Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2011 Dec 26.
Published in final edited form as: J Cereb Blood Flow Metab. 2006 Apr 26;27(1):76–85. doi: 10.1038/sj.jcbfm.9600320

VEGF protein associates to neurons in remote regions following cortical infarct

Ann M Stowe 1, Erik J Plautz 1, Ines Eisner-Janowicz 1, Shawn B Frost 1, Scott Barbay 1, Elena V Zoubina 1, Numa Dancause 1, Michael D Taylor 1, Randolph J Nudo 1
PMCID: PMC3245973  NIHMSID: NIHMS340735  PMID: 16639424

Abstract

Vascular endothelial growth factor (VEGF) is thought to contribute to both neuroprotection and angiogenesis after stroke. While increased expression of VEGF has been demonstrated in animal models after experimental ischemia, these studies have focused almost exclusively on the infarct and peri-infarct regions. The present study investigated the association of VEGF to neurons in remote cortical areas at three days after an infarct in primary motor cortex (M1). Although these remote areas are outside of the direct influence of the ischemic injury, remote plasticity has been implicated in recovery of function. For this study, intracortical microstimulation techniques identified primary and premotor cortical areas in a non-human primate. A focal ischemic infarct was induced in the M1 hand representation, and neurons and VEGF protein were identified using immunohistochemical procedures. Stereological techniques quantitatively assessed neuronal-VEGF association in the infarct and peri-infarct regions, M1 hindlimb, M1 orofacial, and ventral premotor hand representations, as well as non-motor control regions. The results indicate that VEGF protein significantly increased association to neurons in specific remote cortical areas outside of the infarct and peri-infarct regions. The increased association of VEGF to neurons was restricted to cortical areas that are functionally and/or behaviorally related to the area of infarct. There was no significant increase in M1 orofacial region or in non-motor control regions. We hypothesize that enhancement of neuronal VEGF in these functionally related remote cortical areas may be involved in recovery of function after stroke, through either neuroprotection or the induction of remote angiogenesis.

Keywords: VEGF (vascular endothelial growth factor), neuron, stroke, focal cerebral ischemia, stereology, neuroprotection

Introduction

Although stroke is the third leading cause of death and the leading cause of long-term disability among adults in the United States, the understanding of post-stroke physiology is far from comprehensive. One focus of study gaining increased attention is the process and significance of angiogenesis within, and surrounding, the ischemic infarct. Angiogenesis is both the modification of existing vessel diameters and branching patterns and the formation of new blood vessels presumably to increase blood flow to tissue. Experimental data in rodent vascular occlusion models have demonstrated that angiogenesis can occur in the peri-infarct region (Zhang et al, 2000; Wei et al, 2001). It has been suggested that peri-infarct angiogenesis in human stroke survivors is related to long-term recovery (Krupinski et al, 1994).

One of many neurotrophic and angiogenic factors implicated in the process of stroke recovery is vascular endothelial growth factor (VEGF), a protein that is fundamental during adult angiogenesis. The initiation of angiogenesis through VEGF relies on the immediate reduction in blood flow to periinfarct and infarct tissue, resulting in a loss of oxygen delivery to cells (i.e. hypoxia; Marti et al, 2000; Wei et al, 2001). This fall in tissue oxygen pressure prevents the degradation of hypoxia inducible factor 1-alpha subunit, allowing it to dimerize to its beta subunit and promote the transcription of certain target genes, including VEGF (Bergeron et al, 1999). Vascular endothelial growth factor protein interacts with receptors on ischemic endothelial cells in the immediate hours after stroke to produce blood–brain barrier (BBB) leakage, increased vascular permeability, and edema (Zhang et al, 2000). Intravenous (iv) administration of recombinant human (rh) VEGF, at one hour after a middle cerebral artery occlusion (MCAo) in rats, significantly increases blood–brain barrier permeability, edema, and ischemic lesion size. At 2 days after infarct, however, the administration of recombinant human VEGF results in recovery of brain tissue by enhancing angiogenesis, and concomitantly reducing motor and somatosensory deficits during functional recovery. Thus, the role of postinfarct VEGF appears to transform from a permeability factor to an angiogenic factor, which ultimately spares further loss of cortical tissue.

Although VEGF is an ‘endothelial’ factor, its production and distribution under conditions of adult angiogenesis are not limited to endothelial cells. Post-stroke experimental and clinical data have suggested that astrocytes, macrophages/microglial cells, and/or neurons upregulate peri-infarct VEGF protein production (Hayashi et al, 1997; Marti and Risau 1998; Issa et al, 1999; Plate et al, 1999). In an effort to elucidate the functions of VEGF in stroke recovery, new data has stressed the role of VEGF in neuroprotection after ischemic injury (Sun et al, 2003; Wang et al, 2005). In vitro studies have suggested that neurons can produce then use VEGF in an autocrine function to enhance axonal outgrowth and cell survival (Sondell et al, 1999; Ogunshola et al, 2002). Many in vivo infarct studies in rodents have also demonstrated a post-stroke neuroprotective effect of VEGF, often in the absence of the induction of angiogenesis (Yang et al, 2002; Sun et al, 2003). Several investigators have suggested that VEGF may act in multiple functions after stroke, and that processes such as angiogenesis and neuroprotection are not necessarily mutual (Sun et al, 2003).

The investigation into neuronal expression of VEGF in the cerebral cortex has been limited thus far to the peri-infarct regions. While the peri-infarct region is vulnerable to neuronal death after stroke, it may not be the only cortical area that undergoes changes in the pattern of VEGF expression in response to the infarct. A large body of work has demonstrated that biochemical, physiologic, and anatomic modifications occur in cortical areas removed from the site of hypoxic/ischemic injury, and that these changes can aid in functional recovery (Seitz et al, 1998; Liu and Rouiller 1999; Frost et al, 2003; Dancause et al, 2005). These ‘remote’ changes after stroke may be initiated through functional associations with the area of infarct or modifications in behavior.

