Epidural cortical stimulation enhances motor function after sensorimotor cortical infarcts in rats

Experimental Neurology 200 (2006) 356 – 370 www.elsevier.com/locate/yexnr

Epidural cortical stimulation enhances motor function after sensorimotor cortical infarcts in rats DeAnna L. Adkins a,⁎, Peter Campos b , David Quach c , Mark Borromeo c , Kellan Schallert d , Theresa A. Jones a,b a

Institute for Neuroscience, University of Texas at Austin, TX 78712, USA Department of Psychology, University of Texas at Austin, TX 78712, USA School of Biological Sciences, Section of Neurobiology, University of Texas at Austin, TX 78712, USA d Department of Chemistry and Biochemistry, University of Texas at Austin, TX 78712, USA b

c

Received 12 January 2006; revised 22 February 2006; accepted 22 February 2006 Available online 6 May 2006

Abstract This study examined whether epidurally delivered cortical electrical stimulation (CS) improves the efficacy of motor rehabilitative training and alters neuronal density and/or cell proliferation in perilesion cortex following ischemic sensorimotor cortex (SMC) lesions. Adult rats were pretrained on a skilled reaching task and then received partial unilateral SMC lesions and implantation of electrodes over the remaining SMC. Ten to fourteen days later, rats received daily reach training concurrent with anodal or cathodal 100 Hz CS or no stimulation (NoCS) for 18 days. To label newly generated cells, bromodeoxyuridine (BrdU; 50 mg/kg) was administered every third day of training. Both anodal and cathodal CS robustly enhanced reaching performance compared to NoCS controls. Neuronal density in the perilesion cortex was significantly increased in the cathodal CS group compared to the NoCS group. There were no significant group differences in BrdU-labeled cell density in ipsilesional cortex. Staining with Fluoro-Jade-B indicated that neurons continue to degenerate near the infarct at the time when cortical stimulation and rehabilitation were initiated. These data indicate that epidurally delivered CS greatly improves the efficacy of rehabilitative reach training following SMC damage and raise the possibility that cathodal CS may influence neuronal survival in perilesion cortex. © 2006 Elsevier Inc. All rights reserved. Keywords: Ischemia; Motor cortex; Skilled reaching; Cortical stimulation; Behavioral recovery; Bromodeoxyuridine (BrdU); Epidural electrical stimulation; Stroke; Motor learning

Introduction Motor rehabilitation following cortical injury improves motor function and drives restorative neuroplastic changes in denervated brain areas (e.g., Nudo et al., 1996; Jones et al., 1999; Chu and Jones, 2000; Biernaskie and Corbett, 2001). Electrically stimulating the motor cortex also may facilitate recovery of function. In humans, transcranial direct current stimulation (tDCS) and transcranial magnetic stimulation (TMS) of the infarcted hemisphere have been found to acutely improve motor function in patients with chronic motor ⁎ Corresponding author. University of Florida, McKnight Brain Institute, Department of Neuroscience, PO Box 100244, Gainesville, FL 32610-0244, USA. Fax: +1 352 379 2332. E-mail address: [email protected] (D.L. Adkins). 0014-4886/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.expneurol.2006.02.131

impairments (Pascual-Leone et al., 1996; Cantello, 2002; Hummel et al., 2005; Fregni and Pascual-Leone, 2005). In a small clinical trial, epidural cortical stimulation (CS) combined with physical therapy improved motor performance in chronic stroke patients compared to patients who received only physical therapy (Brown et al., 2006). Recent animal studies have demonstrated that cortical stimulation (CS) combined with rehabilitative training (skilled reaching) enhances motor functional outcome compared to rehabilitation alone. Following SMC lesions in rats (Adkins-Muir and Jones, 2003; Kleim et al., 2003; Teskey et al., 2003) and monkeys (Plautz et al., 2003), subdural stimulation of the motor cortex, via chronically implanted electrodes, combined with skilled reaching enhanced forelimb behavioral recovery compared to rehabilitation alone. Enhanced behavioral function in the CS and rehabilitation groups also coincided with increased dendritic plasticity

D.L. Adkins et al. / Experimental Neurology 200 (2006) 356–370

(Adkins-Muir and Jones, 2003), enlarged microstimulationevoked motor maps (Kleim et al., 2003; Plautz et al., 2003), and enhancement in the polysynaptic component of evoked potentials (Teskey et al., 2003) in ipsilesional cortex compared to rehabilitation alone. Although animal studies support the benefit of subdurally delivered CS, projected clinical trials will use epidurally placed electrodes. Therefore, it seemed essential to determine if current delivered through the dura would produce similar results as found with subdurally implanted electrodes. The primary aim of this study was to determine the effectiveness of combining epidural cathodal or anodal 100 Hz monopolar CS with reach training of the impaired forelimb following sensorimotor cortex (SMC) lesions compared to rehabilitation alone. The second aim was to evaluate whether CS plays a role in cellular proliferation or survival sufficient to affect neuronal density in the remaining SMC. Electrical stimulation applied to different brain regions can affect neuronal survival (Maesawa et al., 2004), reduce infarct volume (Glickstein et al., 2003), and enhance cell proliferation (Bengzon et al., 1997; Parent et al., 1998; Scott et al., 1998; Nakagawa et al., 2000). Brain injury alone has been found to alter cellular proliferation in the subventricular zone and hippocampus (for review, see Parent, 2003), and in some studies, these new cells have migrated to the site of injury in striatum and cortex (Magavi et al., 2000; Arvidsson et al., 2002; Parent et al., 2002). These new cells, including neurons, may integrate into existing networks and become functionally active (e.g., Nakatomi et al., 2002). Thus, it seemed reasonable to suspect that, following cortical lesions and the administration of cortical electrical stimulation, there would be an increase in the density of neurons in perilesion cortex as a result of greater survival and/or as a result of production of new neurons. In this study, adult male rats were pre-trained to proficiency on a skilled reaching task, the single pellet retrieval task. They then received unilateral endothelin-1 (ET-1; a vasoconstrictor) induced SMC lesions and epidural implantation of monopolar electrodes over remaining SMC. Ten to fourteen days after surgery, animals underwent 18 days of rehabilitative training with concurrent delivery of either cathodal or anodal 100 Hz stimulation or no stimulation. Animals received injections of bromodeoxyuridine (BrdU) every third day of rehabilitative training. The density of BrdU+ cells in perilesion cortex and striatum was examined using stereological methods, and the phenotype of cortical BrdU cells was analyzed with immunofluorescence and confocal microscopy. Neuronal density was estimated in Nissl-stained sections of the ipsilesional cortex adjacent to and underlying the site of electrical stimulation. In a subset of animals, Fluorojade-B was used to assess the possibility of ongoing neural degeneration at the time of onset of CS and rehabilitative training. Methods Animals Sixty-three adult (3 to 4 months old) male Long–Evans hooded rats were used. Rats were made tame by gentle handling

357

beginning after weaning and were housed on a 12:12-h light: dark cycle. Thirty-one rats were used to assess the effects of CS after unilateral SMC lesions. Rats were moderately food restricted (13–15 g/day) to motivate performance on the reaching task. Animals were randomly assigned to the following groups with the exception that they were matched as closely as possible for pre-operative and pre-rehabilitation performance: (1) 100 Hz cathodal stimulation during training (CSCath, n = 10), (2) 100 Hz anodal stimulation during training (CSAnod, n = 11), and (3) training with no stimulation (NoCS, n = 10). An additional 32 rats were used for a Fluoro-Jade B (FJB) neurodegeneration labeling study. These rats received SMC lesions or sham operations and were sacrificed 24 h (lesion: n = 3), 3 days (sham: n = 6, lesion: n = 9), or 14 days (sham: n = 5, lesion = 9) following surgery. All animal use was in accordance with a protocol approved by the Animal Care and Use Committee of the University of Texas at Austin. Surgical procedures Ischemic damage to the sensorimotor cortex (SMC) was created using endothelin-1, a vasoconstricting peptide, applied to the cortical surface (Macrae et al., 1993; Fuxe et al., 1997). Rats were anesthetized with a cocktail of ketamine (100 mg/kg) and xylazine (10–13 mg/kg). Rats were placed in a stereotaxic apparatus, received an incision at midline of the scalp, and the skull was removed between 0.5 mm posterior and 2.5 mm anterior to Bregma and 3 to 5 mm lateral to midline. Dura was cut with a scalpel parallel to midline from the anterior to posterior edge of the craniectomy. Then, 240 pmol endothelin-1 in 3 μl of sterile saline (0.2 μg/μl) was applied to the cortical surface at a rate of 1 μl/3 min and was left undisturbed for 10 min after the last application. This lesion method has previously been found to produce focal damage to cortical layers I-VI underlying the craniectomy (Fuxe et al., 1997; Adkins-Muir and Jones, 2003; Adkins et al., 2004; Adkins and Jones, 2005). Electrode implantation Ten minutes after the final endothelin-1 application, the craniectomy was enlarged by ~ 1 mm each at the anterior and medial edges to expose perilesion motor cortex for epidural electrode implantation. Each electrode consisted of 0.4 mm wide by 2 mm long parallel platinum wire strip contacts mounted on a 3 mm by 3 mm supporting plate extending from an electrode connector pedestal (Plastics One Inc., Roanoke, VA). The platinum contacts were placed on dura and were orientated approximately parallel to midline. Extensive preliminary data indicated that this electrode placement reliably enables postoperative stimulation-evoked contralesional forelimb and/or upper body and face movement. The electrode was configured so that one current polarity was delivered to the contacts on cortex and the other polarity's current flowed through a small metal disk placed below the skin (contact side up) near Lambda to ground the current (see Fig. 1; Northstar Neuroscience, Inc., Seattle, WA). The craniotomy was covered with gel foam and sealed with SDI Wave dental acrylic (SDI Inc., Bensenville, IL).

