Permanent occlusion of distal branch of middle cerebral artery was performed using an electrocoagulation methodology described previously (Tamura et al., 1981; Xing et al., 2012). In brief, rats were placed in the anesthesia induction box supplied with 3–4% isoflurane at 3 L/min in 100% oxygen. Anesthesia was maintained with 1.5–2.5% isoflurane at 800 mL/min in 100% oxygen, delivered through a nose mask (SurgiVet, Waukesha, WI, United States) during the surgical procedure. The distal striatal branch of MCA was exposed and occluded by unipolar electrocoagulation under an operating microscope. Rectal temperature of the animals was monitored and maintained at approximately 37°C throughout the procedure. Sham-operated animals were performed with the same surgical procedures except for the electrocoagulation of dMCAO. After surgery, the rats were allowed to wake up and evaluated the neurological status as described by previously (Rodriguez-Grande et al., 2013). Rats neither with neurologic deficit nor cortical infarction were excluded from this study.

Animals were intracardially perfused with normal saline, followed by 4% paraformaldehyde (PFA) in phosphate buffer saline (PBS) (0.01 M, pH 7.4) under anesthesia with 10% chloral hydrated administered intraperitoneally (350 mg/Kg). The brains were pos-fixed for 12 h in 4% PFA and then cryoprotected with 10, 20, 30% sucrose in the same fixative overnight. Coronal tissue blocks (bregma 1.7 mm to −5.8 mm) were cut on a freezing microtome (Leica CM1950, Heidelberg, Germany) into 30 μm-thick sections. According to the standard procedure, Nissl staining was performed with 0.1% cresyl violet (MilliporeSigma, Burlington, MA, United States), and then sections were dehydrated with 90 and 100% ethanol and immersed into dimethylbenzene.

Nissl staining was used for infarct volume evaluation. Relative infarct volume was expressed as the percentage of the contralateral hemisphere (Swanson et al., 1990). Briefly, the coronal brain sections from bregma level of 1.0 mm to −1.0 mm were analyzed. Relative infarct volumes at 28 days after dMCAO were evaluated from the contralateral hemisphere volume (Vc) and ipsilateral nonischemic hemisphere volume (Vi) according to the equation (Vc-Vi)/Vc × 100%.

The rats were divided into four groups, sham-operated, dMCAO, vehicle and BMSCs groups for immunohistochemical experiments. Rats were euthanized at 7, 14, and 28 days after dMCAO, respectively (n = 7 in each group). Single-label immunohistochemistry was conducted by the avidin-biotin complex (ABC) peroxidase method (Zhan et al., 2010). Briefly, sections from bregma −4.52 to −6.3 mm were rinsed with 0.01 M PBS, treated with 3% H₂O₂ for 30 min, followed by 5% normal serum for 1 h at room temperature, and then incubated overnight at 4°C with primary antibodies, including rabbit polyclonal anti-nuclear-associated antigen Ki-67 (Ki-67) (1:2000, Abcam, Cat# ab15580, RRID:AB_443209), rabbit polyclonal anti-doublecortin (DCX) (1:9000, Abcam, Cat# ab18723, RRID:AB_732011), rabbit polyclonal anti-tyrosine hydroxylase (TH) (1:3000, MilliporeSigma, Cat# AB152, RRID:AB_390204), mouse monoclonal antibody anti-neuron-specific nuclear-binding protein (NeuN) (1:2000; MilliporeSigma, Cat# MAB377, RRID:AB_2298772) and rabbit polyclonal anti-glial fibrillary acidic protein (GFAP) (1:2000, MilliporeSigma, Cat# AB5804, RRID:AB_11212369). After washing three times with 0.01 M PBS, the sections were incubated with biotinylated secondary IgG antibody for 2 h at room temperature. Following washing with PBS, the sections were incubated with the ABC peroxidase for 30 min at room temperature. The peroxidase reaction was visualized with 0.05% diaminobenzidine and 0.01% hydrogen peroxide. Immunopositive cells in ipsilateral and contralateral substantia nigra compact (SNc) part were quantified in three sections of each animal. The number of immunoreactive cells for TH, NeuN, GFAP, Ki-67, and DCX was counted, respectively. Only cells with reaction products that presented within a clear border were quantified from three non-overlapping fields under a light microscope with ×200 magnification and presented as the average cell number per field on each section using ImageJ software (NIH, Bethesda, MD, United States). Data were assessed in a double-blind procedure.