The present study investigating VEGF protein association to neurons after stroke is novel in the following aspects: First, it extends the examination of VEGF protein expression to a non-human primate model of focal cortical ischemia. With few exceptions (Issa et al, 1999; Nadar et al, 2005), such studies have been limited to rodent models. Second, virtually all of the previous studies of VEGF protein expression have been limited to the infarct and periinfarct tissue. The present study examines VEGF protein association to neurons in physiologically identified cortical areas remote from the site of the infarct. The homology of cortical motor areas between monkeys and humans facilitates this novel investigation in remote, but functionally interrelated, motor areas (Maier et al, 1997; Rizzolatti et al, 2002). Finally, this study represents a novel use of design based stereological techniques in neurophysiologically defined cortical areas to assess sensitive shifts in neuronal-VEGF ratios. The experimental results in non-human primates demonstrate that VEGF increases association to neurons in remote cortical areas that are functionally related to the area of infarct. Although it is not yet possible to determine the precise role of increased VEGF association to neurons remote from the site of injury (e.g., neuroprotection, angiogenesis), and VEGF is only one of many neurotrophic and angiogenic factors upregulated after an infarct, the present data extend our understanding of complex early molecular events during the sub-acute stages of stroke.

Materials and methods

Surgical and Mapping Procedures

All surgical procedures were conducted under aseptic conditions and in accordance with protocols approved by the University of Kansas Animal Care and Use Committee (KU IUCAC protocol no. 2004-1470). Seven adult squirrel monkeys (genus Saimiri; (Nudo and Milliken 1996) were used in this study. Briefly, neurophysiologic procedures were conducted to functionally identify the cortical region for the intended infarct. These procedures utilized intracortical microstimulation (ICMS) mapping techniques to define the motor representations of the primary motor cortex (M1). Then, four monkeys received an experimental ischemic infarct in the primary motor cortex hand area (Experimental group) while three monkeys underwent the motor mapping procedures but did not receive the infarct (Control group). At three days after infarct, all monkeys were killed and histologic procedures were conducted.

Before surgery, hand preference for each animal was assessed to determine the dominant motor hemisphere for the craniotomy procedure (Plautz et al, 2000). On the day of surgery, ketamine was administered as a pre-anesthetic agent (20 mg/kg im), the trachea intubated, and the saphenous vein catheterized for i.v. administration of fluids and drugs. Lactated Ringers with 5% dextrose was continuously administered for the remainder of the procedure (~10 cm3/kg/h i.v.). Physiologic vital signs were constantly monitored and body temperature regulated by a homeothermic blanket. The animal was placed in a stereotaxic frame and halothane/nitrous oxide anesthesia (25% O2; 75% N2O; 1.5 to 2.0% halothane) administered for the duration of the surgical procedure. Penicillin (0.15 ml, 45,000 U) was administered subcutaneously before the craniotomy. Warm mannitol (8 cm3 i.v.; 37°C) was administered to counter cerebral edema. A craniotomy (~1.5 × 1.5 cm2) contralateral to the preferred hand exposed the lateral extent of the central sulcus, and the overlying dura was excised. A plastic chamber was attached around the opening with dental acrylic and filled with sterile silicone oil (dimethylpolysiloxane, Dow 200 fluid).

A magnified image of the cortex was digitally captured using a high-resolution video camera (Cohu) and NIH Image software on a Macintosh computer, transferred to a graphics program (Canvas, Deneba Software) for use during motor microstimulation mapping procedures, and a 500 μm grid superimposed for designation of microelectrode penetration sites. In preparation for ICMS mapping procedures, halothane/nitrous oxide anesthesia was withdrawn and ketamine (~20 mg/kg/h i.v.) or valium (00.1 mg/kg/h i.v.) administered as needed to maintain a stable anesthetic state that abolished voluntary movements but allowed enough muscle tone for ICMS-evoked movements at relatively low current levels. Ketamine was delivered in this way until completion of motor mapping procedures.

Tapered and sharply beveled glass micropipettes (15 to 25 μm external diameter tip) filled with 3.5 mol/L NaCl solution (500 to 800-kΩ impedance) served as microelectrodes. They were introduced perpendicular to the cortical surface at the grid intersections, (carefully avoiding penetration of blood vessels), to a depth of ~1,750 μm (approximately layer V). The ICMS stimulus (13 200-μsecs pulses at 300 Hz, 3.3 msecs pulse interval, 39.6 msecs duration train burst) was repeated at 1Hz and the current intensity slowly increased until a visible movement was evoked (~1 min or less; maximum current = 30 μA). Movements and minimum thresholds were determined and confirmed visually by two observers. No movement at 30 μA was considered ‘no response.’ Motor mapping was completed when the entire border of the M1 hand representation, the ventral premotor (PMv) hand representation, and the lateral border of the M1 hindlimb representation were identified. It should be noted that both the M1 and PMv ‘hand’ representations include digit, wrist, and forearm movements, and might alternatively be termed a ‘distal forelimb’ representation. But for clarity and continuity, the term ‘hand’ will be used for movement representations (Dancause et al, 2005).

To aid in the co-registration of neurophysiologic and histologic data, tapered and beveled glass micropipettes (30 to 70 μm diameter) were used to inject 1 μm polystyrene beads (Polysciences Inc., Warrington, PA, USA) at the caudal and rostral corners of the lateral extent of the M1 hindlimb representation; the medio-caudal, caudo-lateral, medio-rostral, and rostro-lateral corners of the M1 hand representation; and the medial and caudal borders of the PMv hand representation. At each location, a microsyringe pump (Micro4; World Precision Instruments Inc., Sarasota, FL, USA) was used to inject a bolus of polystyrene beads (150 nL) at each of three injection depths (~1750, ~1200, and ~800 μm) below the cortical surface.