358

D.L. Adkins et al. / Experimental Neurology 200 (2006) 356–370

and continued for the duration of the training session. The NoCS group was connected to stimulator cables during performance of the reaching task, but no electric current was delivered. Behavioral methods Reach training apparatus

Fig. 1. Monopolar electrode. A monopolar electrode used to deliver 100 Hz cortical stimulation. Chronic electrodes were implanted so that contacts (white arrows) rested on the dura overlying remaining sensorimotor cortex (SMC). The ground (black arrow) was implanted caudally facing dermis. Current was delivered during performance of the skilled reaching task.

The electrode was secured to the skull using skull screws and cemented with a combination of standard dental cement and 3M Wave UV curing dental acrylic.

The single pellet retrieval task was conducted in a clear Plexiglas box (26 cm long × 28.5 cm high × 16 cm wide) with a 1 cm wide by 20 cm long window at the front of the box, extending from the floor towards the top of the box. The apparatus design (Fig. 2) was adapted from Miklyaeva and Whishaw (1996), McKenna and Whishaw (1999), Peterson and Devine (1963), Withers and Greenough (1989) and is described in more detail in a previous study (see Bury and Jones, 2002). Rats were trained to reach through the window to retrieve a single palatable food piece (45 mg banana-flavored pellet, Bioserve, Inc., Frenchtown, NJ) placed on a shelf (11.7 cm long × 5 cm wide × 3 cm high) in one of two shallow wells aligned 1 cm from each edge of the window. Reaching with the contralesional (impaired) forelimb was effectively enforced by the insertion of an inner chamber wall ipsilateral to the reaching limb and by placement of the pellet in a well opposite the impaired limb. A 2 mm diameter bar was attached to the shelf to discourage rats from scraping pellets into the chamber.

Stimulation procedures Reach training procedure Assessment of movement thresholds Movement threshold was defined as the minimal current necessary to produce visible movements of the forelimb and/or face contralateral to the lesion. On training days 1, 7, 14, and 18, rats were placed into a transparent cylinder and observed while series of 3 s trains of 1 ms pulses of increasing current levels were delivered at a frequency of 100 Hz and with groupappropriate polarity. No stimulation controls (NoCS) did not undergo movement threshold testing because such testing itself can change movement thresholds (Adkins-Muir and Jones, 2003; personal communication by G. C. Teskey) that may indicate changes in the electrophysiological response properties and neural activity patterns of perilesion motor cortical neurons. Daily cortical stimulation levels for rehabilitation were delivered at 50% of the movement threshold. Cortical stimulation with training Ten to fourteen days following surgery, rehabilitation on the single pellet retrieval task was initiated concurrent with cathodal, anodal, or no cortical stimulation. On each day of training, rats were attached to stimulator cables and placed into the reaching chamber. The frequency (100 Hz) and polarity (anodal or cathodal) of CS were chosen based on previous findings indicating these to be particularly effective stimulation parameters (Kleim et al., 2003; Teskey et al., 2003). Stimulation was delivered at 50% of that week's movement threshold. CS was initiated when animals first approached the reaching window

Rats were trained for 60 trials per day or 15 min, which ever occurred first. A reaching trial began with the placement of a pellet in a well and ended when the rat grasped the pellet and brought it to its mouth for consumption (success), dropped the pellet within the chamber before eating it (drop), or either knocked the pellet from the well or missed it after 5 reach attempts (miss). After each trial, rats turned around to eat a single food pellet dropped in the back or side of the cage so that their reaching posture was “reset” for the next trial. Performance was measured as the percent of successes out of the total number of reach attempts [(total successes / total reach attempts) * 100]. Prior to lesion/electrode implantation surgeries, animals were pre-trained to a minimum criterion of 30% success rate for two consecutive days. Eight to twelve days following surgery, all animals received 2 training sessions while attached to electrode cables but without stimulation current. Animals were then trained for 18 consecutive days, while receiving group appropriate CS or no stimulation. Rats then had 1 day of rest followed by two consecutive days of training without CS. To simplify the data presentation and analyses, data were pooled within animals for every two consecutive days of training. Forelimb postural-motor asymmetries The Schallert Cylinder test was used to detect lesion-induced decreases in the proportionate use of the impaired forelimb during upright postural support (e.g., Schallert et al., 2000). Rats were placed into a clear Plexiglas cylinder and videotaped for

D.L. Adkins et al. / Experimental Neurology 200 (2006) 356–370

359

Fig. 2. Skilled reaching task. A rat reaching for (A) and retrieving (B) a pellet in the single-pellet retrieval task.

2 min prior to SMC lesions (Pre-Lesion), the day before the initiation of CS and rehabilitation (Post-Lesion), and following the last day of CS and rehabilitation (Post-Rehab). During slow-motion digital video play back, the first 30 instances of sole use of either forelimb (ipsilateral or contralateral to the lesion) or simultaneous use of both forelimbs (bilateral) for upright support against the cylinder wall were recorded. The % of contralateral forelimb use was computed using the formula: %[(contralateral + bilateral forelimb support / ipsilateral + contralateral + bilateral forelimb support]. BrdU injection protocol Bromodeoxyuridine (BrdU; Sigma, St. Louis, MO) was dissolved in sterile 0.9% NaCl at a concentration of 12.5 mg/ml. Animals received injections of BrdU (50 mg/kg, i.p.) following the completion of every third day of rehabilitative training (days 2, 5, 8, 11, 14, and 17).

Immunohistochemical processing Immunohistochemistry for BrdU was performed on three adjacent sets of 4–5 50 μm free-floating coronal sections including the forelimb representation region of the motor cortex. One set of sections was counterstained with Toluidine blue and used for BrdU-immunoreactive cell density quantification and the other two sets of sections were processed for immunofluorescent double-labeling with neuron-specific nuclear protein (NeuN) to identify mature neurons and with glial fibrillary acidic protein (GFAP) to identify astrocytes. To verify the specificity of antibody labeling, each batch of immunohistochemically processed tissue included a section processed as described below minus the primary antibody (no-primary controls). One section containing the dentate gyrus was also included as a type of positive control for BrdU-immunoreactive (IR) cells. Transmitted light BrdU immunohistochemistry

Histology Animals were transcardially perfused with 0.1 M sodium phosphate buffer followed by 4% paraformaldehyde solution in the same buffer. Rats in the CS experiment were perfused within 24 h after the last training day. Animals used to assess neurodegeneration were perfused 24 h, 3 days, or 14 days after ET-1 induced SMC lesions or sham operations. For both experiments, six rostral–caudal sets of 50 μm coronal sections through the cerebrum were collected and stored in cryoprotectant solution at −20°C before use. For the neurodegeneration experiment, one set of sections was stained with Fluoro-Jade B. In the CS study, one set of sections was stained with Toluidine blue (a Nissl stain) and used for lesion verification, volume estimates, and neuronal density measurements and three sets of adjacent sections were processed for immunohistochemistry, as described below.

