The subjects of these experiments were aged male Sprague-Dawley rats (N = 80; 18–20 months of age; 520–600 g) kept under standard laboratory conditions with free access to food and water. The numbers reported in the results refer to the number of animals that survived the surgery and completed the 8-week testing period. All experiments were approved by the Animal Experimentation Ethics Board of the State of Mecklenburg-Vorpommern as meeting the ethical requirements of the German National Act on the Use of Experimental Animals (approval no LALLF M-V/TSD/7221.3-1.1-040/10) and by the Institutional Animal Care and Use Committee of the Medical University of Craiova.

A scientist was in charge of randomization by (i) group assignment, (ii) surgery assignment; and (iii) treatment assignment.

To evaluate changes in neurological function associated with ischemia, the rats were subjected to a variety of somatosensory, motor, learning, and memory tests before and after surgery. All testing was performed from 9 to 11 a.m. Results obtained before surgery were used to define 100% functionality for each animal on each test, and functional recovery was expressed as percent recovery relative to the pre-surgery baseline.

The rotating pole task assesses coordination and sensorimotor function in the MCAO model. Each rat was tested for its ability to cross a rotating (6 rpm) horizontal rod. The score assessment was done as previously described by our group (Buchhold et al., 2007). Briefly, the time taken for the rat to traverse the rotating cylinder and join a group of rats visible at the finish line was measured. The score assessment was twofold: (i) time (seconds) required to traverse the rotating cylinder and (ii) the score as follows: 0 – rat falls immediately (onto a soft surface); 1 – rat does not walk forward, but stays on the rotarod; 2 – rat walks, but falls before reaching the goal; 3 – rat traverses the rod successfully, but the limbs are used asymmetrically; 4 – the left hindlimb is used less than 50% of the time taken to traverse the rod; 5 – the rat successfully traverses the rod, but with some difficulties; 6 – no mistakes, symmetric movements.

We assessed the asymmetry of sensorimotor deficit of the forelimbs induced by unilateral MCAO by the adhesive tape removal test. In short, sticky patches were applied on the distal hairless parts of the forelimbs and the removal time from both limbs was measured. Three trials were done separately for each limb and the means of the values were noted. If the animal did not remove the tape within 180 s, the timer was stopped. Results are given as time needed to remove the adhesive tape from one forelimb divided by the sum of time needed to remove it from both forelimbs (Popa-Wagner et al., 2010).

The Morris water maze task was used to assess spatial learning and memory. One week before surgery, aged rats were trained to find a submerged platform in a large (180 cm-diameter) pool filled to within 20 cm of the upper edge with water maintained at 26°C. The pool was divided into four compass quadrants (north, south, east, and west). Several visual stimuli were placed in each of the four quadrants. For the acquisition of spatial learning, each animal underwent a block of four trials per day for 7 days. Before the first trial, the rat was placed on the hidden platform for 30 s by the investigator. Each trial consisted of placing the rat in the water at one of the randomly selected four starting locations around the pool perimeter. Each rat was allowed a maximum of 60 s to find the hidden platform and remain on it for 30 s. If a rat failed to find the platform within 60 s, the rat was placed on the platform for 30 s by the investigator. The time and distance required to find the hidden platform during these four acquisition trials were averaged (Tottori et al., 2002). The swim path was recorded by an image analysis system (VideoMot2, TSE, Bad Homburg, Germany) that computed path length and percentage of time spent in each of the four quadrants. Functional restoration of spatial learning and memory was estimated by weekly testing after MCAO and in total for 8 weeks.