After completion of the motor mapping procedure and bead injections, halothane/nitrous oxide anesthesia was reinstated and the plastic chamber removed. Animals in the experimental group (601, 563A, 602, 653) then underwent induction of an ischemic infarct confined to the M1 hand representation through permanent occlusion of cortical vessels supplying the hand area as they entered the cortex. Microforceps connected to a bipolar electrocoagulator were used to occlude the entire vascular bed within the neurophysiologically defined zone (Frost et al, 2003). This technique consistently produces infarcts through all six layers of cerebral cortex and does not extend into white matter (Nudo and Milliken 1996). Control animals (155, 597, 598) did not undergo the infarct procedure. The dura and skull flap were replaced by gelfilm and gel foam, respectively, the opening secured with a dental acrylic cap and skin wound sutured. Topical lidocaine (~1 cm3) and an antibacterial agent (Furazolione) were applied to the skin incision. After gas anesthesia was withdrawn, the monkey was removed from the stereotaxic frame and monitored until alert. Penicillin (0.15 ml, 45,000 U) was again administered subcutaneously.

Histologic Procedures

At 3 days after surgical and neurophysiologic procedures, animals were deeply anesthetized with ketamine (20 mg/kg i.v.) followed by euthasol (1 cm3; 390mg pentobarbital sodium/ml; i.p.) and then perfused through the left ventricle with 0.1 mol/L sodium phosphate buffer, pH 7.25 and 3% paraformaldehyde. The brain was removed, the cerebral cortex separated from the rest of the brain, and each hemisphere tangentially flattened between 2 × 3 in glass slides and kept in 3% paraformaldehyde/0.1 mol/L phosphate-buffered saline for 72 h (Dancause et al, 2005). Each hemisphere was then placed between layers of trimmed biopsy pads in a Supercassette (Surgipath Medical Industries, Richmond, IL, USA), dehydrated, and paraffin embedded. In each animal, the ipsilesional hemisphere was sectioned tangential to the cortical surface at 12 μm on a rotary microtome. In contrast to the brains of many primates, including humans, the brains of squirrel monkeys are lissencephalic (i.e., smooth-brained). The lack of multiple gyri in the frontal cortex was a particular advantage in the histologic techniques used here, in which functional areas (e.g., primary motor and premotor hand areas) could be identified on the cortical surface. In turn, the locations of these regions could be identified in individual tangential sections from the flattened cortex.

Approximately 75 sequential sections were collected, with blocks of 50 consecutive sections reserved for stereology. In each animal, five sections, each 10 sections apart, were selected for stereological analysis and yielded results with acceptable coefficient of errors (see below). The initial section of this series was randomly chosen from the first 10 of the 50 total sections.

Sections were cleared in xylene, rehydrated through graded alcohols, and placed in Antigen Unmasking solution (2.4 L; 100°C; Vector labs, Burlingame, CA, USA). Then sections were rinsed in 10 mmol/L phosphate- buffered saline, pH 7.5, exposed to 3% H2O2 for 5 mins, then blocked with 3.0% normal horse serum, 10 mmol/L phosphate-buffered saline (Vector). Sections were exposed to primary antibodies to stain for (a) neurons, using mouse anti-neuronal nuclear (NeuN) (1:100; Chemicon, Temecula, CA, USA) and (b) VEGF protein using mouse anti-human VEGF-164 (1:100; Sigma, St. Louis, MO, USA) at 37°C. All primary antibodies were suspended in (0.05%NGS; 0.3% Triton; 0.005% Na azide; 10 mmol/L phosphate-buffered saline, 7.5 pH) antibody diluent. Procedures for blocking and secondary labeling followed the Rapid Antibody Procedure in the mouse Vectastain ABC Elite kits (Vector labs). Visualization for neurons used Vector SG stain and VEGF protein used diaminobenzidine stain (Vector labs). Negative controls for section staining consisted of omission of primary antibody. Sections were dehydrated and coverslipped with DPX mounting medium (Fluka, Steinheim, Germany). Adjacent sections to those chosen for the stereological procedures were stained for Nissl substance following standard protocols. Serial sections were cleared and rehydrated, placed in cresyl violet (15 to 20 mins), then rinsed and dehydrated through graded alcohols. Sections were placed in xylene, then coverslipped with DPX mounting medium.

Stereological Analysis (Optical Fractionator Probe)

All histologic sections were analyzed using an Axioplan 2 Microscope (Zeiss, Jena, Germany). Borders of the ICMS-defined motor areas (M1 hand, PMv hand, M1 hindlimb) were identified in every section using the reliable and reproducible location of the polystyrene bead injection sites. The PMv hand representation was chosen as a representative secondary motor area as it is heavily interconnected with M1 hand, does not have a hindlimb representation, and has been shown to undergo anatomic and physiologic changes in response to an M1 infarct (Frost et al, 2003; Shimazu et al, 2004; Dancause et al, 2005). The general location of each of the areas not identified using ICMS techniques (M1 orofacial region, somatosensory cortex, and temporal cortex) was identified using topographical references. In particular, the M1 orofacial region and somatosensory cortex was designated based on published experiments and previous studies conducted in this laboratory (Huang et al, 1988; Dancause et al, 2005). After identification, all areas were outlined using Stereo Investigator (Microbrightfield, Colchester, VT, USA; Figure 1B). Approximate x- and y-distances between sampling sites within sections were determined in each separate neurophysiologic area.

Figure 1.

Figure 1

Open in a new tab

Methodology for the use of stereology in neurophysiologically define cortical areas. (A) Flattened cortex was tangentially sectioned at 12 μm thickness. Note the presence of cingulate (C), parietal (P), frontal (F), and temporal (T) cortices. Central sulcus (CS) is outlined in red. (B) Same section as (A), outlined in Stereo Investigator. The M1 hand representation is masked (black), and the peri-infarct region is depicted as the surrounding black outline. Remote areas that were analyzed are also labeled. Small white box near M1 in (B) is magnified in (C). (C) Red circle outlines the polystyrene bead injection used to identify the M1 hand/peri-infarct region in (B) at the rostrolateral border of M1 hand representation. Note the tissue damage within the area of the injection site. Scale bar: (A and B)=1 cm; (C)=100 μm.