Sections were placed in 0.3% H2O2 in 0.01 M phosphatebuffered saline (PBS) at room temperature for 30 min to exhaust endogenous peroxidase activity. DNA was denatured by treating sections for 15 min at 37°C with 2 N HCL. Sections were rinsed for 10 min in 0.1 M boric acid (pH 8.5) to neutralize the HCL. Sections were then rinsed in PBS and placed for 2 h at room temperature in a solution to block non-specific protein binding. This blocking solution consisted of 0.4% Triton-X and 0.1% bovine serum albumin in PBS with 2% horse serum. Following rinses in PBS, adjacent sections were incubated at 4°C for 48 h in mouse anti-BrdU (1:200, Novocastra, Newcastle, UK). Following incubation, sections were again rinsed and incubated for 1 h in peroxidase-linked avidin–biotin complex (ABC kit, Vector Labs). Immunoreactivity was visualized using 3-3′ diaminobenzidine (DAB) with nickel ammonium sulfate (NAS)

360

D.L. Adkins et al. / Experimental Neurology 200 (2006) 356–370

intensification. Tissue was then lightly counterstained with a Nissl stain (Toluidine blue). Immunofluorescent BrdU double-labeling To identify the phenotype of new cells generated during the rehabilitation period, adjacent sets of 4–5 BrdU immunolabeled sections were also processed for NeuN and GFAP. Primary antibody dilutions used were 1:200 anti-BrdU (monoclonal mouse; Novocastra, Newcastle, UK); 1:200 anti-BrdU (rat monoclonal; Accurate Chemical, Westbury, NY); 1:500 antiNeuN (mouse monoclonal; Chemicon, Temecula, CA); and 1:800 anti-GFAP (rabbit polyclonal; Sigma, St. Louis, MO). Tissue processing for BrdU immunofluorescence was performed as described above with the exception that, following 48h incubation in BrdU primary, tissue was rinsed several times and then incubated with either anti-NeuN or anti-GFAP in a blocking solution consisting of 0.1% bovine serum album (BSA), 0.3% Triton-X, and 2% goat serum in PBS for 2 h at room temperature. At a dilution of 1:500, the secondary antibodies used for immunofluorescence were goat anti-rat Alexa Fluor® 488, goat antirabbit Alexa Fluor® 488, or goat anti-mouse Alexa Fluor® 568 (Molecular Probes, Eugene, OR). After several rinses in PBS, tissue was incubated in appropriate fluorescent secondary antibodies in blocking solution for 1 h in a dark container. Tissue was then rinsed in PBS and mounted onto slides. Fluoro-Jade B processing Fluoro-Jade B (Chemicon, Temecula, CA) is a polyanionic fluorescein derivative that specifically binds to degenerating neurons (Schmued and Hopkins, 2000). Three coronal 50 μm sections including the anterior extent and rostral border of the SMC lesions were selected for processing because these areas represent the approximate focus of the CS delivery in the CS and rehabilitative training portion of this study. Tissue sections were placed onto gelatin subbed slides and dried for 4 days. The Fluoro-Jade B staining protocol was conducted according to the manufacturer's instructions. Slides were immersed in a solution of 1% sodium hydroxide in 80% alcohol for 5 min, and 70% alcohol for 2 min, distilled water for 2 min, and 0.06% potassium permanganate for 30 min on a shaker table. Avoiding direct light from this step forward, slides were immersed into a solution of 0.0004% Fluoro-Jade B (diluted from a 0.01% stock solution) in distilled water for 30 min. Slides were then rinsed several times in distilled water and allowed to dry over night before cover slipping using either Krystalon (EM Science Harlaco, Gibbstown, NJ) or VectaShield (Vector, Burlingame, CA). Anatomical quantification All light-microscopy analyses were performed with the aid of a high-resolution digital camera (DVC Inc., Austin, TX) and Neurolucida software (Microbrightfield Inc., Williston, VT). BrdU-labeled cell phenotyping was performed using a confocal microscope. Excluding volume measurements, all quantitative

anatomical data were obtained in coronal sections of the remaining sensorimotor cortex between the rostral appearance of the genu of the corpus callosum and the appearance of the decussation of the anterior commissure (i.e., between approximately +1.2 and −0.26 mm anterior/posterior relative to Bregma; Paxinos and Watson, 1986). All tissue was coded to conceal the experimental conditions. Neuronal density and number estimation The optical disector method (e.g., Harding et al., 1994) was used to estimate the density of neurons in the lesion penumbral region using a 100× oil objective (final magnification 1680×). Using a total of 48 unbiased sample frames per brain (three 5625 μm2 sample frames in each of 4 zones for each of the 4 sections), neurons in all layers were counted at distances of 250 μm and 500 μm both medial and lateral to the lesion border. Neuronal density (Nv) was calculated with the following equation: Nv = ∑Q− / ∑v(frame), where ∑Q− is the total number of neurons counted and ∑v(frame) is the total sample volume (13,500,000 μm3). Neuron number was estimated by calculating the product of remaining ipsilesional volume and the neuronal density combined across sample zones (e.g., Gundersen, 1986). This latter calculation was performed to aid in the interpretation of the neuronal density results but does not accurately reflect total neuronal number because neuronal density samples include a fraction of the tissue used for volume estimates (as described below). Transmitted light microscopic analysis of BrdU+ cell density The optical disector method, as described above, was used to estimate the density of BrdU-immunoreactive (IR) cells in four ipsilesional regions: remaining SMC, a subregion of SMC medial to the lesion, the corpus callosum, and the dorsolateral striatum. Cell counts were made in 3 coronal sections using a 457,200 μm2 unbiased sample frame and a 40× oil immersion objective (final magnification 880×). BrdU-IR cells were identified as having dark brown/black nuclei (see Fig. 8). Remaining SMC was defined as dorsal motor and sensory cortex lying medial and lateral to the lesion. Samples were taken at 500 μm X–Y stepping distances in a total of 27 samples (9 samples per each section) inclusive of all layers from the medial edge of the ipsilesional motor cortex to the lateral edge of the sensory cortex dorsal to the striatum. A more focused sample was taken in cortex proximal to the lesion. BrdU-IR cell density was estimated in a total of 9 samples taken at 250 μm from the medial extent of the lesion (a sampling zone used for the neuronal density estimates). BrdU-IR cells were also counted in corpus callosum beginning lateral to the dorsal peak of the corpus callosum and continuing laterally at 500 μm stepping distances for a total of 9 samples per brain. Ipsilesional dorsolateral striatal BrdU-IR density measurements included 6 samples per brain. X–Y stepping distances were 500 μm, but sometimes had to be adjusted (to a minimum of 250 μm) because of lesion-induced striatal deformation. Three subjects were removed from BrdU light microscopy analyses due to

D.L. Adkins et al. / Experimental Neurology 200 (2006) 356–370

361

Fig. 3. Cortical stimulation concurrent with rehabilitative training enhances reaching performance. Monopolar cortical stimulation with either 100 Hz cathodal (CSCath) or anodal (CSAnod) current delivered at 50% of movement thresholds enhanced performance relative to unstimulated rehabilitation controls (NoCS) over days of training. The cortical stimulation was administered on post-lesion days 14–32 (correspond to training days 1–18). Data are means ± SEM. *P b 0.05, CSCath significantly different from NoCS; †P b 0.05, CSAnod significantly different from NoCS.

damaged tissue during processing (NoCS, n = 1 and CSAnod, n = 2). BrdU+ cell phenotyping The proportion of new neurons and glia cells was estimated in the CSCath and NoCS groups. These two groups were chosen because significant differences in neuronal density in perilesion cortex were found between these groups (Fig. 7). The physical disector method was used to estimate the density of BrdU+ cells single-labeled and double-labeled with NeuN or GFAP in 4 sample z-series stacks (total height = 15 μm) taken at each of 24 sample sites per brain and obtained using a confocal microscope (Leica SP2 AOBS). The area of the unbiased sample frame was 129,000 μm2. For each of the 4 sections, two z-series were collected each at the medial and lateral edge of the lesion and another two z-series were collected at an additional 500 μm medial to the lesions, inclusive of layers II-VI. Captured confocal images were visualized using LCS light software (Leica Microsystems, Wetzlar, Germany) at a final magnification of 300×. Data are presented as the mean proportion of BrdU-IR cells counted that were co-labeled with NeuN or GFAP.