BMSCs were isolated and cultured as previously described (Shen et al., 2006). Healthy specific pathogen free (SPF) grade, 3 weeks old male SD rats weighing 50–100 g were soaked in 70% ethanol for 3 min, sacrificed by cervical dislocation. Fresh bone marrow was harvested aseptically from femurs and tibias by inserting a syringe fitted with 18-gauge needle into the shaft of the bone and flushing bone marrow out with a low glucose Dulbecco’s Modified Eagle’s Medium [LG-DMEM (Gibco, Carlsbad, CA, United States)]. Then the bone marrow was mechanically dissociated by pipetting repeatedly. Next, the cells were separated from the bone marrow by a centrifuge with 1300 rpm for 3 min and incubated with complete growth medium [LG-DMEM (Gibco) containing 10% FBS (Gibco) and 1% penicillin/streptomycin (HyClone)] at 37°C in 5% CO₂ in a sterile petri dish. After 3 days, nonadherent cells were removed by replacing the medium every three days. The adherent cells after 3–5 passages were prepared for transplantation.

BMSCs were analyzed for the expression of a panel of antigens. Flow cytometry analysis was performed with BMSCs at five passage of culture. Cells were incubated with fluorescent isothiocyanate (FITC)-conjugated mice anti-human CD29, CD45, and phycoerythrin (PE)-conjugated mice anti-human CD34, CD44, and CD90 (1 × 10⁶ cells/100 μL + 20 μL antibody, Becton Dickinson, Pharmingen) for phenotypic characterization. As negative controls, cells were stained with an isotype control antibody. After incubation with the antibody for 25 min, cells were washed with PBS twice and re-suspended in PBS and analyzed using flow cytometer (MoFlo XDP, Beckman Coulter, United States).

The surface antigen expression of BMSCs was identified as CD29⁺, CD44⁺, and CD90⁺, CD34^– and CD45^– (Supplementary Figure 1). The lack of CD34 and CD45 expression suggested that the cell population was depleted of hematopoietic stem cells during sub cultivation by plastic adherence. Thus, the cells used in this investigation were regarded as BMSCs.

At 24 h after dMCAO, rats were randomly selected for BMSCs transplantation. After anesthetizing with 3–4% isoflurane in 100% oxygen, rats were injected slowly with approximately 4 × 10⁶ BMSCs in 1 mL of PBS (0.01 M, pH 7.4) or PBS only as control via the tail vein over a 5 min period.

BMSCs were incubated with 10 μM DiR-PBS (PH7.4) for 30 min at 37°C according to the protocol of Xeno-Light 1, 1-dioctadecyltetramethyl indotricarbocyanine iodide (DiR) (Caliper Lifesciences). At 1 day after dMCAO, rats were injected slowly with approximately 4 × 10⁶ BMSCs-DiR⁺ (DiR-labeled) or BMSCs-DiR^– (DiR-nonlabeled) in 1 mL of PBS via the tail vein under anesthesia with 3–4% isoflurane in 100% oxygen. At 2 h, and 1, 7, 14 days after transplantation, the rats were completely anesthetized with chloral hydrate, in vivo their fluorescence imaging were monitored to assess bio-distribution of DiR⁺ or DiR^– labeled BMSCs by Xenogen IVIS Spectrum imaging system (PerkinElmer, MA, United States) under 745 nm of excitation and 800 nm of emission. Coronal brain slices, 2 mm thick, were obtained using a brain slicer for ex vivo imaging. Fluorescent images of each sample including cortex (−0.4 mm from bregma) and SN (−5.3 mm from bregma) were analyzed and overlaid using Living Image software (Xenogen, Alameda, CA, United States). DiR fluorescence intensity which was presented as an average radiant efficiency was plotted in units of the maximum number of photons per second per centimeter square per steradian (p/s/cm²/sr) (Ruan et al., 2012).

The behavioral tests were performed at 1, 4, 8, 12, 16, 20, 24, and 28 days after dMCAO by an investigator who was blinded to the experimental groups. All rats were familiarized with the testing environment before surgery. In the adhesive removal test, a piece of adhesive paper sized 200 mm² was placed onto the contralesional forepaw (Schallert and Whishaw, 1984; Bouet et al., 2009). The rat performance was assessed by measuring the time needed sense and to remove the adhesives. The time to remove each stimulus from forelimbs was recorded on 3 trials/day. Individual trial was separated by at least 3 min. Before surgical procedure, the animals were trained 1 trial/d for 5 days. Once the rats were able to remove the tape within 2 min, they were subjected to dMCAO. In addition, neurological function was assessed by modified neurological severity scores (NSS) which is a composite of motor, sensory, reflex, and balance tests graded on a scale of 0–18 (normal score, 0; maximal deficit score, 18) (Chen et al., 2001).