Bone marrow was collected from 40 Sprague-Dawley (5 weeks old; Harlan-Winkelmann, Borchen, Germany). Both femurs were aseptically removed and placed in alpha-MEM (Biochrom, Germany) supplemented with 20% fetal bovine serum (FBS, Biochrom, Germany), 1% penicillin/streptomycin (Biochrom, Germany), and 2 mM HEPES (Biochrom, Germany), to which we refer as standard medium. The bone marrow was flushed out with 5 ml standard medium using a 10 ml-syringe and a 23-gage needle and the BM MSC was isolated and cryopreserved as previously described (Sauerzweig et al., 2009). The cells were counted using a Neubauer-hemocytometer and cultured with a density of 2 × 10⁶ whole marrow cells per square centimeter in standard medium with 37°C, 5% CO₂, and 92% relative humidity (RH). After 3 days, non-adherent cells were removed by flushing with 10 ml PBS 0.1 M and the standard medium was replaced. The attached cells were cultured until 80% of confluence was reached. After trypsinization, cells were dissolved in standard medium supplemented with 10% DMSO (Sigma-Aldrich, Germany) and apportioned for cryo-preservation. Immunophenotyping of rBM-MSCs was performed with antibodies against rat antigens CD29, CD45, and CD90.

Mesenchymal stem cells (MSCs) harvested and cultured from normal human bone marrow were purchased from Lonza Research (Cologne, Germany) and cultivated according to manufacturer’s protocol under the same conditions as rat cells. The phenotypes of hBM MSCs used in this study were positive for CD105 and CD166. Cells tested negative for CD14, CD34, and CD45. Cells were tested for purity by flow cytometry and for their ability to differentiate into osteogenic, chondrogenic, and adipogenic lineages by Lonza.

Eighteen hours prior to surgery, rats were deprived of food to minimize variability in ischemic damage that can result from varying plasma glucose levels. Water remained available at all times. In all cases, surgery was performed between 9:00 and 13:00 hours.

Cerebral infarction was induced by transcranial interruption of blood flow by transiently lifting the middle cerebral artery (MCA) with a tungsten hook as previously described (Popa-Wagner et al., 2010). Throughout surgery, anesthesia was maintained by spontaneous inhalation of 1–1.5% halothane in a mixture of 75% nitrous oxide and 25% oxygen. Body temperature was maintained at 37°C by a Homeothermic Blanket System (Harvard Apparatus). The local changes in blood flow were monitored using a laser Doppler device (Perimed, Stockholm, Sweden). A decrease in the laser Doppler signal to <20% of control values was considered to indicate successful MCA occlusion. After 90 min, the tungsten hook was released and the common carotid arteries were re-opened. Surgery was performed under antiseptic conditions to minimize the risk of infection. Subsequent to survival times of 56 days, the rats were deeply anesthetized with 2.5% halothane in 75% nitrous oxide and 25% oxygen, and perfused with neutral buffered saline followed by buffered 4% freshly depolymerized paraformaldehyde. The brain was removed, post-fixed in 4% buffered paraformaldehyde for 24 h, cryoprotected in 15% glycerol prepared in 10 mmol/l phosphate buffered saline, flash-frozen in isopentane, and stored at −70°C until sectioning.

Animals were randomly assigned to three groups as follows: (i) G-CSF group (N = 20) received daily injections of 50 μg/kg BW and in total for 28 days; (ii) G-CSF + BM MSCs (N = 20) group received daily injections of G-CSF at 50 μg/kg BW and in total for 28 days and a single dose of BM MSCs (1 × 10⁶/kg BW) given intravenously; and (iii) control group (N = 20) was given daily the vehicle (5% glucose) for 28 days. Combination treatment was done at 6 h post-stroke (Figure 1, upper panel). To investigate the localization of injected cells, a separate group of aged animals was injected with mesenchymal cells of human origin (hBM-MSCs). Although hBM-MSCs are poorly immunogenic (Chamberlain et al., 2007) and the rats survived just 4 days after administration, to prevent graft rejection the animals were given cyclosporine A (s.c., Sandimmun, Novartis, 10 mg/kg) diluted in Chremophor EL (Sigma).

To label newly generated cells, rats were given bromodeoxyuridine (BrdU; 50 mg/kg body weight, i.p.; Sigma), daily in the first week post-stroke, and every other day in the following weeks post-stroke, for a total period of 4 weeks.