Each section was then assessed with the Optical Fractionator (West et al, 1991). An Optical Fractionator count represents the systematic sampling of neurons for each cortical area, in a series of serial sections, allowing an unbiased estimate of the population number in a volume (Stereo Investigator). One marker was placed on every neuron within the field that had a robust appearance and an intact cell membrane, and another marker was placed on each neuron that also had positive staining for VEGF protein. A minimum of 150 counts was used to establish the density of cells within each physiologic area. The Gundersen coefficient of error (Gundersen et al, 1999) for total neuron counts was between 0.04 and 0.08 for all areas in either control or experimental animals, except for the M1 hand infarct in the experimental animals. This area had a CE range of 0.09 to 0.26, because of the significant loss of neurons.

Statistical Analysis

As each of the physiologic regions selected for examination varied in surface area both within and between animals, raw neuronal population estimates derived using Stereo Investigator (total neurons and neurons that associate with VEGF protein) were converted to neurons per unit volume (mm3) to facilitate comparisons. One-way analysis of variance was used to compare the total neurons/mm3 between groups (experimental and control; arcsine transformed; Zar, 1984) for each physiologic region (JMP statistical software). The ratio of VEGF-associated neurons to total neuronal population was calculated for each physiologic region in each group. Values are reported as mean ± s.d. Statistical outcomes were considered significant for P≤0.05.

Results

Intracortical microstimulation mapping procedures identified the entire extent of the hand representation in M1 and PMv, as well as the lateral border of the distal hindlimb representation in M1 (Figure 1B; PMv does not contain a hindlimb representation (Preuss et al, 1996; Frost et al, 2003). Immediately after the infarct procedure, the borders of the infarct were clearly delineated by visual observation of occluded surface vasculature and subsequent blanching of infarcted cortical tissue. Based on this criterion, the ischemic infarct was confined exclusively to the M1 hand representation. On removal of the brain at three days after infarct, the infarct contained blood-filled surface vessels in the cortical tissue within M1. Histologic examination of bead injection sites verified that the lesion was confined to the M1 hand representation (Figure 1C). In both immunostained and cresyl violet-stained sections, the infarct was further characterized by a reduced volume of cortical tissue and the presence of pathological neurons, as indicated by shrunken, irregularly shaped nuclei (Figure 2A, 2B, 2D, and 2E).

Figure 2.

Figure 2

Open in a new tab

VEGF protein association to neurons in cortical gray matter. Representative photomicrographs of tangential histologic sections through the cerebral cortex in infarct and control animals. Sections depicted in (B, E, and I) underwent cresyl violet staining. All other sections underwent immunohistochemical staining for VEGF protein (diaminobenzidine; brown) and NeuN (Vector SG; gray). (A and B) Low-power photomicrographs of adjacent sections along the border of the M1 hand representation. Asterisks mark the same blood vessels in the two sections; black line indicates the border of the infarct (infarcted tissue above the line). Note sparse staining of neurons within the infarct. (C) Comparable section along the border of the M1 hand representation in cerebral cortex of a control animal (no infarct). (D and E) Higher magnification of infarcted tissue in (A and B), respectively. Arrow in (D) indicates a neuron that has positive staining for VEGF protein, but a shrunken and pathologic nucleus. This neuron would not be counted as positive for stereological estimations. (F) Higher magnification of section from the control M1 hand area shown in (C). Arrowhead points to a neuron with minimal VEGF protein association, but this, and all neurons identified by arrowheads in this figure, would be counted as positive for stereological estimations. (G and H) Higher magnification of the peri-infarct tissue shown below the black line in (A). Arrowheads identify neurons with positive staining for VEGF protein. Black line in (G) encloses the walls of a blood vessel with positive VEGF staining. Asterisk in (H) marks a non-neuronal cell type with positive staining for VEGF protein. (I) Higher magnification of the peri-infarct tissue below the black line (B). (J and K) Low- and higher-power photomicrographs (respectively) through the PMv hand representation from the section depicted in (A). Black line in (J) outlines the injection site used for identification of the area. Arrowhead in (K) identifies a PMv neuron with positive staining for VEGF protein. (L) Comparable magnification (compared with K) of the PMv hand representation in cortex from the control animal illustrated in (C). Scale bar: (A–C, and J)=100 μm; (D–I and K, L)=25μm.

In both infarcted and non-infarcted tissue, NeuN staining resulted in a dark gray nucleus, with a white nucleolus sometimes visible, as well as lighter gray staining of the cytoplasm (Figure 2D, 2F, 2H, and 2K). Neuronal staining patterns were similar to those in adjacent sections in neurons stained using cresyl violet techniques (Figure 2E and 2I). Staining patterns for VEGF protein appeared to be predominantly neuronal. Vascular endothelial growth factor protein was spatially confined to a peri-cytoplasmic and membrane location in cells that also positively stained for NeuN. Sections stained in the absence of primary antibodies had no discernable staining patterns (data not shown).

Neuronal Population Estimates Three Days after Infarct

Neurons (NeuN-positive cells) and neurons associated with VEGF (NeuN-positive cells that also positively stained for VEGF protein) were systematically counted using the Optical Fractionator. Within each cortical area, an estimate of total population count was calculated from the analysis of serial sections (Table 1). Table 2 depicts population density estimates for both neurons and neurons associated with VEGF protein that were identified using immunohistochemical techniques. Total neurons/mm3 within the M1 hand representation (i.e., the target of the infarct) were significantly lower in the experimental versus the control group (Figure 3). The mean (± s.d.) density of neurons for the M1 hand representation in control animals was 42,924 ± 5,728 neurons/mm3 while the mean number in experimental animals was 5,502 ± 4,223 neurons (F = 99.64, P < 0.001). These means reflect an 87.2% reduction of neurons in the M1 hand area at 3 days after the ischemic infarct. There were no significant differences in neuronal densities between control and experimental groups in any of the other cortical regions (the peri-infarct, M1 hindlimb, M1 orofacial, PMv hand, somatosensory region, and temporal regions; P > 0.05; Figure 3).

Table 1.

