fluorescein dextran (FD), red represents Fluoro Ruby (FR), and dark blue corresponds to Phase-

olous vulgaris-leucoagglutinin (PHA-L). The corticofugal trajectories of labeled axons emanating from these injection sites were studied in the crus cerebri of the cerebral peduncle. The central sulcus, arcuate sulcus, and cingulate sulcus have been opened (see dashed line) to show the cortex lining the banks of each sulci. For abbreviations, see list.

persed uniformly between the centromedial subsector and the lateral part of the medial subsector. At inferior levels of the midbrain, fibers continued to migrate medially and overlap with the other motor area projections intensified

(Figs. 4F,G, 5E,F). At this level, labeled fibers occupied nearly the entire medial half of the crus cerebri (Figs. 4G, 5F). As the LPMCd and LPMCv fibers approached and proceeded through the region of the midbrain-pontine

Fig. 2. Plate of low-power digital photomicrographs illustrating representative examples of injection sites in different cortical motor areas following tissue processing for immunohistochemical visualization. A: Coronal section depicting the injection site of LYD (blue reaction product) in the arm representation of M1 in case SDM 54. The anatomical orientation (bottom right) of this panel also applies to B–D. B: Coronal section depicting the BDA injection site in the LPMCv (brown reaction product) and the FD injection site (blue reaction product) in the LPMCd in case SDM 57. C: Coronal section depicting the BDA injection site in the LPMCv (brown reaction product) in case

SDM 61. D: Coronal section depicting the FD injection site in M2 (blue reaction product) in case SDM 54. Black arrows identify a coalesced bundle of fibers arching inferiorly from the FD injection site. E: Horizontal section depicting the BDA injection site in M3 (brown reaction product) in case SDM 15. F: Coronal section depicting the BDA injection site in M4 (brown reaction product) in case SDM 43. A portion of this injection site also involved the dorsal region of area 23b of the cingulate gyrus. The arrow identifies a labeled fiber bundle emerging from the injection site. For abbreviations, see list. Scale bar ⫽ 1 mm in E (applies to A–F).

The Journal of Comparative Neurology. DOI 10.1002/cne

156

R.J. MORECRAFT ET AL. TABLE 1. Injection Parameters and Arm Representations Involved in Each Monkey1

Case no. SDM 15

SDM 19 SDM 23

SDM 39 SDM 43 SDM 54

SDM 57

SDM 61

1

Area injected

Tracer/injections

Total vol. (␮L)

Postinjection survival (days)

M1 arm LPMCv arm M2 arm M3 arm M4 arm M1 arm M2 arm M1 arm LPMCd arm M2 arm M3 arm M4 arm M3 arm/head M3 arm/head M4 arm/head/23b M1 arm M2 arm LPMCd arm M1 arm LPMCd arm LPMCv arm M1 arm LPMCd arm LPMCv arm

PHA-L/4 FD/4 LYD/2 BDA/2 FR/2 FD/4 FR/2 PHA-L/3 FD/3 FR/2 BDA/2 LYD/2 BDA/3 FD/3 BDA/3 LYD/3 FD/3 BDA/3 LYD/3 FD/3 BDA/3 LYD/3 FD/3 BDA/3

1.2 1.2 0.4 0.4 0.4 0.8 0.6 1.2 1.2 0.8 0.4 0.4 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2

30

26 29

medial subsector but spread to involve the lateral half of the medial subsector (Figs. 3C, 4C–E). In terms of density, the fibers were similarly distributed in these regions. Very few fibers were found to occupy the medial region of the medial subsector. At inferior midbrain levels, the main fiber bundle extended medially, thus involving the entire portion of the medial and centromedial subsectors (Figs. 3F, 4F,G). Fiber bundle density maintained a uniform distribution across both subsectors. Finally, through the midbrain-pontine isthmus, the M2 fiber bundle occupied the medial region of the peduncle and overlapped extensively with all of the fiber systems studied (Fig. 4H).

32 33

Trajectory of M3 arm fibers through the crus cerebri

33

At superior and mid-levels of the midbrain, the corticofugal projection from the arm area of M3 occupied the medial and centromedial subsectors of the crus cerebri of the cerebral peduncle (Figs. 3D, 4A–E). Upon descent, they were heavily distributed in the medial subsector and lightly dispersed in the medial region of the centromedial subsector. Thus, the M3 projection maintained the most medial position of all the fiber pathways studied at superior and middle levels of the crus cerebri. At inferior midbrain levels including the midbrain-pontine isthmus, M3 fibers continued to occupy this position and were extensively intermingled with fibers from all of the other motor areas (Figs. 3F, 4F–H).

32

33

For abbreviations, see list.

isthmus, the fibers continued to occupy the medial half of the peduncle and overlapped extensively with corticofugal projection fibers from all of the other motor areas studied (Figs. 4H, 5G).