Magnetic resonance imaging (MRI) was used to visualize the infarct volume for both groups at day 3 and at day 48 after stroke. MRI measurements were performed on a 7-T Bruker ClinScan magnet with a 20 cm inner bore, capable of 290 mT/m in 250 μs (Bruker BioSpin MRI, Ettlingen, Germany). Images were received by a 2 × 2 phased array RF coil, designed specifically for rat brain studies that was placed directly on the skull. The animals were anesthetized during imaging to minimize discomfort. Respiratory rate was monitored, and isoflurane concentrations were varied between 1.5 and 2.0% to keep the respiratory rate between 35 and 45/min. After positioning the animal’s head, quantitative T2 measurements were performed with a multislice spin-echo sequence with 25 slices of 0.7 mm thickness and a matrix size 640 × 640 pixels, a repetition time (TR) of 4330 ms, and an echo time (TE) of 45 ms.

T2WI lesion volumes were determined using the image processing software Medical Image Processing, Analysis and Visualization (MIPAV, version 3.0, National Institutes of Health, Bethesda, MD, USA). After optimal adjustment of contrast, the edge of the lesion was traced manually on each of the 25 coronal slices, which completely covered the MCA territory in all animals. The areas of hyperintensity were then summed and multiplied by the slice thickness to calculate lesion volumes.

To assess the size of the infarct induced by focal ischemia, we introduced several years ago NeuN immunostaining (Badan et al., 2003). Every 20th free-floating section of 25 μm was immunostained for NeuN to cover the entire infarcted volume, which was then calculated as the sum of the partial areas using the MicroBrightField (Colchester, VT, USA) system. Integration of the resulting partial volumes (partial areas × no. of sections × section thickness × section intervals) yielded the total volume of the ipsilateral hemisphere along with the total volume of the infarct.

Cryostat, free-floating sections of 25 μm were fixed in 4% paraformaldehyde for 15 min and then washed extensively with PBS and processed using an automatic staining machine for floating sections Tingomat 501¹. After incubation in 50% formamide/2× SSC for 2 h at 60°C, sections were washed again, first in 2× SSC and then in 10× PBS. After denaturization in 2 N HCL at 37°C for 40 min, sections were made neutral by adding 0.1 M borate buffer (pH 8.5). Thereafter, sections were incubated with guinea pig anti-doublecortin (DCX; Millipore, Germany) overnight at 4°C followed by secondary biotinylated antibodies (Dianova, Hamburg, Germany) and visualized with streptavidin Alexa 488 (Life Technologies, Karslruhe, Germany). Finally, sections were incubated with rat anti-BrdU antibody (1:2000, AbD Serotec, Puchheim, Germany). BrdU-positive cells were visualized by incubating with Cy3-conjugated donkey anti-rat IgG (H + L) (1:3000). For phenotyping in the periinfarcted area and beyond the fibrotic scar, sections were triple-immunolabeled with rabbit anti-laminin-specific antibodies (1:5000, Sigma, Munich), mouse anti-NeuN antibodies (1:500, Millipore, Germany), mouse anti-RECA (1:1000, Millipore, Germany), and rat anti-BrdU-specific antibodies (1:3000; Serotec, UK). The antigen–antibody complexes were visualized with donkey anti-mouse Cy3-conjugated antibodies (1:2000), donkey anti-rabbit Cy2-conjugated antibodies (1:3000), and donkey anti-rat Cy5-conjugated antibodies (1:2000), respectively.

For phenotyping of transplanted cells, the tissue was incubated with a rabbit anti-mouse NeuN (1:1000, Novus Biologicals, UK) or mouse anti-human CD166 (1:1000, antibodies-on-line, Aachen, Germany) or mouse anti-human nuclei (ABIN361360 antibodies-on-line, Aachen, Germany) or mouse anti-human CD105 antibodies at 4°C overnight. At the next day, sections were rinsed with PBS and incubated with Alexa Fluor^® 568 goat anti-rabbit IgG or Alexa Fluor^® 594 goat anti-mouse IgG for CD105. Finally, the signal was amplified utilizing an anti-mouse polymer-based secondary detection system (Histofine polymer-HRP, Nichirei, Japan) diluted 1:10 in PBS containing 1% normal goat serum and 0.3% Tween 20. After washing in PBS, sections were stained with tyramide-FITC. After final rinsing, sections were brought to Superfrost Plus slides and mounted using PVA/DABCO-containing medium. To visualize nuclei, some sections were counterstained with DAPI.