Lesion size was inferred from volume measurements of remaining non-necrotic/non-gliotic ipsilesional cortex and the volume difference between ipsi- and contralesional cortices using Nissl-stained coronal sections. The volumes of ipsilateral and contralateral striatum were determined in order to assess potential striatal damage or atrophy. Measurements included the first section caudal to the appearance of the forceps minor of the corpus callosum (∼+2.7 mm anterior to Bregma) and eight additional caudal sections (to approximately 3.3 mm posterior to Bregma), each section was 600 μm apart. In each section, all remaining cortex and striatum were outlined using Neurolucida tracing software option at a final magnification of 17× to obtain area. This sampling scheme focuses the volume estimates in the SMC region, but does not limit the measurement to SMC because it is not possible to define the boundaries of this cortical

Lesion reconstruction and volume estimates The extent of each lesion was reconstructed onto schematic coronal sections adapted from Paxinos and Watson (1986) and the lesion placement was assessed based on the cytoarchitecture of the remaining cortical tissue and macrostructural landmarks. Individual reconstructions where then overlaid onto one template for each of the three stimulation groups to determine the outer boundaries of all lesions combined, the smallest lesion and largest lesion within each group.

Fig. 4. Movement thresholds. Movement thresholds were defined as the lowest current necessary to evoke movements. CSCath and CSAnod had a significant reduction in movement thresholds over days. There were no significant differences between groups. Data are means ± SEM.

362

D.L. Adkins et al. / Experimental Neurology 200 (2006) 356–370

Systematic point counting (reviewed in Mayhew, 1992) was then used to estimate the area of tissue outlined. Using a 40× dry object (final magnification of 800×), Fluoro-Jade B positive cells within the outline area were then counted. The density of cells was calculated as follows: Nv = ∑Q−/[a(p) × ∑P × T] where ∑Q− is the sum of Fluoro-Jade B stained cells counted, a(p) is the linear distance between points adjusted for magnification, ∑P is the sum of points falling within the measured area, and T is average section thickness (50 μm). There were no differences between sham subgroups perfused at the two time points (F(1,9) = 0.57, P N 0.05) and these groups were pooled for subsequent analyses. Fig. 5. Impaired forelimb use for postural support. Following the lesions, all groups had a reduction in the proportional use of the impaired forelimb for postural support behaviors, as assessed with the cylinder test. The CSAnod and NoCS groups, but not the CSCath group, returned to more symmetrical limb use at the post-rehabilitation time point (Post-Rehab). *CSCath significantly different from NoCS and CSAnod, P b 0.05.

region with precision. Volume was calculated using the Cavalieri method (Gundersen et al., 1988), as the product of the summed area and the distance between section planes. Analysis of neuronal degeneration using Fluoro-Jade B Fluoro-Jade B positive cells in perilesion SMC were counted in each of 3 sections in an exhaustive, rather than a systematic random, sampling scheme, because of the relative scarcity of degenerating neurons in sham-operates and at later time points following SMC lesions. At low magnification (final magnification 17×), the region of dorsal motor and sensory cortex lying medial and lateral (respectively) to the lesion was identified and outlined with the aid of Adobe PhotoShop software (Adobe Systems Inc., San Jose, CA).

Statistical analyses Behavioral data were analyzed using SPSS (SPSS, Inc., Chicago, IL) repeated-measures analyses of variance (ANOVAs). Volume and BrdU+ cell and neuronal density estimations were analyzed using one-way ANOVAs. Fisher LSD post hoc analyses were used when needed to further analyze group differences and behavioral performance on individual days. Bivariate correlations were used to assess the relationship between the immediate post-lesion and the early-training period (reaching success pooled over rehabilitation days 1–4, before groups separated in reaching performance) and the late-training period (reaching success pooled over rehabilitative training day 5 to the end of training) and cortical and striatal volume. Significance levels were set at α level 0.05. Results Rehabilitative reaching performance As shown in Fig. 3, the three groups had similar post-lesion declines in performance on the single pellet reaching task. Both

Fig. 6. SMC lesion extent and placement. The placement of lesions was similar between groups. Within each group, black lines indicate the maximal extent of all lesions combined, solid black areas indicate the largest lesion and solid grey areas indicate regions of damage common to all subjects. Numbers to the right indicate approximate coordinates in mm relative to Bregma.

D.L. Adkins et al. / Experimental Neurology 200 (2006) 356–370 Table 1 Cortical and striatal volume NoCS

CSAnod

CSCath

Cortex

Ipsilesional 101.42 ± 2.69 100.12 ± 2.88 98.01 ± 1.56 Contralesional 113.79 ± 1.72 114.10 ± 2.51 117.67 ± 1.48 Contra–Ipsi 12.37 ± 2.03 a 13.98 ± 0.37 a 19.66 ± 2.16 a, b Striatum Ipsilesional 35.58 ± 0.91 33.63 ± 0.97 34.01 ± 1.13 Contralesional 35.72 ± 1.45 34.29 ± 0.89 34.16 ± 0.77 Contra–Ipsi 0.15 ± 1.10 0.66 ± 0.90 0.14 ± 0.69 3

Data are mean ± SEM volume or volume difference (Contra–Ipsi) in mm . a P b 0.05 significantly different from NoCS. b P b 0.05 significantly different from CSAnod.

cathodal (CSCath) and anodal (CSAnod) delivery of 100 Hz stimulation during training significantly enhanced reaching performance over days of training relative to training alone (NoCS). By the end of training, both CS groups were performing at or above pre-lesion levels of performance whereas NoCS rats had more moderate improvements. In two-way ANOVAs, there were significant effects for Day (F(11,308) = 19.28, P b 0.0000001) and Group by Day interaction (F(22,308) = 2.13, P b 0.01) and a non-significant effect of Group (F(2,28) = 2.72, P = 0.08). Post hoc analysis revealed a significant overall effect of group between CSCath and NoCS (P b 0.05) and a significant difference on several days of reach training, as shown in Fig. 3. There were significant differences on several days of reach

363

training between CSAnod and NoCS (Ps b 0.05), but the overall group difference failed to reach significance (P = 0.08). There were no significant differences between CSCath and CSAnod (Ps N 0.05). Stimulation-induced movement threshold The level of electrical current necessary to evoke contralateral movements declined over days of testing for both CSCath and CSAnod groups (see Fig. 4). Repeated measures ANOVA revealed that there was a significant main effect of Day of threshold testing (F(3,57) = 7.13, P b 0.001). There were no significant differences between CSCath and CSAnod in the rate of decline [Group × Day (F(3,57) = 0.52, P N 0.05)] or in the overall movement threshold levels [Group effects (F(1,19) = 1.91, P N 0.05)]. Lesion-induced forelimb asymmetries The Schallert Cylinder test is sensitive to impairments in the use of the forelimb for postural support created by cortical, subcortical, and spinal cord injury (e.g., Schallert et al., 2000). As shown in Fig. 5, all three groups had a moderate but significant reduction in the proportionate use of the impaired forelimb 10–14 days after SMC lesions compared to pre-lesion

Fig. 7. Perilesion neuronal density in Nissl stained sections. (A–C) Photomicrographs of Nissl-stained neurons medial to SMC lesions. Representative images are from (A) NoCS, (B) CSAnod, and (C) CSCath. Scale bar is 25 μm. (D) CS groups had a greater density of neurons remaining in perilesion cortex. *CSCath significantly different from NoCS, P b 0.05.