Trajectory of M2 arm fibers through the crus cerebri

Trajectory of M4 arm fibers through the crus cerebri

The course of the M2 projection through the crus cerebri was studied in four cases (Figs. 1, 2D, 3C,F, Table 1). We found the M2 projection to be located in the same general position of the crus cerebri as the LPMCd and LPMCv projections. For example, at superior levels of the midbrain, the corticofugal projection from the arm area of M2 occupied primarily the centromedial region of the crus cerebri of the cerebral peduncle (Fig. 4A,B). Limited involvement was noted in the medialmost part of the central subsector and the lateralmost part of the medial subsector. As the fibers continued their descent through midlevels of the midbrain, they involved primarily the centro-

The descending course of the M4 projection in the ccCP was similar to the M1 course, as evidenced by the extensive overlap between the two pathways in case SDM 23 (Fig. 4). For example, at superior and mid-levels of the midbrain, the corticofugal projection from the arm area of M4 occupied the central subsector of the cerebral peduncle and the lateral segment of the centromedial subsector (Figs. 3D, 4A–E). Occasionally, M4 fibers extended more laterally in the central subsector than fibers originating from M1 and, in some sections, occupied the most medial region of the centrolateral subsector (Figs. 3D, 4E). As the M4 fibers continued their descent, they maintained a central position until reaching the inferior region of the pe-

Fig. 3. Plate of digital photomicrographs illustrating examples of immunohistochemically labeled fiber bundles in the crus cerebri. In each panel the asterisks denote the general boundary between the anatomical subdivisions, which are identified in C. In each panel the inset(s) show an enlarged image of labeled fibers from the location depicted by the arrow. A: Representative section through superior levels of the midbrain in SDM 57 depicting BDA-labeled fibers (brown reaction product) and LYD-labeled fibers (blue reaction product). The main distribution of BDA-labeled fibers is outlined by the red dashed line and that of LYD fibers by the blue dashed line. B: Representative section through superior levels of the midbrain in case SDM 57. Depicted are BDA-labeled fibers (brown reaction product) and FDlabeled fibers (blue reaction product). Fibers from both injections sites were located in the centromedial and medial sectors and overlapped extensively. The main distribution of BDA- and FD-labeled fibers is outlined by the black dashed line. C: Representative section through mid-levels of the midbrain in case SDM 54. Depicted are BDA-labeled fibers (brown reaction product) and FD-labeled fibers (blue reaction product). Fibers from both injection sites were located in the centromedial and medial sectors of the ccCP and overlapped extensively.

The main distribution of BDA- and FD-labeled fibers is outlined by the black dashed line. D: Representative section through mid-levels of the midbrain in SDM 43. Depicted are BDA-labeled fibers (brown reaction product) and FD-labeled fibers (blue reaction product). The main distribution of BDA-labeled fibers is outlined by the red dashed line and that of FD fibers by the blue dashed line. E: Representative section through inferior levels of the midbrain in case SDM 54. Depicted are BDA-labeled fibers (brown reaction product) and LYDlabeled fibers (blue reaction product). Fibers from both injection sites overlapped extensively. The main distribution of BDA- and LYDlabeled fibers is outlined by the black dashed line. F: Representative section through inferior levels of the midbrain in case SDM 23. Depicted are BDA-labeled fibers (brown reaction product) and FRlabeled fibers (blue reaction product). Fibers from both injection sites were located in the medial and centromedial sectors with overlap. The main distribution of BDA-labeled fibers is outlined by the red dashed line and that of the FR fibers by the blue dashed line. For abbreviations, see list. The anatomical orientation depicted in the bottom right of E applies to A–F. Scale bar ⫽ 2 mm in F (applies to A–F).

The Journal of Comparative Neurology. DOI 10.1002/cne

ARM REPRESENTATION IN THE CEREBRAL PEDUNCLE

.

Figure 3

157

The Journal of Comparative Neurology. DOI 10.1002/cne

Fig. 4. Representative line illustrations through the midbrain showing superior levels (A,B), mid-levels (C–E), inferior levels (F,G), and the midbrain-pontine isthmus (H) in case SDM 23. Illustrated are the corresponding trajectories of the five pathways studied through

the ccCP. Each descending fiber bundle is identified by a color-coded outline, and the cortical origin of each bundle is identified in the key on the bottom right of the figure. For abbreviations, see list.

The Journal of Comparative Neurology. DOI 10.1002/cne

ARM REPRESENTATION IN THE CEREBRAL PEDUNCLE

Figure 4

(Continued)

159

The Journal of Comparative Neurology. DOI 10.1002/cne

Fig. 5. Representative line illustrations through the midbrain showing superior levels (A,B), mid-levels (C,D), inferior levels (E,F), and the midbrain-pontine isthmus (G) in case SDM 57. Illustrated are the corresponding trajectories of the three pathways studied through

the ccCP. Each descending fiber bundle is identified by a color-coded outline, and the cortical origin of each bundle is identified in the key on the bottom right of the figure. For abbreviations, see list.

The Journal of Comparative Neurology. DOI 10.1002/cne

ARM REPRESENTATION IN THE CEREBRAL PEDUNCLE

Figure 5

(Continued)

161

The Journal of Comparative Neurology. DOI 10.1002/cne

162

R.J. MORECRAFT ET AL.

duncle. At this level the fiber bundle migrated medially to occupy the central and centromedial subregions and overlapped extensively with all other motor area projections (Fig. 4F,G). A light distribution of fibers was also found in the medial subsector (Fig. 4G). From a comparative perspective, the M4 projection in case SDM 43 was similar to that found in case SDM 23, with the exception of labeled fibers extending more laterally than found in case SDM 23. For example, corticofugal labeling also occurred in the lateral region of the centrolateral sector (Fig. 3D). In consideration of the similarities and differences in these cortical injection sites, the lateral extension of the projection in case SDM 43 may be attributable to injection site involvement of the medial wall or of the gyral portion of the cingulate cortex (i.e., area 23b).