In this study, retrograde tracing were employed to label the neurons between the infarct cortex and the ipsilateral SN. The attenuated (Bartha) strain of PRV-152, which was constructed to express EGFP (a gift from Enquist L. W., Princeton University), was used. PRV-152 viruses were grown in pig kidney (PK15) cells and stored at -80°C. The final titers deters determined in PK15 cells were 1 × 10^8–9 PFU for PRV-152 (Xu et al., 2006). Under anesthesia with 3–4% isoflurane, 3 μL PRV-152 was injected to the ipsilateral SNr of rats (5.4 mm posterior to the bregma, 2.1 mm lateral to the midline, and 8.3 mm below the dura) stereotactically at 7 days of dMCAO. The needle was left in place for an additional 30 min before it was removed. A fresh stock of virus was thawed for each injection.

Triple-fluorescent immunohistochemistry was performed as previously described (Zuo et al., 2018). Sections were pre-incubated with 5% normal goat serum (containing 0.2% Triton X-100) for 1 h at room temperature, and then incubated overnight at 4°C with mixtures of primary antibodies: rabbit anti-GFP (1:800, MilliporeSigma, Cat# MAB3580, RRID:AB_2313783), rabbit polyclonal anti-Ki-67 (1:1000, Abcam, Cat# ab15580, RRID:AB_443209), rabbit polyclonal anti-DCX (1:5000, Abcam, Cat# ab18723, RRID:AB_732011) and mouse monoclonal antibody anti-NeuN (1:1000; MilliporeSigma, Cat# MAB377, RRID:AB_2298772). After rinsing in 0.01 M PBS, the sections were incubated for 1 h at room temperature with the following secondary antibodies: Cy3-conjugated goat anti-mouse IgG antibody (1:100; MilliporeSigma, Cat# AP124C, RRID:AB_11213281) and FITC-conjugated goat anti-rabbit antibody (1:100; MilliporeSigma, Cat# AP307F, RRID:AB_92652). Then, sections were PBS-washed and mounted with mounting medium containing 4′, 6-diamidino-2-phenylindole (DAPI). Slides were analyzed with a confocal laser microscope (Leica Microsystems, Wetzlar, Hessen, Germany).

Rats were anesthetized with 10% chloral hydrate (3.5 mg/kg, i.p.) and perfused transcardially with normal saline. According to previous study (Jackson-Lewis and Przedborski, 2007), the striata were removed quickly (within 30 s) from rats after dMCAO with or without BMSCs transplantation and placed immediately into ice-cold saline before being stored at -80°C for high performance liquid chromatography-tandem mass spectrometric (HPLC-MS/MS) analysis.

According to the protocol described by Man et al. (2019), the concentrations of dopamine (DA), 3, 4-dihydroxyphenylacetic acid (DOPAC), and homovanillic acid (HVA) in the striata were analyzed by HPLC-MS/MS (1290-6460C, Agilent Technologies, CA, United States) in National Analytical Center of China (NACC), Guangzhou. Briefly, the striata were extracted in 2 mL of 100% methanol by homogenizer and centrifuged at 4500 g for 5 min at 4°C. This process was repeated for 3 times. The supernatant was transferred to clean vials and evaporated to dryness under thermostat water bath at 80°C, and then reconstituted in 500 μL of 20% methanol, followed by thoroughly filtered through microporous membrane filters (0.22 μm) prior to analysis. Reference samples of DA, DOPAC and HVA were obtained from Sigma-Alorich (St. Louis, MO, United States). The samples were placed in an Agilent G1367A well plate autosampler and were chilled to 4°C. A 5 μL aliquot of sample was injected an InfinityLab Poroshell 120 Phenyl Hexyl column (4.6 × 50 mm, 2.7 μm column (Agilent Technologies, CA, United States) with a C18 Security Guard Cartridges (2.0 mm × 4.0 mm, Phenomenex, CA, United States). The mobile phase A was acetonitrile, and B was 0.1% formic acid aqueous solution. For DA, the initial mobile phase was consisted of mobile phase A and B (3:97, v/v), increased to 90% mobile phase A at 2.1 min and kept for 1.4 min. Then the mixture was reversed back to 3% mobile phase A at 3.6 min and kept for 1.4 min. For DOPAC and HVA, the initial mobile phase was consisted of mobile phase A and B (5:95, v/v), increased to 95% mobile phase A at 3.5 min and kept for 1.5 min. Then the mixture was reversed back to 5% mobile phase A at 5.1 min and kept for 0.9 min. Mass spectrometric analysis for DA was operated in the positive electrospray ionization (ESI) mode, while the analysis for DOPAC and HVA was in the negative ESI mode (Agilent Technologies, CA, United States). Quantification of the analyte was achieved by multiple reaction monitoring (MRM) with the transitions of m/z 153.9–136.9, 167–123, and 181–137 for DA, DOPAC and HVA, respectively. Data were analyzed using Agilent Mass Hunter (Agilent Technologies, CA, United States) with results in tissue expressed in ng/mg.