364

D.L. Adkins et al. / Experimental Neurology 200 (2006) 356–370

Table 2 Neuronal density in perilesion cortex

Medial Lateral Proximal Distal

NoCS

CSAnod

CSCath

4.97 ± 0.23 5.54 ± 0.15 5.20 ± 0.20 5.30 ± 0.28

5.40 ± 0.15 5.79 ± 0.28 5.67 ± 0.13 5.52 ± 0.16

5.95 ± 0.29 a 5.85 ± 0.28 5.84 ± 0.2 a 5.96 ± 0.18 a

Data are mean ± SEM 104 neurons per mm3 per location relative to the lesion. a P b 0.05 CSCath significantly different from NoCS.

levels (Ps b 0.00001). In two-way ANOVAs, there was a significant interaction effect of Group × Day (F(4,56) = 3.19, P b 0.02) and main effect of Day (F(2,56) = 22.83, P b 0.00001), but no significant effect of Group (P N 0.05). All three groups had a moderate, but significant decrease in the proportionate use of the impaired forelimb following SMC lesions (“Post-Lesion”) compared to pre-lesion levels of use (Ps b 0.05). The CSAnod and NoCS groups returned to more symmetrical use of the forelimbs following rehabilitative training (“Post-Rehab”) compared to the post-lesion testing day (Ps b 0.05). In contrast, the CSCath group did not increase the proportionate use of their impaired forelimb following rehabilitation compared to their post-lesion testing day (P N 0.05). Additionally, the CSCath group used their impaired forelimb proportionately less following rehabilitation compared to the other two groups (Ps b 0.05). Histology results Lesion size and volume estimates of cortex and striatum All lesions produced damage to the SMC, including the forelimb representation area, and were similar in placement

between groups (Fig. 6). As shown in Table 1, there were no significant differences between groups in the volume of cortex or striatum in either hemisphere. Volume measurements inclusive of the sensorimotor region of the cortex revealed a similar amount of remaining ipsilesional cortex in all three groups (F(2,28) = 0.38, P N 0.05). Likewise, groups did not significantly differ in volume of contralesional cortex (F(2,28) = 1.14, P N 0.05), ipsilesional striatum (F(2,28) = 1.05, P N 0.05), and contralesional striatum (F(2,28) = 0.65, P N 0.05). There was, however, a significant difference between groups in the volume difference between remaining cortex of the infarcted hemisphere and the contralesional cortex (F(2,28) = 4.14, P b 0.05). Post hoc contrasts revealed that CSCath had a significantly greater difference between contralesional and infarcted cortex compared to NoCS (P b 0.05) and CSAnod (P b 0.05). The significant difference between contralesional and infracted cortical volume was a combined effect of CSCath tending to have the greatest contralesional volume and the least ipsilesional cortical volume. Bivariate correlation revealed that there was no significant relationship between reaching performance during either early (days 1–4) or late (days 5–18) training days and interhemispheric cortical or striatal volume differences (Ps N 0.05). Perilesion neuronal density and number measurements As shown in Fig. 7, CS groups had a greater density of neurons remaining in perilesion cortex, an effect that only reached significance in the CSCath group compared to NoCS. ANOVA inclusive of all 3 groups indicated a tendency for there to be an overall group effect for neuronal density in perilesion cortex (F(2,28) = 3.05, P = 0.06). In post hoc analysis, CSCath had significantly greater overall perilesion neuronal density

Fig. 8. BrdU labeling in ipsilesional sensorimotor cortex. (A) Photomicrograph of an ipsilesional coronal section processed for BrdU immunohistochemistry. (B) High magnification of BrdU-labeled cells from cells medial to the lesion (arrows).

D.L. Adkins et al. / Experimental Neurology 200 (2006) 356–370 Table 3 Density of BrdU-IR cells

Ipsilesional SMC Medial SMC Dorsolateral striatum Corpus callosum

NoCS

CSAnod

CSCath

564.90 ± 134.98 893.55 ± 307.31 274.01 ± 101.99 752.20 ± 168.50

671.20 ± 139.73 778.19 ± 190.34 326.63 ± 154.70 694.71 ± 131.08

637.42 ± 75.97 701.27 ± 55.08 170.64 ± 32.57 460.13 ± 57.21

Data are means ± SEM cells per mm3.

compared to the NoCS group (P b 0.05). As shown in Table 2, this effect was found in cortex medial, proximal, and slightly more distal to the lesion. CSAnod tended to have greater neuronal density proximal to the lesion (P = 0.08) compared to NoCS. There were no significant differences between CSAnod and NoCS distal to the lesion or in combined measures of perilesion neuronal density (Ps N 0.05). There were also no significant differences in neuronal density proximal or distal to the lesion or all areas combined between CSCath and CSAnod (Ps N 0.05). There was no significant group effect in the estimated total number of neurons in the remaining perilesion cortex (F(2,28) = 0.80, P N 0.05), although the CSCath did have, on average, a greater number of neurons compared to

365

NoCS. The mean ± SEM 104 number of neurons was 534.31 ± 27.80 in NoCS, 559.50 ± 20.12 in CSAnod, and 578.97 ± 26.06 in CSCath. (Note that because of the difficulty in precisely defining the boundaries of the SMC, volume measure used in this calculation included a larger cortical region than used for the neuronal density estimates, which may have diluted significant group differences in the estimated total number of neurons in the remaining ipsilesional cortex.) Density of newly proliferated cells Starting on the second day of rehabilitation, animals received injections of BrdU on this and every subsequent third day to label proliferating cells (see Fig. 8). BrdU+ cells were seen throughout the remaining SMC dorsal to the striatum and within the corpus callosum and striatum (see Table 3). There was no evidence for increased cellular proliferation as a result of CS. There were no significant differences between groups in the density of BrdU-IR cells in any of the structures examined: the ipsilesional SMC (F(2,25) = 0.21, P N 0.05), the subregion of SMC medial to the lesion (F(2,25) = 0.32, P N 0.05), the dorsolateral striatum (F(2,25) = 0.51, P N 0.05), or the corpus

Fig. 9. Proportion of BrdU-IR cells co-labeled with markers of neurons (NeuN) and astrocytes (GFAP). (A) Confocal images of BrdU (blue) and NeuN or GFAP (red) positive cells. Arrows indicate a BrdU cell co-labeled with NeuN or GFAP. Scale bar is 40 μm. (B) There were no differences between the CSCath and NoCS group in the proportion of newly proliferated cells co-labeled as neurons or glia. Data are means ± SEM.

366

D.L. Adkins et al. / Experimental Neurology 200 (2006) 356–370

callosum F(2,25) = 1.17, P N 0.05). The measurement medial to the lesion was a region in which a significant difference in neuronal density between CSCath and NoCS groups was found (Table 2). It should be noted that although there were no significant differences in BrdU-IR density, the CSCath tended to have reduced mean BrdU-IR cell densities in the medial-tothe-lesion cortex, corpus callosum, and dorsolateral striatum compared to the other groups. The lack of significant effects in these areas could be due to relatively high variance within the CSAnod and NoCS groups. Proportion of BrdU-IR cells double labeled with NeuN and GFAP We also investigated whether there were differences in the proportion of new mature neurons or glia in the CSCath group compared to the NoCS group (Fig. 9). These two groups were chosen because CSCath, but not CSAnod, had a significantly greater density of neurons compared to NoCS. The analysis indicated that the proportion of BrdU-IR cells co-labeled with

the neuron marker NeuN (F(1,18) = 0.006, P N 0.05) and GFAP (F(1,18) = 0.26, P N 0.05) did not differ between CSCath and NoCS (Fig. 8). These results suggest that the increased neuronal density in the CSCath compared to NoCS group is unlikely to be a result of the addition of new neurons. Lesion induced neurodegeneration As shown in Fig. 10, there was a significant difference in the density of Fluoro-Jade B stained cells in the neocortex of groups examined at 24 h, 3 days, and 14 days after SMC lesions compared with sham-operates. Significant neurodegeneration was evident at each time point following SMC lesions compared to sham controls (Ps b 0.05). As was expected, the density of degenerating neurons was greatest at 24 h after SMC lesion compared to 3 (P b 0.000001) and 14 (P b 0.000001) days after the lesion. There were no significant changes in the density of degenerating neurons between days 3 and 14 postlesion (P N 0.05). Although we sampled all of dorsal cortex in the SMC region, the majority of the Fluoro-Jade B cells were

Fig. 10. Density of Fluoro-Jade B-stained degenerating neurons. Photomicrographs of Fluoro-Jade B-stained neurons in the sensorimotor cortex (A, D) after a sham operation and (B, E) 24 h or (C, F) 14 days after unilateral SMC lesions. Arrows indicate degenerating neurons. (G) The greatest density of Fluoro-Jade B labeled neurons was found 24 h after SMC lesions, but labeled neurons remained significantly elevated at 3 and 14 days post-surgery compared to shams. Data are means ± SEM. *Significantly different from Sham, Ps b 0.00001. †Significantly different from 24 h, Ps b 0.01.