Microvascular density was quantitated using the “hot spot” analysis covering 30% of the infarcted area. Briefly, hot-spots, i.e., regions with a high density of CD105-positive microvessels in humans and CD31-capillaries in rats, were identified using a 40× objective and were then counted using 20× objective, corresponding to a microscopic field of 0.7386 mm². Counting was done by two independent observers and the results are expressed as means ± SD (Buga et al., 2014).

For light microscopy, a Nikon Eclipse (Nikon, Duesseldorf, Germany) was used. Confocal microscopy images were acquired using a Zeiss LSM710 laser-scanning confocal system with spectral detection capabilities, and Zen 2010 software version 6.0 (Carl Zeiss Microscopy GmbH, Jena, Germany) was used for image acquisition and analysis. Excitation light was provided by 488, 543, and 634 nm laser lines; fluorescence emission was detected at 500–530 nm for FITC (green), 550–600 nm for rhodamin (red) and 650–710 nm for Cy5 (blue) in separate tracks, using a confocal aperture of 1 Airy unit. Some of the images were acquired as z-stacks and 3D reconstruction was performed.

The main effects of treatment and time as well as interactions of the two factors, on recovery, were evaluated using a mixed-model analysis of variance (MANOVA) for each measure, with treatment as between-subjects variables and time as a within-subject variable. The non-parametric data analysis was conducted using the Kruskal–Wallis test that is designed for multiple independent measures followed by a Bonferroni correction for alpha errors (SPSS Inc., Chicago, IL, USA). A p value of <0.05 was considered to be statistically significant. Water maze analysis was done using the Wilcoxon rank sum test.

To facilitate feeding in aged animals during the first 3 days post-stroke, we fed them with moistened, soft pellets. Following infarction, all rats had diminished performance on the first 3 days post-surgery, which was at least partially attributable to the surgical procedure itself. The mortality rate was almost identical, 10% for each group. All the above mentioned animals died between day 3 and day 10 post-stroke.

The adhesive tape removal test probes for differences between forelimbs in cutaneous sensitivity and sensorimotor integration after stroke. As compared to pre-operative, trained animals, rMCAo animals demonstrated a marked difference in post-operative performance for the left (affected) forelimb. By day 3, post-stroke animals started recuperation and reached significant recovery of function by day 14 in the G-CSF group only (Figure 1B, filled red squares). The combination between G-CSF and BM MSC showed no additional or showed a similar although non-significant trend improvement of recuperation of sensorimotor function (Figure 1C, filled green squares).

After an abrupt decline in performance on the rotarod on day 3 post-stroke, control rats began improvement and recovered to 47% by day 56 (Figure 1D, filled black circles). Of the treated groups, the best recovery was shown by group treated with G-CSF alone that reached 72% of the pre-surgery value (baseline) (Figure 1D, red squares) followed by the group treated with G-CSF + BM MSC that reached 58% of the pre-stroke value (Figure 1E, blue squares).

Representative swim paths are shown in Figures 2A–C and included the start of training (−7d), the pre-surgery path pattern (0d), first testing after stroke (+7d), and the final testing (+56d). Over the pre-stroke training period of 7 days, rats learned to locate and climb onto the hidden platform and performance improved significantly during this time. In all groups, the path became shorter as the training sessions progressed (Figures 2A–C). Because of the skull injury, we avoided testing the animals in the first week post-stroke.