All compared data are expressed as mean ± SD. and analyzed using SPSS 13.0 (SPSS, Inc., Chicago, IL, United States). Statistical significance was determined by two-tailed Student’s t-test or one-way ANOVA, followed by LSD or Tamhane’s T2 post hoc test. p < 0.05 was considered statistically significant.

Progressive neuronal damage in the ipsilateral SNc was characterized by the reduced numbers of TH⁺ and NeuN⁺ cells after dMCAO (Figure 1). Compared with the sham-operated group, the numbers of TH⁺ cells at 1, 2, and 4 weeks were significantly decreased in the ipsilateral and contralateral SNc after dMCAO (Figures 1AVI,X,XIV,BI), but this decrease in the ipsilateral SNc was more obvious. Compared to the contralateral and sham-operated groups, the number of NeuN⁺ cells in the ipsilateral SNc displayed a significant reduction after dMCAO (Figures 1AVII,IX,XV,BII). On the contrary, the number of GFAP⁺ cells in the ipsilateral SNc was increased as early as in the first week, peaking at 4 weeks after dMCAO compared with the contralateral and sham-operated groups. These astrocytes cells were characterized by their typical hypertrophic shape with thickened process (Figures 1AVIII,XII,XVI,BIII).

Biodistribution of BMSCs-DiR⁺ was monitored within 14 days by IVIS imaging system. The in vivo fluorescence imaging showed that the transplanted BMSCs-DiR⁺ could migrate from peripheral blood to the ipsilateral brain after dMCAO (Figure 2A). The ex vivo fluorescence imaging demonstrated that BMSCs-DiR⁺ mainly existed in the ipsilateral cortex and SN (Figure 2B). Within the BMSCs-DiR⁺ transplanted dMCAO groups, there were no significant differences in the fluorescence intensities in the ipsilateral cortex and SN across all time points except the first day post-transplantation intravenously. No fluorescent signals were detected in the brain region of BMSCs-DiR^– groups both in vivo and ex vivo (Figure 2C). These results illustrated that intravenous transplanted BMSCs can migrate to the ipsilateral cortex and SN after focal cortical infarction.

Immunohistochemical study showed that the Ki-67⁺ and DCX⁺ cells in sham-operated group were expressed in SNc (Figures 3AI–III), and these cells in the ipsilateral SNc stained deeply both in vehicle and BMSCs groups at 2 days after dMCAO (Figures 3AV–IX). In contrast to the sham-operated group, the number of Ki-67⁺ cell was significantly increased in the ipsilateral SNc at 2 days, and 1, 2, and 4 weeks after dMCAO with or without BMSCs transplantation, and it also increased in the contralateral SNc at 1, 2, and 4 weeks after dMCAO with or without BMSCs transplantation. Notably, BMSCs transplantation increased obviously the number of Ki-67⁺ cells in the ipsilateral SNc at 2 days, 1, 2, and 4 weeks after dMCAO when compared with vehicle groups (Figure 3B). Similarly, in comparison with the sham-operated group, the number of DCX⁺ cells in the ipsilateral SNc was increased at 2 days, 1, 2, and 4 weeks post-dMCAO. Furthermore, transplantation with BMSCs apparently increased the number of DCX⁺ cells in the ipsilateral SNc at 2 days, 2 and 4 weeks when compared with vehicle groups (Figure 3C). In comparison to the contralateral SNc, the number of ipsilateral BMSCs-DiR⁺ cells was significantly increased, with a peak at 2 days after dMCAO (Figure 3D). In addition, in the BMSCs group, the triple-labeled immunofluorescence assay showed that BMSCs-DiR⁺ cells colocalized with Ki-67⁺, DCX⁺, and TH⁺ cells in the ipsilateral SN at 1 week after dMCAO (Figures 3EI–XII). These results indicated that transplanted BMSCs could proliferate and differentiate into mature neurons in the SN after dMCAO and stimulate endogenous neurogenesis.