Stereological results as estimated by the optical fractionator

Animals ID M1 hand Peri-infarct PMv hand M1 hindlimb M1 orofacial Somatosensory temporal
Control 155 Neurons with VEGF 49 535 110 347 24 861 57 172 N/A 36 390 47 888
Total neurons 702 343 1 392 479 258 089 724 942 N/A 777 617 1 176 267
597 Neurons with VEGF 95 363 170 398 37 595 90 409 34 758 66 367 82 388
Total neurons 947 171 1 683 939 341 906 1 226 325 444 053 1 099 454 1 453 033
598 Neurons with VEGF 66 092 112 369 18 314 82 902 24 419 34 249 54 416
Total neurons 571 704 1 102 627 183 142 796 582 206 205 494 383 899 306
Infarct 601 Neurons with VEGF 10 768 167 170 40 731 147 684 29 622 43 935 45 540
Totalneurons 32 304 1 118 187 233 806 1 045 849 432 701 783 982 721 050
563A Neurons with VEGF 58 940 257 507 76 074 157 687 47 250 70 973 32 226
Total neurons 176 821 1 382 910 452 893 731 402 498 459 606 822 527 490
602 Neurons with VEGF 11 305 236 847 84 249 242 662 55 042 46 645 59 896
Total neurons 28 264 1 402 027 457 192 1 523 139 615 581 407 931 1 550 259
653 Neurons with VEGF 51 608 207 934 54 329 178 541 44 108 50 841 51 583
Total neurons 124 059 858 722 295 313 1 022 553 470 672 497 413 1 081 856
Open in a new tab

Table shows total neuron and neuron with VEGF estimates for each physiological area from each animal at 3 days following a focal infarct in the M1 hand representation. Neuron and VEGF-neuron data represent cell population estimates calculated by Stereo Investigator from serial section population counts for each animal. M1 orofacialcial region of animal 155 was not present in the histological sections (NA).

Table 2.

Population densities and percent ratios for VEGF-associated neurons and total neurons in physiological areas

M1 hand (infarct) Peri-infarct PMv hand M1 hindlimb M1 orofacial Somatosensory temporal
Control (n = 3) VEGF-neurons/mm3 4 116 4 523 4 377 3 589 5 089 4 181 3 519
Neurons/mm3 42 924 47 982 42 637 42 026 53 716 71 702 66 613
Percent ratio 9.6 9.4 10.2 8.6 9.8 5.9 5.3
Distance (mm) 2.58 5.06 6.91 4.28 5.85 16.55
Infarct (n = 4) VEGF-neurons/mm3 2 086 9 737 8 517 8 491 6 159 6 823 2 786
Neurons/mm3 5 502 53 466 47 930 50 988 73 513 75 143 55 989
Percent ratio 37.1 18.7 17.8 17.3 8.7 9.7 5.3
Distance (mm) 2.61 6.59 6.67 5 5.49 16.41
Open in a new tab

Total estimates are given for neurons and neurons that associate with VEGF protein in each of the physiologic areas for control and experimental groups (population/mm3). These estimates represent the mean for all animals per group. Percent ratio is the total neuron estimate divided by the VEGF/neuron estimate (mean ± s.d.). Distance (mm) was measured from the center of the M1 hand representation to the center of each of the remaining physiologic areas. Peri-infarct distance represents an additional mean of four distances (0°, 90°, 180°, 270°) as it surrounds the M1 hand representation.

Figure 3.

Figure 3

Open in a new tab

Neuronal population estimates in physiologic areas. Population estimates are shown as total neurons/mm3 and represent the mean (mean ± s.d.) of population numbers from all animals in either experimental or control groups as calculated by Stereo Investigator. Significant variation between experimental and control populations was identified only in M1 hand area, as identified by asterisk (one-way analysis of variance, P<0.05).

Vascular Endothelial Growth Factor Protein Association to Neurons 3 Days after Infarct

In general, a relatively small proportion ( <10%) of neurons associated with VEGF protein in control animals (Table 2). The mean proportion of neurons associated with VEGF protein was higher in experimental animals in each of the regions except for the orofacial region and temporal cortex (Figure 4). Group differences in the density of VEGF protein association to neurons were examined for each of the seven cortical regions. Although there was a significant loss of neurons in the infarcted M1 hand area (see above), the association of VEGF protein with neurons increased significantly, from 9.6% ± 2.3% in the control group to 37.1% ± 4.4% in the experimental group (F = 106.96; P < 0.0001). This difference represented over a fourfold increase in neurons associated with VEGF within the infarct zone. In addition, VEGF-associated neurons were significantly increased in the peri-infarct region (9.4% ± 1.3% to 18.7% ± 4.0%, control versus experimental; F = 17.87; P < 0.01), the M1 hindlimb representation (8.6% ± 1.3% to 17.3% ± 3.2%; F = 21.79; P < 0.01) and in the PMv hand representation (10.2% ± 0.7% to 17.8% ± 0.8%; F = 174.00; P < 0.0001). There were no significant differences in VEGF protein association between experimental and control groups in the M1 orofacial region, the somatosensory cortex, or temporal cortex (P > 0.05; Figure 4).

Figure 4.

Figure 4

Open in a new tab

Percent ratios of VEGF association to neurons. Percent ratios are based on cell populations derived from stereological analysis of neurons. The ratio represents the amount of neurons that associate with VEGF protein to total neuron estimations (mean ± s.d.). Exact ratios are given in Table 2. Note that, though there is a significant increase in VEGF association within the M1 hand representation infarct, the total neuron population for this area was significantly reduced (87.2% fewer neurons) versus all other areas depicted (one-way analysis of variance, P<0.05).

Discussion

Stroke has become a leading cause of adult disability, and great effort is being expended in developing new treatments to aid in behavioral recovery. To further this aim, there needs to be a better understanding of the early physiologic responses of the intact regions of the cortex that are not directly damaged by the ischemic/hypoxic injury. As an initial step, this study investigated VEGF as one of many possible neurotrophic and angiogenic factors that alter cortical activity in response to an infarct, and determined its association to neurons throughout the ipsilesional hemisphere in a non-human primate model of focal cortical ischemia.