Summary of results At superior and mid-levels of the midbrain, arm representation occupied the medial three-fifths of the crus cerebri (Fig. 6). M1 and M4 fibers were mainly located in the central region of the crus cerebri and M3 fibers medially. Fibers originating from the M2, LPMCd, and LPMCv were located between, with significant overlap. As the fiber pathways descended, they gradually became more diffuse and expanded their region of occupation in the medial dimension, with the exception of M3, which maintained a medial orientation throughout its entire course of passage. Upon reaching inferior levels of the midbrain, corticofugal projections from all six cortical arm representations occupied the medial half of the crus cerebri, and overlap was extensive (Fig. 6). This commixed organization continued as the fibers passed through the midbrain-pontine isthmus to enter the superior pontine tegmentum (Fig. 4H, 5G). The condensed and overlapping nature of fiber organization through the midbrain-pontine isthmus suggests a vulnerable region for comprehensive disruption of corticofugal projection fibers from the arm areas of the frontal and cingulate motor cortices.

DISCUSSION Midbrain injury typically occurs as a consequence of posterior circulation ischemia and has long been recognized from the general perspective of distinctive categorical symptoms. Occulomotor disorders, ataxia, and hemiparesis typically accompany anteriomedial and anterolateral vascular lesions, whereas sensory dysfunction accompanies lateral lesions (Caplan, 1980; Tuttle and Reinmuth, 1984; Alverez-Sabin et al., 1991; Bogousslavsky et al., 1994; Hommel and Besson, 2001; Kim and Kim, 2005). The symptom of motor paresis has historically been associated with lesions affecting the crus cerebri of the cerebral peduncle (ccCP), and it has been customary to consider the destruction of the corticofugal projection from the precentral gyrus of the frontal lobe as the causative factor. However, published accounts of the location of the projection from M1 are inconsistent (Meyers, 1951; Walker, 1955; Bucy and Keplinger, 1961; Bucy et al., 1964; Brodal, 1981; Quinones-Hinojosa et al., 2005), and the clinical profile following cerebral peduncle insult suggests a more complicated organization than is supported by the available data. Further underscoring this enigma are the highly variable outcomes that follow midbrain injury, ranging from severe, persistent hemiparesis to ex-

traordinary levels of motor recovery characterized by functional independence. These clinical observations have led to the suspicion that corticofugal motor representation, or what has traditionally been described as “pyramidal tract” representation, may be more widespread in the ccCP than classically interpreted (Bucy et al., 1964; Kim and Kim, 2005). Our results offer strong support for this supposition (Fig. 6) and provide new information that is of practical value for interpreting the clinical consequences of isolated midbrain damage and predicting motor recovery outcomes following localized upper brainstem injury. Our findings also lend insight to predicting which regions of the motor cortex may contain intact corticofugal circuitry and therefore be identified as candidate targets for therapeutic interventions designed to enhance the motor recovery process following the compromise of distinct midbrain subsectors.

Localization of arm representation in the crus cerebri of the cerebral peduncle Lateral cortical pathways. Our observations are largely in accord with those of Levin (1936), Barnard and Woolsey (1956), and Schmahmann and colleagues (2004), who found fibers from the arm representation of M1 located in the central region of the crus cerebri at superior levels in the monkey (Figs. 4A–E, 5A–D, 6). In terms of human observations, our findings are similar to the central location of fibers arising from the precentral cortical region as determined by dissection (Ross, 1980) and the marked presence of localized Wallerian degeneration associated with small lesions seated in the lateral precentral gyrus (Waragai et al., 1994). Our observations also complement human intraoperative stimulation mapping observations, which show that direct surface stimulation of the central region of the crus cerebri gives rise to arm movement (Walker, 1955; Bucy, 1957; Quinones-Hinojosa et al., 2005). Our analysis extends these findings by demonstrating a more expansive representation of projection fibers at middle and inferior levels of the ccCP including a gradual shift of M1 fibers in the descending dimension to the central and centromedial region at middle levels and then to the central, centromedial, and medial regions at inferior levels (Figs. 4 – 6). Furthermore, at the midbrainpontine isthmus, we found that M1 fibers overlapped extensively with fibers from all of the other cortical arm areas studied (Figs. 4H, 5G, 6). We found that the projections from the LPMCd and LPMCv followed the identical descending course in cases SDM 57 (Fig. 5) and SDM 61, in which both motor area pathways were studied in the same experimental brain (Fig. 1). A consideration of cases SDM 23 and SDM 54 (Fig. 1), which demonstrated that the LPMCd projection and the M2 projection overlapped (Figs. 3C, 4), suggests that the pathways from the LPMCd (area 6D), LPMCv (area 6V), and M2 (area 6m) follow the same descending course through the crus cerebri. We also noted that across our cases, all area 6 projections were located medially to the M1 projection, with some overlap (Figs. 4 – 6). These findings complement the experimental work of previous investigators who noted degenerating fibers in the centromedial peduncular region following ablation of the periarcuate portion of the LPMCd and in the central subsector following ablation of M1(Levin, 1936; Bernard and Woolsey, 1956; Kuypers, 1981). However, in addition to these observations, we found labeled LPMCd fibers located me-