D.L. Adkins et al. / Experimental Neurology 200 (2006) 356–370

found either in the lesion core (24 h and 3 days) or in cortex directly bordering the lesion (see Figs. 10A–F). Discussion The main finding of this study is that 100 Hz cathodal and anodal cortical electrical stimulation (CS) applied epidurally adjacent to SMC lesions during motor rehabilitative training robustly enhances functional recovery of skilled reaching performance. Thus, like CS administered directly to the pial surface of the cortex (Adkins-Muir and Jones, 2003; Kleim et al., 2003; Plautz et al., 2003; Teskey et al., 2003), CS administered through the dura appears to be extremely effective in improving forelimb motor skill function in rats following cortical infarcts, at least when combined with rehabilitative training. The results also suggest that cathodal CS may contribute to greater perilesion neuronal density compared to unstimulated rehabilitation controls. Previous findings that kindling increases neurogenesis (reviewed in Parent, 2003; Bengzon et al., 1997; Parent et al., 1998; Scott et al., 1998; Nakagawa et al., 2000) and that, following brain injury, newly proliferated cells can migrate to the lesion site (Magavi et al., 2000; Arvidsson et al., 2002; Parent et al., 2002), raised the possibility that CS could enhance cell proliferation around the lesion. However, CS did not increase neural proliferation following SMC lesions compared to unstimulated controls. There were no significant group differences in the density of BrdU-labeled cells or in the proportion of newly generated neural or glial cells in the CSCath compared to the NoCS group, as assessed by double-labeling BrdU positive cells with NeuN and GFAP. Thus, newly proliferated cells are not likely to be responsible for the increased neuronal density in perilesion cortex found following cathodal CS. Data from a second experiment indicate that neuronal degeneration in perilesion cortex is in a temporal position to be impacted by the CS. We found that at 14 days post-SMC lesion (the time when CS and rehabilitative training are initiated) there is a population of degenerating neurons in and around the lesion, as shown in Fluoro-Jade B-stained tissue. Thus, it is possible that the greater density of neurons is a result of a greater survival of vulnerable neurons in the CSCath group. Although this study was not designed to directly investigate whether 100 Hz CS was neuroprotective, other studies have indicated that electrical stimulation can protect neurons during insult (Glickstein et al., 2003; Maesawa et al., 2004), reduce lesion volume, and elevate γ-aminobutyric acid (GABA) levels in ipsilesional cortex (Gan et al., 2005), facilitate peripheral axonal regrowth (Brushart et al., 2002), and reduce excitotoxicity (e.g., Canavero et al., 2003; Franzini et al., 2003). CS may also contribute to more effective neuronal activation. In vitro studies have shown that low-amplitude extracellular electrical stimulation of cortical (Purpura and McMurtry, 1965; Nowak and Bullier, 1998) and hippocampal (Bikson et al., 2004) neurons depolarizes dendrites and lowers membrane potential and thus increases the likelihood of neural firing potentials. Human studies with transcranial direct current stimulation of the motor cortex have shown that enhancements of motor evoked potentials are

367

dependent upon NMDA receptor efficacy (Liebetanz et al., 2002; Nitsche et al., 2003a,b). NMDA-dependent neuronal activity can facilitate cell survival and decrease apoptotic cell death (Catsicas et al., 1992; Sherrard and Bower, 1998; West et al., 2002) and refine synaptic circuitry (Katz and Shatz, 1996; Behar et al., 1999; Cohen-Cory, 2002). If CS increases functional activation of neurons in perilesion cortex, possibly via NMDA receptor activation, then it could be that greater neuronal density in the CSCath group is a result of CS inducing vulnerable neurons in the penumbra to be more functionally activated and effectively connected which, in turn, could contribute to neuronal survival. We have previously found that bipolar CS increases the surface density of dendrites in perilesion cortex (Adkins-Muir and Jones, 2003) and others have found that CS enlarges movement representation areas (motor maps; Kleim et al., 2003; Plautz et al., 2003). Motor map expansion has been associated with synaptogenesis (Kleim et al., 2004). CS also was found to increase the polysynaptic component of evoked motor cortical potentials (Teskey et al., 2003) that has been associated with LTPlike changes (Monfils and Teskey, 2004). Although there is evidence that CS enhances functional activation and dendritic structural plasticity, which may lead to neuronal survival, greater survival may not be a requirement to induce behavioral improvements or may only be a minor contributor. The CSAnod group performed significantly better than the NoCS group in the absence of significantly greater neuronal density. At this time, there is no clear explanation for why the CSCath group, but not the CSAnod, had greater neuronal density in perilesion cortex compared to NoCS. Although there were no significant differences between current levels necessary to induce movement thresholds nor in performance levels in the rehabilitative training task between the two CS groups, there may be subtle polarity-specific differences in mechanisms and/or the efficacy between cathodal and anodal stimulation. These findings also raise the possibility that an earlier onset of CS, corresponding to a greater cell death rate, may have even greater effects on the density of remaining neurons. An alternative interpretation of the reduced neuronal density is that cathodal CS causes a sublethal degeneration of neuronal or glial processes adjacent to the lesion such that the volume of tissue surrounding neuronal somata is reduced. That CS might create greater loss of neuronal processes seems unlikely because, as mentioned, CS has previously been found to increase the surface density of dendritic processes in cortex adjacent to the lesion (Adkins-Muir and Jones, 2003) and others have found increased movement representations (Kleim et al., 2003; Plautz et al., 2003) and evoked responses (Teskey et al., 2003) in this region following CS. Although it is not a very precise estimate of neuron number, there was a non-significant tendency for the CSCath group to have more overall remaining neurons than the NoCS group. The CSCath group also had greater interhemispheric cortical volume differences compared with the NoCS group (Table 1). However, this effect was not found in a subsequent study of animals receiving a similar regimen of cathodal CS treatment, suggesting that it is unlikely to be a consequence of the CS (Adkins et al., 2005). It seems more likely that the CSCath rats

368

D.L. Adkins et al. / Experimental Neurology 200 (2006) 356–370

were administered lesions which tended to be slightly larger. Future studies should directly estimate CS effects on apoptotic cell death and neuronal degeneration to more directly and sensitively assess its effects on neuronal survival. We also sought to determine if cathodal versus anodal current would be more effective in improving behavioral function. A previous study by Kleim et al. (2003) comparing 100 Hz cathodal and anodal CS concurrent with rehabilitation found that, early during training, cathodal CS induced greater reaching success compared to anodal CS, although at the end of 10 days of training there were no longer significant differences between the two groups (Kleim et al., 2003). Animal (Purpura and McMurtry, 1965; Nowak and Bullier, 1998) and human (e.g., Nitsche and Paulus, 2000, 2001; Nitsche et al., 2003a,b) studies have also shown differences in the electrophysiological and behavioral consequences between cathodal and anodal cortical stimulation using shorter durations of stimulation, higher amplitudes, and lower frequencies of current. In the present study, we did not find differences in performance of the reaching task between CSCath and CSAnod at any time-point and both polarities effectively enhanced performance compared to NoCS. The discrepancy in the effects of anodal CS found between the present study and Kleim et al. (2003) may be the result of methodological differences. Current in our study was delivered epidurally while Kleim et al. (2003) delivered current via subdurally implanted electrodes. Kleim et al. (2003) initiated CS and rehabilitation 3 days after cortical infarcts, whereas, in the present study, training was initiated at 10–14 days post-lesion, a time that corresponds to anodal CS-induced behavioral improvements in the study by Kleim et al. (2003). It is possible that anodal current delivered near infarcted tissue is more effective at later time points. Despite these differences, at the end of rehabilitation in both studies, 100 Hz anodal current induced greater motor recovery compared to unstimulated controls. Likewise, in vitro studies using diffuse electrical stimulation (Purpura and McMurtry, 1965; Nowak and Bullier, 1998) and in vivo human studies using transcranial direct current (e.g., Nitsche and Paulus, 2000, 2001; Nitsche et al., 2003a,b) have repeatedly demonstrated opposing effects between cathodal and anodal stimulation on membrane potentials and effects on the amplitude of motor evoked potentials. Although, in the present study, we used higher frequency and lower amplitudes of current over a longer period of time than used in previous in vitro and transcranial direct current studies, it cannot be ruled out that the two electrode polarities are differentially activating neurons as well. Despite these unanswered questions, this study and previous studies by Kleim et al. (2003) suggest that even if anodal stimulation is suboptimal, it has the capacity to improve function compared to no stimulation. The improvements in skilled reaching performance in CS versus NoCS rats did not generalize to another measure of sensorimotor behavior, the Schallert cylinder test. Other studies have indicated that post-injury improvements following rehabilitation may be task-specific. For example, amphetamine administration after brain injury has been shown to accelerate recovery (reviewed in Martinsson and Eksborg, 2004; Feeney, 1998). However, if rats with motor cortex damage are immobilized during amphetamine administration, they do not show