In agreement with results in rodent studies, VEGF protein significantly increased association to neurons in the infarct (i.e. M1 hand representation) and peri-infarct regions at 3 days after infarct induction (Hayashi et al, 1997; Yang et al, 2002). Unique to the present study, however, was the ability to evaluate changes in functionally defined remote motor areas that are thought to be involved in recovery of function (Frost et al, 2003; Nudo 2003; Dancause et al, 2005). The results indicate that VEGF protein significantly increased association to neurons in both the PMv hand and M1 hindlimb representations, but not in two areas closer in proximity to the area of infarct, the M1 orofacial region, and the somatosensory region. These results reveal a previously unappreciated complexity of VEGF neuronal association in spared cortical tissue after stroke. This research is novel, not only because of the examination of remote, intact tissue in a non-human primate species, but because it employed stereology as design based quantitative technique to assess neuronal protein associations.

The functional role of the increased association of VEGF protein to neurons in brain regions remote from the site of damage is not well known. Based on the existing rodent literature, one might presume that VEGF plays a neuroprotective role, especially at early time periods. Remote areas, such as the PMv hand region, lose reciprocal connectivity with the area of infarct (Dancause et al, 2005). Intracortical neurons may therefore require neuroprotection, either because of the sustained loss of retrograde neurotrophin transport from terminations in the infarct, or the loss of reciprocal excitatory input from neurons originating in the infarct (Skaper and Walsh 1998; Stein 1998). It should be noted, though, that increased VEGF protein association might also be deleterious. Vascular endothelial growth factor could initiate vascular leakage and edema in remote areas, as it has already been shown to exacerbate periinfarct edema when exogenously administered during acute stroke in rodents (Zhang et al, 2000). The increased association of VEGF protein to neurons in the PMv hand area occurs despite isolation from the influence of the actual ischemic insult.

While neurons in the PMv hand area may initiate molecular cascades that release VEGF for immediate neuroprotection, VEGF may also serve long-term functions within the area. Perhaps most provocatively, VEGF may play an angiogenic role during chronic stages of stroke, and thus contribute directly or indirectly to functional recovery. Studies have demonstrated that the PMv hand representation undergoes neurophysiologic and neuroanatomic reorganization after ischemic injury to M1 in this model of stroke (Frost et al, 2003; Dancause et al, 2005). The physiologic and anatomic changes in PMv that occur in the first few weeks and months after an infarct in M1 most likely impose greater metabolic demands on the local cell populations. These plasticity-associated demands could, in turn, induce angiogenesis, as has been seen in the motor cortex of rats under the metabolic stress of prolonged exercise (Swain et al, 2003).

To summarize, recent research suggests the possibility of ‘dual’ roles for VEGF in both neuroprotection and angiogenesis (Manoonkitiwongsa et al, 2004). It is difficult to determine, in this single series of experiments early after infarct, through which stimulus (e.g., interruption of interconnectivity, loss of neurotrophin transport, etc.) neurons in the PMv hand representation prompt the significant increase in VEGF protein association or if it is functionally linked to later neuroanatomic and physiologic plasticity.

In these experiments, analysis of VEGF association with neurons in the M1 hindlimb representation served as a control for plasticity in the PMv hand representation. Unexpectedly, neuronal VEGF protein levels were significantly increased in this physiologic area as well, even though the M1 hindlimb representation is not heavily interconnected to the M1 hand representation (Hatanaka et al, 2001). One possible initiator of increased VEGF association in the M1 hindlimb region may be an increase in collateral blood flow. Doppler blood flow analysis of the motor areas in the squirrel monkey showed that the medial aspect of the M1 hindlimb representation includes arteries supplied by the anterior cerebral artery (unpublished observations). Perhaps compensatory changes in collateral anterior cerebral artery blood flow play a mechanical role in the induction of VEGF protein expression in the area, and this is reflected in the increased VEGF association to neurons.

Both the infarct and peri-infarct regions had significant increases in VEGF protein association to neurons at 3 days after infarct. Using an electrocoagulation method to produce local and discrete vascular occlusion in this study, only 11% of neurons remained in the infarct core at 3 days after infarct, while there was no significant loss of neurons in the peri-infarct region. However, both the infarct and peri-infarct regions had significant increases in VEGF protein association to neurons. Of the neurons within the infarct zone that were considered viable (based on immunohistochemical staining), 37.1% associated with VEGF protein, a greater than fourfold increase over the 9.6% of neurons associating with VEGF in control tissue. Neurons within the peri-infarct region underwent a twofold increase in association with VEGF protein (from 9.4% to 18.7%). This finding in peri-infarct tissue is similar to previous results in rodent models, and provides further support for the concept that increases in neuronal VEGF are independent of events resulting in early cell death. To underscore the importance of post-infarct VEGF, Issa et al (1999) found sustained increases in peri-infarct neuronal VEGF in human gray matter as far removed as 150 days post-infarct.

While the present non-human primate model is well suited to monitoring ipsilesional remote changes in functionally related motor areas after an infarct, there are limitations to its use. Foremost is the inability to accurately assess the layer depth for each tangential section. Regression analysis showed no trends of VEGF protein association to neurons by depth, though the data was not presented because of the inaccuracy of depth assignment. Alternatively, coronal or saggital sections could have been obtained that would have made laminar analysis more accurate, but would have rendered the spatial preservation of the ICMS-derived maps difficult to maintain. The use of the vascular electrocoagulation techniques for infarct induction may alter VEGF production within the infarct zone in a yet uncharacterized manner (Nudo and Milliken 1996), but because there is not a significant loss of neurons in the peri-infarct region, this method of infarct induction should not inadvertently affect remote areas.

In addition to remote areas in the ipsilesional cortex, neuronal VEGF levels should be assessed in the contralateral primary and premotor cortices, as both experimental and clinical data support the role of the contralateral cortex in recovery of function. The contralateral hemisphere was not examined in the present study, since neurophysiologic identification of cortical regions would have required an excessively long surgical procedure. Finally, conclusions regarding the physiologic and anatomic consequences of VEGF upregulation in remote areas are still premature. It will be important to use immunohistochemical and in situ mRNA hybridization methods in this model to determine the effects of VEGF neuronal association on cellular metabolism and neuronal function.