The Journal of Comparative Neurology. DOI 10.1002/cne

ARM REPRESENTATION IN THE CEREBRAL PEDUNCLE

163

Fig. 6. Summary diagram illustrating the main findings of our study. At superior levels, fibers from M1 and M4 occupied the central subsector of the ccCP, and fibers from M3 resided medially (top right). The main distribution of LPMCd, LPMCv, and M2 fibers occupied the centromedial region and overlapped extensively. Progressing inferiorly, fiber bundles in the central and centromedial sectors gradually extended medially, and the overlap increased. At inferior midbrain levels, all the fiber bundles overlapped (bottom left). The widespread distribution of corticofugal motor projections in the ccCP may account

for the favorable levels of motor recovery that accompany subtotal midbrain injury. The topography of anteromedial (AM) and anterolateral (AL) vascular territories, which may be compromised in midbrain occlusions (depicted on the left of each horizontal section), is based on the interpretations of Duvernoy (1999), Hommel and Besson (2001), and Kim and Kim (2005). On the bottom right is a threedimensional model of the descending pathways from superior to inferior levels of the ccCP. For abbreviations, see list.

dial to this position at middle and inferior midbrain levels (Figs. 3C–E, 4C–E). Our observations are also the first to localize corticofugal representation from the arm region of the LPMCv, which, as noted, overlapped significantly with labeled fibers from the LPMCd (Figs. 3B, 5, 6). In addition, we found that the lateral premotor cortex projection migrated medially and overlapped extensively with all of the

motor area projections studied at inferior midbrain levels and the midbrain-pontine isthmus (Figs. 3E, 5F,G). Medial cortical pathways. Our investigation of the corticofugal projection from the arm region of three distinct motor areas on the medial wall of the hemisphere revealed that these pathways, as a group, occupied an extensive portion of the crus cerebri. In total, this projec-

The Journal of Comparative Neurology. DOI 10.1002/cne

164

R.J. MORECRAFT ET AL.

tion system, whose origin is supplied by the anterior cerebral artery, which is rarely involved in cortical strokes, was distributed across the medial three-fifths of the crus cerebri (Figs. 4, 6). The most lateral projection arose from the caudal cingulate motor cortex (M4), and the most medial projection arose from the rostral cingulate motor cortex (M3), with the supplementary motor cortex (M2) projection located in between. The positioning of the projection from M3 was characterized by its consistent occupation of the medial and centromedial subsectors throughout its descending course. Classically, the medial region of the peduncle was presumed to contain only tegmental and frontopontine fibers (Beck, 1950; Bucy et al., 1966); however, our findings demonstrate the inclusion of corticospinal fibers because terminals from the M3 projection were found in the spinal cord in cases SDM 23, SDM 39, and SDM 43. Fibers from M2, as well as those from the LPMCd and LPMCv, were located primarily lateral to M3 fibers, with some overlap (Figs. 3F, 4 – 6). In contrast, the projection from M4 was positioned mainly in the central subsector, following the general course of the M1 projection (Figs. 3D, 4, 6). Notably, the close anatomical relationship between descending fiber bundles from M4 and M1 also occurs in the internal capsule (Morecraft et al., 2002). Lastly, all medial wall projections overlapped extensively with all the lateral motor area projections at inferior levels and the midbrain-pontine isthmus, suggesting a common and potentially vulnerable region of fiber passage.

Clinical considerations Although there have been few systematic studies focusing on the clinical consequences and syndromes associated with isolated midbrain injury (Bucy et al., 1966; Walker and Richter, 1966; Bogousslavsky et al., 1994; Kim and Kim, 2005), there are an abundant number of clinical reports demonstrating that hemiparesis is a common functional consequence following disruption of the midbrain cerebral peduncle when the confirmed lesion includes the crus cerebri. For instance, hemorrhagic lesions occupying the medial region of the crus cerebri (Dabby et al., 2004; Kim and Kim, 2005), medial and central regions (Dehaene and Dom, 1982; Duncan and Weindling, 1995), central region (Ho, 1982), and central and lateral regions (Hommel et al., 1990; Chia, 1991; Gaymard et al., 1991; Park et al., 1997; Messe et al., 2001; Dabby et al., 2004; Kim and Kim, 2005) lead to acute hemiparesis. However, what is intriguing in this body of literature are the numerous accounts of patients attaining remarkable levels of motor recovery over time. Similarly, only transient impairment of motor function was observed in many patients following surgical interruption of various subregions of the human crus cerebri that was once performed for the treatment of intractable movement disorders (Walker, 1949, 1955; Meyers, 1951, 1956a,b; Hamby, 1953; Broager, 1955; Bucy and Keplinger, 1961; Bucy et al., 1964; Maspes and Pagni, 1964). The significant level of motor recovery observed following human pedunculotomy was subsequently observed in a classic series of non-human primate experiments examining the effects of peduncular transection on motor behavior (Bucy et al., 1966; Walker and Richter, 1966). When these clinical observations are taken into account, our findings suggest that the widespread distribution of arm-related corticofugal fibers in the ccCP may provide a