As previously shown, aged rats need more time to recover behaviorally after stroke than young animals (Lindner, 1997; Tottori et al., 2002; Lindner et al., 2003; Buchhold et al., 2007). Consequently, the path length required to reach the platform in the third quadrant reached a maximum by day 7 post-stroke. After 7 days, the animals began recovering in this test. The best recovery was seen for the G-CSF group that showed significant improvement of spatial reference-memory between days 21 and 49 in the second quadrant (Figure 2D). However, in the third quadrant, the performance was temporarily improved between days 14 and 28 in the group treated with G-CSF + BM MSC as compared to the control group (Figure 2E).

Representative MRI data for the two rats closest to the mean for each group are shown in Figures 3A–F. Cortical lesion and brain edema at day 3 post-stroke, as defined by the region of T2 hyperintensity, was not significantly reduced by any treatment (Figures 3A–C). The G-CSF group showed, however, less edema as compared with the combination treatments or controls (Figure 3J, red bar). The second MRI done at day 48 post-stroke revealed much smaller infarcts suggesting that brain edema has had a large contribution to the T2 hyperintensity in aged animals (Figures 3D–F).

Immunohistochemical staining of the infarct area at day 56 using an anti-NeuN antibody (Figures 3G–I) showed that the infarct volumes were largely similar to those measured by MRI at day 48, indicating that the infarct volume had stabilized by day 48. Consequently, the infarct volume as measured by MRI (Figure 3K) or immunohistochemistry (Figure 3L) was not significantly different between the groups at day 48 post-stroke.

In the ipsilateral hemisphere, the injected human BM MSCs were detected in the corpus callosum as shown for CD166-positive cells (Figure 4A, arrows) and CD105-positive cells (Figure 4F, arrows). In our model the cell probably entered the injured brain via the lateral ventricle as shown by the CD166-positive cells (Figure 4B). A fraction (about 1%) of the injected CD166- and CD105-positive cells reached the infarcted area (Figures 4C,E, arrows) where they were intermingled with surviving or degenerating neuronal nuclei (Figure 4C, arrowheads). Noteworthy was also the presence of immunopositivity for human nuclei (Figure 4D, arrows) that were dispersed between the rat nuclei in the infarcted area (Figure 4D, arrowheads).

Next, we investigated the presence of the early neuronal marker doublecortin by immunofluorescence in the lateral ventricle region. To this end, the proliferating cells were labeled by injecting animals with BrdU. To our surprise, at day 56 post-stroke none of the DCX⁺ cells in the SVZ of control animals co-localized with BrdU-labeled nuclei. Instead, the BrdU-positive nuclei were distributed mainly in the “pinwheel” architecture of the ventricular epithelium (Liebner et al., 2008; Gajera et al., 2010) (Figure 5A). The DCX⁺ cells occupied an adjacent, distinct position (Figure 5A, arrows). Some of the DCX⁺ migrated away from the ventricular wall (Figure 5B). In agreement with the previous results, we noted vigorous neurogenesis with many DCX⁺ (arrows) co-localizing with BrdU nuclei in the G-CSF-treated animals (Figure 5C; arrowheads) and animals treated with G-CSF + BM MSC (Figure 5D, arrows).

In regions adjacent to the infarct scar, we found numerous BrdU⁺ nuclei in the endothelium of newly formed blood vessels in the formerly infarct core (Figure 5E, green). The border to the healthy brain region was abruptly demarcated to the left by NeuN-positive nuclei (Figure 5E, NeuN shown in red). Beyond the formerly infarct core, we noted vigorous sprouting angiogenesis as evidenced by RECA/BrdU double positive blood vessels (Figure 5F, violet) as well as numerous BrdU⁺ nuclei in the newly formed endothelium (Figure 5F, blue) and reconstruction of the basal lamina (Figure 5F, green) during the resolution phase of angiogenesis. By number of laminin/BrdU co-localizations, the density of the newly formed blood vessels was significantly higher (threefold, p = 0.01) in the brains of animals treated with the combination G-CSF + BM MSC as compared to controls and G-CSF alone (Figure 5G).