improved motor function compared to rats permitted to freely move in their home-cages following amphetamine treatment (Schmanke et al., 1996). In the present study, CS likely modulates neural activity while motor improvements are behaviorally driven through task-specific motor practice. It is not surprising, then, to find that cortical stimulation did not further enhance use of the impaired forelimb during upright exploratory behaviors, given that CS was never intentionally coupled with practice in these behaviors. This may indicate that CS needs to be coupled with task-specific practice in order to more effectively drive CS activated neurons which may increase the likelihood of improving function during repetitively performed tasks. In addition, the CSCath group had more prolonged asymmetries in forelimb postural support behavior than the NoCS group. It is possible that these data are a result of larger lesions in the CSCath group. However, there were no significant correlations between interhemispheric volume differences and the behavioral measures. It is also possible that greater reliance on the ipsilesional forelimb in the CSCath group induced slight experience-dependent increases in volume of the contralesional cortex (see Table 1), which exaggerated interhemispheric volume differences. In conclusion, this study indicates that epidural 100 Hz monopolar cathodal and anodal cortical stimulation combined with rehabilitation robustly enhances performance in skilled reaching after unilateral ischemic SMC lesions. Cathodal stimulation also increases neuronal density in perilesion motor cortex compared to unstimulated animals, but does not differentially affect the density of newly proliferated neurons or glia cells in ipsilesional cortex. Future studies are needed to further elucidate the mechanism of action and behavioral generalizability of CS combined with rehabilitation. Studies are currently being conducted to assess the duration of enhanced functional recovery following the combination of CS and rehabilitative training after the cessation of CS and daily training. Acknowledgments We thank Northstar Neuroscience for supplying the electrodes and for technical assistance. This study was supported by NIH Grants MH64586 and RR020700. References Adkins, D.L., Jones, T.A., 2005. d-Amphetamine enhances skilled reaching after ischemic cortical lesions in rats. Neurosci. Lett. 380, 214–218. Adkins, D.L., Voorhies, A.C., Jones, T.A., 2004. Behavioral and neuroplastic effects of focal endothelin-1 induced sensorimotor cortex lesions. Neuroscience 128, 473–486. Adkins, D.L., Hsu, E., Jones, T.A., 2005. The efficacy of cortical electrical stimulation combined with motor rehabilitation after cortical ischemia depends on the magnitude of impairment Program No. 548.9. 2005 Abstract Viewer/Itinerary Planner. Washington, DC: Society for Neuroscience, 2005. Online. Adkins-Muir, D.L., Jones, T.A., 2003. Cortical electrical stimulation combined with rehabilitative training: enhanced functional recovery and dendritic plasticity following focal cortical ischemia in rats. Neurol. Res. 25, 780–788.

D.L. Adkins et al. / Experimental Neurology 200 (2006) 356–370 Arvidsson, A., Collin, T., Kirik, D., Kokaia, Z., Lindvall, O., 2002. Neuronal replacement from endogenous precursors in the adult brain after stroke. Nat. Med. 8, 963–970. Behar, T.N., Scott, C.A., Greene, C.L., Wen, X., Smith, S.V., Maric, D., Liu, Q.Y., Colton, C.A., Barker, J.L., 1999. Glutamate acting at NMDA receptors stimulates embryonic cortical neuronal migration. J. Neurosci. 19, 4449–4461. Bengzon, J., Kokaia, Z., Elmer, E., Nanobashvili, A., Kokaia, M., Lindvall, O., 1997. Apoptosis and proliferation of dentate gyrus neurons after single and intermittent limbic seizures. Proc. Natl. Acad. Sci. U. S. A. 94 (19), 10432–10437. Biernaskie, J., Corbett, D., 2001. Enriched rehabilitative training promotes improved forelimb motor function and enhanced dendritic growth after focal ischemic injury. J. Neurosci. 21, 5272–5280. Bikson, M., Inoue, M., Akiyama, H., Deans, J.K., Fox, J.E., Miyakawa, H., Jefferys, J.G., 2004. Effects of uniform extracellular DC electric fields on excitability in rat hippocampal slices in vitro. J. Physiol. 557 (Pt 1), 175–190. Brown, J.A., Lutsep, H.L., Weinand, M., Cramer, S.C., 2006. Motor cortex stimulation for the enhancement of recovery from stroke: a prospective, multicenter safety study. Neurosurgery 58, 464–473. Brushart, T.M., Hoffman, P.N., Royall, R.M., Murinson, B.B., Witzel, C., Gordon, T., 2002. Electrical stimulation promotes motoneuron regeneration without increasing its speed or conditioning the neuron. J. Neurosci. 22 (15), 6631–6638. Bury, S.D., Jones, T.A., 2002. Unilateral sensorimotor cortex lesions in adult rats facilitate motor skill learning with the unaffected forelimb and traininginduced dendritic structural plasticity in the motor cortex. J. Neurosci. 22, 8597–8606. Canavero, S., Bonicalzi, V., Paolotti, R., Castellano, G., Greco-Crasto, S., Rizzo, L., Davini, O., Maina, R., 2003. Therapeutic extradural cortical stimulation for movement disorders: a review. Neurol. Res. 25 (2), 118–122. Review. Cantello, R., 2002. Applications of transcranial magnetic stimulation in movement disorders. J. Clin. Neurophysiol. 19 (4), 272–293. Review. Catsicas, M., Pequignot, Y., Clarke, P.G., 1992. Rapid onset of neuronal death induced by blockade of either axoplasmic transport or action potentials in afferent fibers during brain development. J. Neurosci. 12 (12), 4642–4650. Chu, C.J., Jones, T.A., 2000. Experience-dependent structural plasticity in cortex heterotopic to focal sensorimotor cortical damage. Exp. Neurol. 166, 403–414. Cohen-Cory, S., 2002. The developing synapse: construction and modulation of synaptic structures and circuits. Science 298, 770–776. Review. Feeney, D.M., 1998. Mechanisms of noradrenergic modulation of physical therapy: effects on functional recovery after cortical injury. In: Goldstein, L.B. (Ed.), Restorative Neurology: Advances in Pharmacotherapy for Recovery after Stroke. Futura, Armonk, NY, pp. 35–78. Franzini, A., Ferroli, P., Dones, I., Marras, C., Broggi, G., 2003. Chronic motor cortex stimulation for movement disorders: a promising perspective. Neurol. Res. 25, 123–126. Fregni, F., Pascual-Leone, A., 2005. Transcranial magnetic stimulation for the treatment of depression in neurologic disorders. Curr. Psychiatry Rep. 7 (5), 381–390. Fuxe, K., Bjelke, B., Andbjer, B., Grahn, H., Rimondini, R., Agnati, L.F., 1997. Endothelin-1 induced lesions of the frontoparietal cortex of the rat. A possible model of focal cortical ischemia. NeuroReport 8, 2623–2629. Gan, P., Cheng, J.S., Ng, Y.K., Ling, E.A., 2005. Role of GABA in electroacupuncture therapy on cerebral ischemia induced by occlusion of the middle cerebral artery in rats. Neurosci. Lett. 383 (3), 317–321. Glickstein, S.B., Ilch, C.P., Golanov, E.V., 2003. Electrical stimulation of the dorsal periaqueductal gray decreases volume of the brain infarction independently of accompanying hypertension and cerebrovasodilation. Brain Res. 994 (2), 135–145. Gundersen, H.J.G., 1986. Stereology of arbitrary particles. A review of unbiased number and size estimators and the presentation of some new ones, in memory of William R. Thompsen. J. Microsc. 143, 3–45. Gundersen, H.J., Bendtsen, T.F., Korbo, L., Marcussen, N., Mbller, A., Nielsen, K., Nyengaard, J.R., Pakkenberg, B., Sorensen, F.B., Vesterby, A., West, M.J., 1988. Some new, simple and efficient stereological methods and their use in pathological research and diagnosis. APMIS 96, 379–394. Harding, A.J., Halliday, G.M., Cullen, K., 1994. Practical considerations for the use of the optical disector in estimating neuronal number. J. Neurosci. Methods 51, 83–89.