favorable situation for pervasive fiber sparing. This, in turn, may play a major role in contributing to the extraordinary degree of upper extremity recovery that accompanies subtotal peduncular injury. Of particular interest to our findings is a recent systematic study that evaluated the clinical features of 40 patients with pure midbrain infarction (Kim and Kim, 2005). In this comprehensive effort, the authors found limb weakness to be present in more than half (55%) of the entire patient population, with 23% of the patients presenting with definitive hemiparesis. In terms of motor deficits, three patient groups were identified based on lesion location. They included an anteromedial lesion group; an anterolateral lesion group; and a combined (anteriomedial ⫹ anterolateral) lesion group (see Fig. 1 of Kim and Kim, 2005). Although we cannot assess with certainty the degree of peduncular involvement in these cases, it is possible to draw some intriguing parallels between our findings and the prevalence of hemiparesis as well as the severity of motor deficit associated with each patient subgroup. Interestingly, there was a distinctive trend indicating that the percentage of patients presenting with definitive hemiparesis increased from the anteromedial group, to the anterolateral group, to the combined lesion group, as did the level of motor severity. For example, of the 40 patients, 18 had anteriomedial infarctions; only 4 (22%) of these presented with mild levels of limb weakness and none with definitive hemiparesis. Depending on the extent of peduncular involvement, this lesion has the potential to disrupt motor fibers located medially in the ccCP while sparing those located more centrally (Fig. 6). Notably, the potential lack of involvement of M1 fibers, which gives rise to the densest corticospinal projection (Murray and Coulter, 1981; Nudo and Masterson, 1990; Dum and Strick, 1991, 1996; Galea and Darian Smith, 1994), as well as others in the vicinity of the central region, would correlate well with the absence of definitive hemiparesis reported in the anteromedial patient group. In contrast, 11 of the 40 patients were diagnosed with anterolateral lesions; 8 (73%) of the 11 patients presented with limb weakness. Of these, five (45%) were diagnosed with mild motor weakness and three (27%) with severe, definitive motor paresis. According to our findings, this lesion has the potential to compromise the centromedial and central subsectors of the ccCP, which contained a significant number of corticofugal motor fibers including projections from the LPMCd, LPMCv, M2, M4, and M1. Such a scenario would correlate with the increased prevalence, as well as severity, of motor deficits found in the anterolateral lesion group versus the anteromedial group. Finally, 6 of the 40 patients had a combined anteromedial and anterolateral injury, and all 6 (100%) presented with limb weakness. Of the six patients, two (33%) were diagnosed with mild limb weakness and four (67%) with severe limb weakness. According to our results, the combined lesion has the potential to heavily disrupt corticofugal fibers from all six cortical arm areas, which would theoretically yield significant deficits in motor function such as those characterized by this patient group. An additional significant finding from this study included the follow-up observations, which revealed that 36 (95%) of the surviving 38 patients progressed to a “functionally independent” status, suggesting remarkable levels of motor recovery. Indeed, recovery may be supported

The Journal of Comparative Neurology. DOI 10.1002/cne

ARM REPRESENTATION IN THE CEREBRAL PEDUNCLE by spared fibers located lateral to anteromedial lesions, medial to anterolateral lesions, or within the territory of a combined lesion. In consideration of the correlates drawn between our findings and the neurological symptomology accompanying anteromedial and anterolateral midbrain injury, it becomes increasingly clear that the widespread distribution of corticofugal fibers found in our non-human primate experiments may also exist in the human crus cerebri. Although such a distribution is speculative at the present time, further developments in ancillary methodology, including white matter tractography, hold great promise for the evaluation of these potential anatomic relationships in the human brain. To assist in unraveling this organization, our findings provide a formidable template to initiate the vigilant study of motor deficits accompanying small, well-defined peduncular lesions immediately following midbrain injury, as well as throughout the recovery process.

Technical considerations It should be noted that corticofugal organization in the ccCP was evaluated in our study by using experimental cases with injections of varying volumes and varying tracer types (Table 1). However, we consistently found the topography of labeled fibers associated with a single motor area to be similar across all experimental cases. For example, with respect to differences in both volume and tracer type, we injected 0.4 ␮l of LYD into M2 in case SDM 15, 0.8 ␮l of FR into M2 in case SDM 23, and 1.2 ␮l of FD into M2 in case SDM 54 (Table 1). Despite these differences, we consistently found the main distribution of labeled fibers in the centromedial subsector at superior and middle midbrain levels, with gradual spread in the medial dimension at inferior midbrain levels. Similarly, cases SDM 15 and SDM 23 received 0.4 ␮l of BDA each, and SDM 39 received 1.2 ␮l of BDA; all three M3 cases gave rise to labeling in the medial subsector and medial region of the centromedial subsector. Thus, differences in tracer volume did not change the results in topography. In our M4 cases, we injected the same volume but different tracers into in cases SDM 15 and SDM 23 and found that both tracers labeled fibers in the central anatomical subsector of the ccCP at superior and middle midbrain levels with gradual spread in the medial dimension inferiorly. The only difference noted in the M4 cases was with SDM 43, which gave rise to labeling more laterally in the ccCP than found in cases SDM 15 and SDM 23. We attributed this slight difference in case SDM 43 to injection involvement of area 23b, which lies on the vertical surface of the cingulate gyrus. We also found that the corticofugal fiber distribution yielded the same topographic results after injection of identical volumes of two different tracers in our LPMCd cases (SDM 54, SDM 57, and SDM 61) and M3 cases (SDM 39 and SDM 43). Thus, although it is likely that different tracers and injected volumes may produce varying results in the total number of individually labeled corticofugal axons, as well as the topographical distribution and density of labeled subcortical terminal-like particles, the general topography of each corticofugal projection as it passed through the ccCP was the same in all our experiments conducted on the same motor area. It is worth raising the point that most of our experimental animals survived for 29 –33 days after tracer injection to minimize the effect that survival time would have on these experiments (Table 1). Furthermore,

165

as pointed out previously, our findings are in close agreement with published accounts of the corticofugal projection in the ccCP from M1 and the LPMCd obtained following the application of ablation and autoradiography methodologies (Levin, 1936; Bernard and Woolsey, 1956; Kuypers, 1981; Schmahmann et al., 2004). This reoccurring anatomical characteristic may be attributable to the coalesced nature of fiber bundling through this ventrally isolated brainstem passageway.

CONCLUSIONS In this report, we provide a detailed description of frontal and cingulate arm representation in the crus cerebri of the non-human primate (Fig. 6). Our observations challenge the classic view of somatotopical organization in this brainstem region by suggesting that arm representation resides medially in the ccCP, as well as centrally, as traditionally interpreted. Comparatively speaking, our findings demonstrate that the various descending pathways studied occupy widespread parts of the crus cerebri and overlap extensively. Thus, both of these structural features may underlie favorable levels of motor recovery that accompany subtotal midbrain injury. An enhancement of our understanding of the organization of arm representation in this critical brain region is of significant value for improving clinical diagnosis and assisting in the interpretation of upper extremity deficits following localized posterior vascular occlusion or related upper brainstem injury. Our findings also form a basis for predicting clinical outcomes, designing future studies aimed to enhance our understanding of motor deficits associated with well-defined peduncular lesions, and forming a foundation to study methodically the process of long-term recovery of residual motor disability following localized midbrain injury.

LITERATURE CITED Alvarez-Sabin J, Montalban J, Tintore M, Codina A. 1991. Pure sensory stroke due to midbrain haemorrage. J Neurol Neurosurg Neuropsychiatry 54:843. Barbas H, Pandya DN. 1987. Architecture and frontal cortical connections of the premotor cortex (area 6) in the rhesus monkey. J Comp Neurol 256:211–228. Barnard JW, Woolsey CN. 1956. A study of localization in the corticospinal tracts of the monkey and rat. J Comp Neurol 105:25–50. Barnes S. 1901. Degenerations in hemiplegia: with special reference to a ventrolateral pyramidal tract, the accessory fillet and Pick’s bundle. Brain 24:463–501. Bauer G, Gerstenbrand F, Rumpl E. 1979. Variants of the locked-in syndrome. J Neurol 221:77–91. Beck E. 1950. The origin, course and termination of the frontopontine tract in the human brain. Brain 73:368 –391. Biber MP, Kneisley LW, LaVail JH. 1978. Cortical neurons projecting to the cervical and lumbar enlargements of the spinal cord in young adult rhesus monkeys. Exp Neurol 59:492–508. Binkofski F, Seitz RJ, Arnold S, Classen J, Benecke R, Freund HJ. 1996. Thalamic metabolism and corticospinal tract integrity determine motor recovery in stroke. Ann Neurol 39:460 – 470. Binkofski F, Seitz RJ, Hackla¨nder T, Pawelec D, Mau J, Freund HJ. 2001. Recovery of motor functions following hemiparetic stroke: a clinical and magnetic resonance-morphometric study. Cerebrovasc Dis 11:273–281. Bogousslavsky J, Maeder P, Regli F, Meuli R. 1994. Pure midbrain infarction: clinical syndromes, MRI, and etiologic patterns. Neurology 44: 2032–2040. Broager B. 1955. Cerebral pedunculotomy in four cases of hyperkinesia.

The Journal of Comparative Neurology. DOI 10.1002/cne

166 Three cases of choreoathetosis and one of rhythmical tremor. Acta Psychiatr Neurol Scand 30:107–114. Brodal A. 1981. Neurological anatomy in relation to clinical medicine, 3rd ed. Oxford: Oxford University Press. p 185–186. Bucy PC. 1957. Is there a pyramidal tract? Brain 80:376 –392. Bucy PC, Keplinger JE. 1961. Section of the cerebral peduncles: preliminary report. Arch Neurol 5:132–139. Bucy PC, Keplinger JE, Siqueira EB. 1964. Destruction of the pyramidal tract in man. J Neurosurg 21:385–398. Bucy PC, Ladpli R, Ehrlich A. 1966. Destruction of the pyramidal tract in monkey. The effects of bilateral section of the cerebral peduncles. J Neurosurg 25:1–20. Caplan LR. 1980. Top of the basilar syndrome. Neurology 30:72–79. Catman-Berrevoets CE, Kuypers HGJM. 1976. Cells of origin of cortical projections to the dorsal column nuclei, spinal cord and bulbar reticular formation. Neurosci Lett 3:245–252. Chia L-G. 1991. Locked-in syndrome with bilateral ventral midbrain infarcts. Neurology 41:445– 446. Dabby R, Watemberg N, Lampl Y, Eilam A, Rapaport A, Sadeh M. 2004. Pathological laughter as a symptom of midbrain infarction. Behav Neurol 15:73–76. Dehaene I, Dom R. 1982. A mesencephalic locked-in syndrome. J Neurol 227:255–259. Dum RP, Strick PL. 1991. The origin of corticospinal projections from the premotor areas in the frontal lobe. J Neurosci 11:667– 689. Dum RP, Strick PL. 1996. Spinal cord terminations of the medial wall motor areas in macaque monkeys. J Neurosci 16:6513– 6525. Duncan GW, Weindling SM. 1995. Posterior cerebral artery stenosis with midbrain infarction. Stroke 26:900 –902. Duvernoy HM. 1999. Human brain stem vessels, 2nd ed. Springer: New York. Fridman EA, Hanakawa T, Chung M, Hummel F, Leiguarda R, Cohen LG. 2004. Reorganization of the human ipsilesional premotor cortex after stroke. Brain 127:747–758. Fries W, Danek A, Scheidtmann K, Hamburger C. 1993. Motor recovery following capsular stroke: role of descending pathways from multiple motor areas. Brain 116:369 –382. Galea MP, Darian-Smith I. 1994. Multiple corticospinal neuron populations in the macaque monkey are specified by their unique cortical origins, spinal terminations, and connections. Cereb Cortex 4:166 –194. Gaymard B, Saudeau D, de Toffol B, Larmande P, Autret A. 1991. Two mesencephalic lacunar infarcts presenting as Claude’s syndrome and pure motor hemiparesis. Eur Neurol 31:152–155. Hamby W. 1953. The surgical treatment of dystonia musculorum deformans. J Neurosurg 10:490 – 495. He SQ, Dum RP, Strick PL. 1995. Topographic organization of corticospinal projections from the frontal lobe; motor areas on the medial surface of the hemisphere. J Neurosci 15:3284 –3306. Ho K-L. 1982. Pure motor hemiplegia due to infarction of the cerebral peduncle. Arch Neurol 39:524 –526. Hommel M, Besson G. Midbrain infarcts. 2001. In Bogousslavsky J, Caplan LR, editors. Stroke syndromes, 2nd ed. Cambridge, UK: Cambridge University Press. p 512–519. Hommel M, Besson G, Pollak P, Kahane P, Le Bas JF, Perret J. 1990. Hemiplegia in posterior cerebral artery occlusion. Neurology 40:1486 – 1499. Karp JS, Hurtig HI. 1974. “Locked-in” state with bilateral midbrain infarcts. Arch Neurol 30:176 –178. Kim JS, Kim J. 2005. Pure midbrain infarction: clinical, radiologic, and pathologic findings. Neurology 64:1227–1232. Kim JS, Lee JH, Im JH, Lee MC. 1995. Syndromes of pontine base infarction: a clinical-radiological correlation study. Stroke 26:950 –955. Kuypers HGJM. 1981. Anatomy of the descending motor pathways. In Brooks V, editor. Handbook of physiology, sec I, The nervous system, vol II, Motor control, Pt I. Bethesda: American Physiological Society. p 567– 666. Lang CE, Schieber MH. 2003. Differential impairment of individuated finger movements in humans after damage to the motor cortex or the corticospinal tract. J Neurophysiol 90:1160 –1170. Levin PM. 1936. The efferent fibers of the frontal lobe of the monkey, Macaca mulatta. J Comp Neurol 63:369 – 419. Maspes PE, Pagni CA. 1964. Surgical treatment of dystonia and choreoathetosis in infantile cerebral palsy by pedunculotomy. Pathophysio-