The schematic diagram in Figure 4A represented that cortical infarction marked with arrowhead was induced by dMCAO and the secondary damage of ipsilateral SN marked with asterisk occurs after dMCAO. Nissl-staining assay in Figure 4B clearly showed that ischemic foci were confined within the cortex at 4 weeks after dMCAO. The relative infarction volumes at 4 weeks after dMCAO were evaluated in Figure 4C. There was no significant difference in the relative infarct volume between BMSCs and vehicle treatment groups after dMCAO. To assess the long-term functional deficits and possible recovery with or without transplanted BMSCs treatment after dMCAO, the adhesive-removal test and NSS were carried out. The mean time to remove the adhesive from the forepaws was significantly shorter in the BMSCs group than that in the vehicle group after 16 days of dMCAO (Figure 4D). Furthermore, compared with the vehicle group, rats in the BMSCs group exhibited higher NSS after 16 days of dMCAO (Figure 4E). These data suggest that BMSCs treatment could improve the neurological functional outcome after dMCAO.

To determine whether transplanted BMSCs offer neuroprotection in the ipsilateral SN after focal cortical infarction, neuronal damage and astrocytic activation were evaluated in the SNc at 1 and 4 weeks after dMCAO, respectively. Compared with sham-operated and contralateral groups, the numbers of TH⁺ and NeuN⁺ cells were significantly decreased, while GFAP⁺ cells were increased in the ipsilateral SNc at 1 and 4 weeks after dMCAO. As expected, after BMSCs treatment, the numbers of TH⁺ cell at 4 weeks and NeuN⁺ cell at 1 and 4 weeks were evidently increased in the ipsilateral SNc and GFAP⁺ cell were decreased at 1 and 4 weeks after dMCAO when compared to vehicle groups (Figures 5A,B). Furthermore, we detected the concentrations of DA and its metabolites, DOPAC and HVA, in the striata at 1 and 4 weeks after dMCAO with or without BMSCs treatment. Compared with sham-operated and contralateral groups, the concentration of DA was significantly decreased in the ipsilateral striatum at 1 week after dMCAO. Transplantation of BMSCs apparently increased the concentration of DA in the ipsilateral striata at 1 and 4 weeks after dMCAO (Figure 5CI). Similarly, the concentrations of DOAPC and HVA were significantly decreased in the ipsilateral striatum at 1 week after dMCAO when compared with sham-operated and contralateral groups. As expected, transplantation of BMSCs prevented the decline of the concentration of DOAPC and HVA in the ipsilateral striatum at 1 week after dMCAO (Figures 5CII,III). These data showed that BMSCs transplantation could mitigate the secondary damage through promoting dopaminergic neuronal survival and alleviating astrocytic activation, thereby increasing contents of DA and its metabolites, in the ipsilateral SN after dMCAO.

To investigate whether BMSCs transplantation enhanced cortico-striatum-nigral connection after dMCAO, PRV-152 for retrograde tracing was injected stereotaxically to the ipsilateral SN with or without BMSCs transplantation at 7 days of dMCAO (Figure 6). Figure 6A is the schematic diagram showed that PRV-152 was injected to the ipsilateral SNr of rats, and the regions of interest of cortex, striatum and SNr. After PRV-152 was injected into the ipsilateral SNr, PRV-152 would reach to the regions of interest aforementioned by the way of retrograde axoplasmic transport in sham-operated and dMCAO groups (Figure 6B). PRV-152⁺ staining in the ipsilateral cortex, striatum and SNr were detected by immunofluorescence assay. Figures 6C–E showed that PRV-152⁺ cells were characterized by PRV-152⁺ staining around the nucleus (Figures 6C–E). In sham-operated rats, few PRV-152⁺ cells were visible in the ipsilateral cortex, striatum and SNr (Figures 6CI–XII). In the vehicle group, PRV-152⁺ cells mainly distributed in the ipsilateral cortex, at 4 weeks after focal cortical infarction (Figures 6DI–XII). Importantly, compared with sham-operated and vehicle groups, the number of PRV-152⁺ cells in BMSCs treatment group was obviously increased in the ipsilateral cortex, striatum and SNr at 4 weeks after focal cortical infarction (Figures 6EI–XII,F). These results suggested that the intravenous transplantation of BMSCs potentially enhanced cortico-striatum-nigral connections after dMCAO.