369

Hummel, F., Celnik, P., Giraux, P., Floel, A., Wu, W.H., Gerloff, C., Cohen, L.G., 2005. Effects of non-invasive cortical stimulation on skilled motor function in chronic stroke. Brain 128 (Pt 3), 490–499. Jones, T.A., Chu, C.J., Grande, L.A., Gregory, A.D., 1999. Motor skills training enhances lesion-induced structural plasticity in the motor cortex of adult rats. J. Neurosci. 19, 10153–10163. Katz, L.C., Shatz, C.J., 1996. Synaptic activity and the construction of cortical circuits. Science 274 (5290), 1133–1138. Review. Kleim, J.A., Bruneau, R., VandenBerg, P., MacDonald, E., Mulrooney, R., Pocock, D., 2003. Motor cortex stimulation enhances motor recovery and reduces periinfarct dysfunction following ischemic insult. Neurol. Res. 25, 789–793. Kleim, J.A., Hogg, T.M., VandenBerg, P.M., Cooper, N.R., Bruneau, R., Remple, M., 2004. Cortical synaptogenesis and motor map reorganization occur during late, but not early, phase of motor skill learning. J. Neurosci. 24 (3), 628–633. Liebetanz, D., Nitsche, M.A., Tergau, F., Paulus, W., 2002. Pharmacological approach to the mechanisms of transcranial DC-stimulation-induced aftereffects of human motor cortex excitability. Brain 125 (Pt 10), 2238–2247. Macrae, I.M., Robinson, M.J., Graham, D.I., Reid, J.L., McCulloch, J., 1993. Endothelin-1-induced reductions in cerebral blood flow: dose dependency, time course, and neuropathological consequences. J. Cereb. Blood Flow Metab. 13, 276–284. Maesawa, S., Kaneoke, Y., Kajita, Y., Usui, N., Misawa, N., Nakayama, A., Yoshida, J., 2004. Long-term stimulation of the subthalamic nucleus in hemiparkinsonian rats: neuroprotection of dopaminergic neurons. J. Neurosurg. 100 (4), 679–687. Magavi, S.S., Leavitt, B.R., Macklis, J.D., 2000. Induction of neurogenesis in the neocortex of adult mice. Nature 405, 951–955. Martinsson, L., Eksborg, S., 2004. Drugs for stroke recovery: the example of amphetamines. Drugs Aging 21 (2), 67–79. Review. Mayhew, T.M., 1992. A review of recent advances in stereology for quantifying neural structure. J. Neurocytol. 21 (5), 313–328. Review. McKenna, J.E., Whishaw, I.Q., 1999. Complete compensation in skilled reaching success with associated impairments in limb synergies, after dorsal column lesion in the rat. J. Neurosci. 19, 1885–1894. Miklyaeva, E.I., Whishaw, I.Q., 1996. HemiParkinson analogue rats display active support in good limbs versus passive support in bad limbs on a skilled reaching task of variable height. Behav. Neurosci. 110, 117–125. Monfils, M.H., Teskey, G.C., 2004. Skilled-learning-induced potentiation in rat sensorimotor cortex: a transient form of behavioural long-term potentiation. Neuroscience 125 (2), 329–336. Nakagawa, E., Aimi, Y., Yasuhara, O., Tooyama, I., Shimada, M., McGeer, P.L., Kimura, H., 2000. Enhancement of progenitor cell division in the dentate gyrus triggered by initial limbic seizures in rat models of epilepsy. Epilepsia 41, 10–18. Nakatomi, H., Kuriu, T., Okabe, S., Yamamoto, S., Hatano, O., Kawahara, N., Tamura, A., Kirino, T., Nakafuku, M., 2002. Regeneration of hippocampal pyramidal neurons after ischemic brain injury by recruitment of endogenous neural progenitors. Cell 110, 429–441. Nitsche, M.A., Paulus, W., 2000. Excitability changes induced in the human motor cortex by weak transcranial direct current stimulation. J. Physiol. 527, 633–639. Nitsche, M.A., Paulus, W., 2001. Sustained excitability elevations induced by transcranial DC motor cortex stimulation in humans. Neurology 57, 1899–1901. Nitsche, M.A., Fricke, K., Henschke, U., Schlitterlau, A., Liebetanz, D., Lang, N., Henning, S., Tergau, F., Paulus, W., 2003a. Pharmacological modulation of cortical excitability shifts induced by transcranial direct current stimulation in humans. J. Physiol. 553 (Pt 1), 293–301. Nitsche, M.A., Schauenburg, A., Lang, N., Liebetanz, D., Exner, C., Paulus, W., Tergau, F., 2003b. Facilitation of implicit motor learning by weak transcranial direct current stimulation of the primary motor cortex in the human. J. Cogn. Neurosci. 15 (4), 619–626. Nowak, L.G., Bullier, J., 1998. Axons, but not cell bodies, are activated by electrical stimulation in cortical gray matter: I. Evidence from chronaxie measurements. Exp. Brain Res. 118, 477–488. Nudo, R.J., Wise, B.M., SiFuentes, F., Milliken, G.W., 1996. Neural substrates for the effects of rehabilitative training on motor recovery after ischemic infarct. Science 272, 1791–1794.

370

D.L. Adkins et al. / Experimental Neurology 200 (2006) 356–370

Parent, J.M., 2003. Injury-induced neurogenesis in the adult mammalian brain. Neuroscientist 9 (4), 261–272. Review. Parent, J.M., Janumpalli, S., McNamara, J.O., Lowenstein, D.H., 1998. Increased dentate granule cell neurogenesis following amygdala kindling in the adult rat. Neurosci. Lett. 247 (1), 9–12. Parent, J.M., Vexler, Z.S., Gong, C., Derugin, N., Ferriero, D.M., 2002. Rat forebrain neurogenesis and striatal neuron replacement after focal stroke. Ann. Neurol. 52, 802–813. Pascual-Leone, A., Peris, M., Tormos, J.M., Pascual, A.P., Catala, M.D., 1996. Reorganization of human cortical motor output maps following traumatic forearm amputation. NeuroReport 7 (13), 2068–2070. Paxinos, G., Watson, C., 1986. The Rat Brain in Stereotaxic Coordinates, 2nd ed. Academic Press, Sydney. Peterson, G.M., Devine, J.V., 1963. Transfer on handedness in the rat resulting from small cortical lesions after limited forced practice. J. Comp. Physiol. Psychol. 56, 752–756. Plautz, E.J., Barbay, S., Frost, S.B., Friel, K.M., Dancause, N., Zoubina, E.V., Stowe, A.M., Quaney, B.M., Nudo, R.J., 2003. Post-infarct cortical plasticity and behavioral recovery using concurrent cortical stimulation and rehabilitative training: a feasibility study in primates. Neurol. Res. 8, 801–810. Purpura, D.P., McMurtry, J.G., 1965. Intracellular activities and evoked potential changes during polarization of motor cortex. J. Neurophysiol. 28, 166–185. Schallert, T., Fleming, S.M., Leasure, J.L., Tillerson, J.L., Bland, S.T., 2000. CNS plasticity and assessment of forelimb sensorimotor outcome in uni-

lateral rat models of stroke, cortical ablation, parkinsonism and spinal cord injury. Neuropharmacology 39, 777–787. Schmanke, T.D., Avery, R.A., Barth, T.M., 1996. The effects of amphetamine on recovery of function after cortical damage in the rat depend on the behavioral requirements of the task. J. Neurotrauma 13, 293–307. Sherrard, R.M., Bower, A.J., 1998. Role of afferents in the development and cell survival of the vertebrate nervous system. Clin. Exp. Pharmacol. Physiol. 25 (7–8), 487–495. Review. Schmued, L.C., Hopkins, K.J., 2000. Fluoro-Jade B: a high affinity fluorescent marker for the localization of neuronal degeneration. Brain Res. 874 (2), 123–130. Scott, B.W., Wang, S., Burnham, W.M., De Boni, U., Wojtowicz, J.M., 1998. Kindling-induced neurogenesis in the dentate gyrus of the rat. Neurosci. Lett. 248 (2), 73–76. Teskey, G.C., Flynn, C., Goertzen, C.D., Monfils, M.H., Young, N.A., 2003. Cortical stimulation improves skilled forelimb use following a focal ischemic infarct in the rat. Neurol. Res. 25, 794–800. West, A.E., Griffith, E.C., Greenberg, M.E., 2002. Regulation of transcription factors by neuronal activity. Nat. Rev., Neurosci. 3 (12), 921–931. Review. Withers, G.S., Greenough, W.T., 1989. Reach training selectively alters dendritic branching in subpopulations of layer II–III pyramids in rat motorsomatosensory forelimb cortex. Neuropsychologia 27, 61–69.