While we have demonstrated that VEGF significantly increases association to neurons in remote motor areas after an ischemic infarct, the functional significance of this finding has not yet been determined. The occurrence of post-stroke plasticity in the premotor areas, as well as the behavioral role of the premotor areas during functional recovery, is well established (Seitz et al, 1998; Liu and Rouiller 1999; Frost et al, 2003; Dancause et al, 2005). Further experiments, however, should attempt to determine the influence of VEGF protein in these processes. In particular, the present non-human primate model would be well suited to understanding the ultimate behavioral outcome of these early events, and the effects of modulating VEGF protein using pharmacothera peutic interventions.

Acknowledgments

We thank Robert Cross, Erica Blumberg, and Phuong Nguyen for assistance in data acquisition and analysis. We also thank Dr Lutz Slomianka, University of Zurich, Switzerland, for help in establishing the stereology protocols.

Footnotes

American Heart Association Predoctoral Fellowship (Stowe, PI) and NIH/NINDS R01 NS 30853-10 (Nudo, PI).

References

  1. Bergeron M, Yu AY, Solway KE, Semenza GL, Sharp FR. Induction of hypoxia-inducible factor-1 (HIF-1) and its target genes followingfocal ischaemia in rat brain. Eur J Neurosci. 1999;11(12):4159–70. doi: 10.1046/j.1460-9568.1999.00845.x. [DOI] [PubMed] [Google Scholar]
  2. Dancause N, Barbay S, Frost SB, Plautz EJ, Chen D, Zoubina EV, et al. Extensive cortical rewiring after brain injury. J Neurosci. 2005;25:10167–79. doi: 10.1523/JNEUROSCI.3256-05.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Frost SB, Barbay S, Friel KM, Plautz EJ, Nudo RJ. Reorganization of remote cortical regions after ischemic brain injury: a potential substrate for stroke recovery. J Neurophysiol. 2003;89:3205–14. doi: 10.1152/jn.01143.2002. [DOI] [PubMed] [Google Scholar]
  4. Gundersen HG, Jensen EB, Kieu K, Nielsen J. The efficiency of systematic sampling in stereology—reconsidered. J Microsc. 1999;193:199–211. doi: 10.1046/j.1365-2818.1999.00457.x. [DOI] [PubMed] [Google Scholar]
  5. Hatanaka N, Nambu A, Yamashita A, Takada M, Tokuno H. Somatotopic arrangement and corticocortical inputs of the hindlimb region of the primary motor cortex in the macaque monkey. Neurosci Res. 2001;40:9–22. doi: 10.1016/s0168-0102(01)00210-3. [DOI] [PubMed] [Google Scholar]
  6. Hayashi T, Abe K, Suzuki H, Itoyama Y. Rapid induction of vascular endothelial growth factor gene expression after transient middle cerebral artery occlusion in rats. Stroke. 1997;28:2039–44. doi: 10.1161/01.str.28.10.2039. [DOI] [PubMed] [Google Scholar]
  7. Huang CS, Sirisko MA, Hiraba H, Murray GM, Sessle BJ. Organization of the primate face motor cortex as revealed by intracortical microstimulation and electrophysiological identification of afferent inputs and corticobulbar projections. J Neurophysiol. 1988;59:796–818. doi: 10.1152/jn.1988.59.3.796. [DOI] [PubMed] [Google Scholar]
  8. Issa R, Krupinski J, Bujny T, Kumar S, Kaluza J, Kumar P. Vascular endothelial growth factor and its receptor, KDR, in human brain tissue after ischemic stroke. Lab Invest. 1999;79:417–25. [PubMed] [Google Scholar]
  9. Krupinski J, Kaluza J, Kumar P, Kumar S, Wang JM. Role of angiogenesis in patients with cerebral ischemic stroke. Stroke. 1994;25:1794–8. doi: 10.1161/01.str.25.9.1794. [DOI] [PubMed] [Google Scholar]
  10. Liu Y, Rouiller EM. Mechanisms of recovery of dexterity followingunilateral lesion of the sensorimotor cortex in adult monkeys. Exp Brain Res. 1999;128:149–59. doi: 10.1007/s002210050830. [DOI] [PubMed] [Google Scholar]
  11. Maier MA, Olivier E, Baker SN, Kirkwood PA, Morris T, Lemon RN. Direct and indirect corticospinal control of arm and hand motoneurons in the squirrel monkey (Saimiri sciureus) J Neurophysiol. 1997;78:721–33. doi: 10.1152/jn.1997.78.2.721. [DOI] [PubMed] [Google Scholar]
  12. Manoonkitiwongsa PS, Schultz RL, McCreery DB, Whitter EF, Lyden PD. Neuroprotection of ischemic brain by vascular endothelial growth factor is critically dependent on proper dosageand maybe compromised by angiogenesis. J Cereb Blood Flow Metab. 2004;24:693–702. doi: 10.1097/01.WCB.0000126236.54306.21. [DOI] [PubMed] [Google Scholar]
  13. Marti HH, Risau W. Systemic hypoxia changes the organ-specific distribution of vascular endothelial growth factor and its receptors. Proc Natl Acad Sci U S A. 1998;95:15809–14. doi: 10.1073/pnas.95.26.15809. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Marti HJ, Bernaudin M, Bellail A, Schoch H, Euler M, Petit E, et al. Hypoxia-induced vascular endothelial growth factor expression precedes neovascularization after cerebral ischemia. Am J Pathol. 2000;156:965–76. doi: 10.1016/S0002-9440(10)64964-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Nadar SK, Blann A, Beevers DG, Lip GY. Abnormal angiopoietins 1&2, angiopoietin receptor Tie-2 and vascular endothelial growth factor levels in hypertension: relationship to target organ damage (a sub-study of the Anglo-Scandinavian Cardiac Outcomes Trial (ASCOT)) J Intern Med. 2005;258:336–43. doi: 10.1111/j.1365-2796.2005.01550.x. [DOI] [PubMed] [Google Scholar]