R.J. MORECRAFT ET AL. logical observations and therapeutic results. J Neurosurg 21:1076 – 1086. Meienberg O, Mumenthaler M, Karbowski K. 1979. Quadriparesis and nuclear oculomotor palsy with total bilateral ptosis mimicking coma. A mesencephalic locked in syndrome? Arch Neurol 36:708 –710. Mettler FA. 1942. Neuroanatomy. St. Louis: CV Mosby. Messe SR, Shin RK, Liu GT, Galetta SL, Volpe NJ. 2001. Oculomotor synkinesis following a midbrain stroke. Neurology 57:1106 –1107. Meyers R. 1951. The human frontocorticopontine tract: functional inconsequence of its surgical interruption. Neurology 1:341–356. Meyers R. 1956a. Results of bilateral intermediate midbrain crusotomy in seven cases of severe athetotic and dystonic quadriparesis. Am J Phys Med 35:84 –105. Meyers R. 1956b. Physiological and therapeutic effects of bilateral intermediate midbrain crusotomy for atheto-dystonia (17 cases). Surg Forum 6:486 – 488. Morecraft RJ, Tanji J. 2007. Cingulofrontal interactions and cingulate skeletomotor areas. In Vogt BA, editor. Cingulate neurobiology and disease: infrastructure, diagnosis, treatment. Oxford: Oxford University Press, in press. Morecraft RJ, Van Hoesen GW. 1992. Cingulate input to the primary and supplementary motor cortices in the rhesus monkey: evidence for somatotopy in areas 24c and 23c. J Comp Neurol 322:471– 489. Morecraft RJ, Louie JL, Schroeder CM, Avramov K. 1997. Segregated parallel inputs to the brachial spinal cord from the cingulate motor cortex in the monkey. NeuroReport 8:3933–3938. Morecraft RJ, Louie JL, Herrick JL, Stilwell-Morecraft. 2001. Cortical innervation of the facial nucleus in the non-human primate: a new interpretation of the effects of stroke and related subtotal brain trauma on the muscles of facial expression. Brain 124:176 –208. Morecraft RJ, Herrick JL, Stilwell-Morecraft KS, Louie JL, Schroeder CM, Ottenbacher JG, Schoolfield MW. 2002. Localization of arm representation in the corona radiata and internal capsule in the non-human primate. Brain 125:176 –198. Morecraft RJ, Cipolloni PB, Stilwell-Morecraft KS, Gedney MT, Pandya DN. 2004. Cytoarchitecture and cortical connections of the posterior cingulate and adjacent somatosensory fields in the rhesus monkey. J Comp Neurol 469:37– 69. Morecraft RJ, McNeal DW, Stilwell-Morecraft KS, Gedney M, Ge J, Schroeder CM, Van Hoesen GW. 2007. Amygdala interconnections with the cingulate motor cortex in the rhesus monkey. J Comp Neurol 500:134 – 165. Muakkassa KF, Strick PL. 1979. Frontal lobe inputs to primate motor cortex: evidence for four somatotopically organized “premotor” areas. Brain Res 177:176 –182. Murray EA, Coulter JD. 1981. Organization of corticospinal neurons in the monkey. J Comp Neurol 195:339 –365. Newton JM, Ward NS, Parker GJ, Deichmann R, Alexander DC, Friston KJ, Frackowiak RS. 2006. Non-invasive mapping of corticofugal fibres from multiple motor areas—relevance to stroke recovery. Brain 129: 1844 –1858. Nudo RJ, Masterson RB. 1990. Descending pathways to the spinal cord, III. Sites of origin of the corticospinal tract. J Comp Neurol 296:559 – 583. Park SA, Sohn YH, Kim WC. 1997. Locked-in syndrome with bilateral peduncular infarct. J Neuroimag 7:126 –128. Pendl G, Vorkapic P. 1988. Microsurgery of midbrain lesions. Acta Neurochir Suppl 42:130 –136. Pineiro R, Pendlebury ST, Smith S, Flitney D, Blamire AM, Styles P, Matthews PM. 2000. Relating MRI changes to motor deficit after ischemic stroke by segmentation of functional motor pathways. Stroke 31:672– 679. Quinones-Hinojosa A, Lyon R, Du R, Lawton MT. 2005. Intraoperative motor mapping of the cerebral peduncle during resection of a midbrain cavernous sinus malformation: technical case report. Neurosurgery 54:ONS439. Raghavan P, Krakauer JW, Gordon AM. 2006. Impaired anticipatory control of fingertip forces in patients with pure motor or sensorimotor lacunar syndrome. Brain 129:1415–1425. Ross ED. 1980. Localization of the pyramidal tract in the internal capsule by whole brain dissection. Neurology 30:59 – 64. Schmahmann JD, Pandya DN. 2006. Fiber pathways of the brain. New York: Oxford University Press.

The Journal of Comparative Neurology. DOI 10.1002/cne

ARM REPRESENTATION IN THE CEREBRAL PEDUNCLE Schmahmann JD, Rosene DL, Pandya DN. 2004. Motor projections to the basis pontis in rhesus monkey. 478:248 –268. Shelton FNAP, Reding MJ. 2001. Effect of lesion location on upper limb motor recovery after stroke. Stroke 32:107–112. Standring S. 2005. Gray’s anatomy: the anatomical basis of clinical practice, 39th ed. London: Elsevier Churchill Livingstone. p 342–347. Tuttle PV, Reinmuth OM. 1984. Midbrain hemorrhage producing pure sensory stroke. Arch Neurol 41:794 –795. Walker AE. 1949. Cerebral pedunculotomy for the relief of involuntary movements; hemiballismus. Acta Psychiatr Neurol 24:723–729. Walker AE. 1952. Cerebral pedunculotomy for the relief of involuntary movements II. parkinsonian tremor. J Nerv Ment Dis 116:766 –775. Walker AE. 1955. Cerebral pedunculotomy for involuntary movements. Surg Gynecol Obstet 100:716 –720. Walker AE, Richter H. 1966. Section of the cerebral peduncle in the monkey. Arch Neurol 14:231–240. Waragai M, Watanabe H, Iwabuchi S. 1994. The somatotopic localisation of

167

the descending cortical tract in the cerebral peduncle: a study using MRI of changes following Wallerian degeneration in the cerebral peduncle after a supratentorial vascular lesion. Neuroradiology 36:402– 404. Ward NS, Newton JM, Swayne OBC, Lee L, Thompson AJ, Greenwood RJ, Rothwell JC, Frackowiak RSJ. 2006. Motor system activation after subcortical stroke depends on corticospinal system integrity. Brain 29:809 – 819. Weiller C, Ramsay SC, Wise RJS, Friston KJ, Frackowiak RSJ. 1993. Individual patterns of functional reorganization in the human cerebral cortex after capsular infarction. Ann Neurol 33:181–189. Wenzelburger R, Kopper F, Frenzel A, Stolze H, Klebe S, Brossmann A, Kuhtz-Buschbeck J, Go¨lge M, Illert M, Deusch G. 2005. Hand coordination following capsular stroke. Brain 128:64 –74. Woolsey CN, Settlage PH, Meyer DR, Sencer W, Pinto Hamuy T, Travis AM. 1952. Patterns of localization in precentral and “supplementary” motor areas and their relation to the concept of a premotor area. Res Publ Assoc Res Nerv Ment Dis 30:238 –264.