  16. Nudo RJ. Functional and structural plasticity in motor cortex: implications for stroke recovery. Phys Med Rehabil Clin N Am. 2003;14(1 Suppl):S57–76. doi: 10.1016/s1047-9651(02)00054-2. [DOI] [PubMed] [Google Scholar]
  17. Nudo RJ, Milliken GW. Reorganization of movement representations in primary motor cortex followingfocal ischemic infarcts in adult squirrel monkeys. J Neurophysiol. 1996;75:2144–9. doi: 10.1152/jn.1996.75.5.2144. [DOI] [PubMed] [Google Scholar]
  18. Ogunshola OO, Antic A, Donoghue MJ, Fan SY, Kim H, Stewart WB, et al. Paracrine and autocrine functions of neuronal vascular endothelial growth factor (VEGF) in the central nervous system. J Biol Chem. 2002;277:11410–5. doi: 10.1074/jbc.M111085200. [DOI] [PubMed] [Google Scholar]
  19. Plate KH, Beck H, Danner S, Allegrini PR, Wiessner C. Cell type specific upregulation of vascular endothelial growth factor in an MCA-occlusion model of cerebral infarct. J Neuropathol Exp Neurol. 1999;58:654–66. doi: 10.1097/00005072-199906000-00010. [DOI] [PubMed] [Google Scholar]
  20. Plautz EJ, Milliken GW, Nudo RJ. Effects of repetitive motor training on movement representations in adult squirrel monkeys: role of use versus learning. Neurobiol Learn Mem. 2000;74:27–55. doi: 10.1006/nlme.1999.3934. [DOI] [PubMed] [Google Scholar]
  21. Preuss TM, Stepniewska I, Kaas JH. Movement representation in the dorsal and ventral premotor areas of owl monkeys: a microstimulation study. J Comp Neurol. 1996;371:649–76. doi: 10.1002/(SICI)1096-9861(19960805)371:4<649::AID-CNE12>3.0.CO;2-E. [DOI] [PubMed] [Google Scholar]
  22. Rizzolatti G, Fogassi L, Gallese V. Motor and cognitive functions of the ventral premotor cortex. Curr Opin Neurobiol. 2002;12:149–54. doi: 10.1016/s0959-4388(02)00308-2. [DOI] [PubMed] [Google Scholar]
  23. Seitz RJ, Hoflich P, Binkofski F, Tellmann L, Herzog H, Freund HJ. Role of the premotor cortex in recovery from middle cerebral artery infarction. Arch Neurol. 1998;55:1081–8. doi: 10.1001/archneur.55.8.1081. [DOI] [PubMed] [Google Scholar]
  24. Shimazu H, Maier MA, Cerri G, Kirkwood PA, Lemon RN. Macaque ventral premotor cortex exerts powerful facilitation of motor cortex outputs to upper limb motoneurons. J Neurosci. 2004;24:1200–11. doi: 10.1523/JNEUROSCI.4731-03.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Skaper SD, Walsh FS. Neurotrophic molecules: strategies for designing effective therapeutic molecules in neurodegeneration. Mol Cell Neurosci. 1998;12:179–93. doi: 10.1006/mcne.1998.0714. [DOI] [PubMed] [Google Scholar]
  26. Sondell M, Lundborg G, Kanje M. Vascular endothelial growth factor has neurotrophic activity and stimulates axonal outgrowth, enhancing cell survival and Schwann cell proliferation in the peripheral nervous system. J Neurosci. 1999;19:5731–40. doi: 10.1523/JNEUROSCI.19-14-05731.1999. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Stein DG. Brain Injury and Theories of Recovery. In: Goldstein LB, editor. Restorative Neurology: Advances in Pharmacotherapy for Recovery After Stroke. Armonk, NY: Futura Publishing Co. Inc; 1998. pp. 1–34. [Google Scholar]
  28. Sun Y, Jin K, Xie L, Childs J, Mao XO, Logvinova A, et al. VEGF-induced neuroprotection, neurogenesis, and angiogenesis after focal cerebral ischemia. J Clin Invest. 2003;111:1843–51. doi: 10.1172/JCI17977. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Swain RA, Harris AB, Wiener EC, Dutka MV, Morris HD, Theien BE, et al. Prolonged exercise induces angiogenesis and increases cerebral blood volume in primary motor cortex of the rat. Neuroscience. 2003;117:1037–46. doi: 10.1016/s0306-4522(02)00664-4. [DOI] [PubMed] [Google Scholar]
  30. Wang Y, Kilic E, Kilic U, Weber B, Bassetti CL, Marti HH, et al. VEGF overexpression induces post-ischaemic neuroprotection, but facilitates haemodynamic steal phenomena. Brain. 2005;128(Part 1):52–63. doi: 10.1093/brain/awh325. [DOI] [PubMed] [Google Scholar]
  31. Wei L, Erinjeri JP, Rovainen CM, Woolsey TA. Collateral growth and angiogenesis around cortical stroke. Stroke. 2001;32:2179–84. doi: 10.1161/hs0901.094282. [DOI] [PubMed] [Google Scholar]
  32. West MJ, Slomianka L, Gundersen HJ. Unbiased stereological estimation of the total number of neurons in the subdivisions of the rat hippocampus using the optical fractionator. Ant Rec. 1991;231:482–97. doi: 10.1002/ar.1092310411. [DOI] [PubMed] [Google Scholar]
  33. Yang ZJ, Bao WL, Qiu MH, Zhang LM, Lu SD, Huang YL, et al. Role of vascular endothelial growth factor in neuronal DNA damage and repair in rat brain following a transient cerebral ischemia. J Neurosci Res. 2002;70:140–9. doi: 10.1002/jnr.10380. [DOI] [PubMed] [Google Scholar]
  34. Zar JH. Biostatistical Analysis. 2. Englewood Cliffs, NJ: Simon & Schuster Co; 1984. [Google Scholar]
  35. Zhang ZG, Zhang L, Jiang Q, Zhang R, Davies K, Powers C, et al. VEGF enhances angiogenesis and promotes blood-brain barrier leakage in the ischemic brain. J Clin Invest. 2000;106:829–38. doi: 10.1172/JCI9369. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES