Eighty adult male C57BL/6 mice, 11–13 weeks of age, body weight 25–30 g, were obtained from Shanghai laboratory animal research center and housed in the experimental animal center of Shanghai Tongji hospital. All the mice were housed under 12 h light/dark cycles, temperature 22 ± 1°C and provided with available food and water ad libitum. All procedures were performed following the national institutes of health for the care and use of laboratory animals (NIH publication No. 8023) and were approved by the animal ethics committee of Shanghai Tongji hospital. All efforts were made to minimize the suffering of animals used and the number of animals needed for this study.

MSCs were isolated and harvested as follows. Briefly, fetal mice were obtained by aseptic cesarean delivery from 13.5 d pregnant mice and flushed with phosphate-buffered saline (PBS). The body of fetal mice was clearly isolated and cut into pieces <1 mm³, followed by trypsinization at 37°C for 5 min, with DMEM used to neutralize trypsin activity. The suspension was centrifuged, resuspended with Dulbecco modified Eagle medium (DMEM, Gibco) supplemented with 10% fetal bovine serum (FBS, Gibco) and cultured in 10 cm dish at 37°C with 5% CO₂ in an incubator. Nonadherent cells were removed every 48 h and primary cells were subcultured 1:3 when 80%-90% confluence. On the third pass (P3) they were trypsinized and counted before administration to the experimental mice.

Identification of MSCs was performed by flow cytometry to examine the surface markers CD11b-PE (1:200, Biolegend), CD29-FITC (1:200, Biolegend), CD34-FITC (1:200, Biolegend), CD44-PE (1:200, Biolegend), CD45-APC (1:200, Biolegend), and SCA1-APC (1:200, Biolegend). Cells (1 × 10⁶) were incubated in a 200 μl buffer (PBS with 1 μl of CD11b-PE, CD29-FITC, CD34-FITC, CD44-PE, CD45-APC, SCA1-APC, and isotope control antibodies, respectively) for 30 min at 4°C in the dark followed by three washes with PBS. Next, the cells were resuspended in 200 μl PBS and analyzed in an FACS instrument (BD AccuriC6).

On the third pass (P3) with 60–80% confluence the cells were trypsinized and plated into 12-well plates at a density of 1 × 10⁴/mL. The cells were cultured in DMEM supplemented with 10% FBS, 100 U/mL penicillin, 1 mg/mL streptomycin in combination with either osteogenic induction medium, including dexamethasone (0.1 μmol/L), β-sodium glycerophosphate (10 mmol/L) and ascorbic acid phosphate (50 μmol/L), or adipogenic induction medium, including dexamethasone (1 μmol/L), insulin (5 or 10 mg/L), IBMX (0.5 mmol/L) and indomethacin (200 μmol/L). The medium was replaced every 3 days for 2 weeks. Calcium deposits and fat droplet formation were observed by Alizarin Red S staining and oil red O staining, respectively.

Transient MCAO model was employed as focal cerebral ischemia/reperfusion model following a previously described method (Haddad et al., 2013). The filament was removed after 1 h occlusion to induce reperfusion. Sham group received the same procedure without the induction of the filament. Mice receiving surgery showing a Longa score from 1 to 3 points at 6 h post-stroke were included in this study (Longa et al., 1989). Those animals that showed lack of neurologic deficit (n = 4), brain hemorrhage (n = 3), or death after ischemia (n = 5) were excluded from the study. Animals were randomly divided into the following 5 groups (n = 16 per group): Sham, MCAO plus saline, MCAO plus borneol, MCAO plus MSCs and MCAO with borneol plus MSCs.

At 24 h after MCAO induction, animals were anesthetized with 1% sodium pentobartital and then subjected to intravenous injections of 5 × 10⁵ MSCs in 0.2 ml sterile PBS solution through caudal vein. MCAO plus saline group received the same procedure except for the injection of the same amount of saline. Borneol was dissolved in 5% Tween 80 and given to mice by gavage at 200 mg/kg 3 days before MCAO induction until the day when the mice were sacrificed.

Neurological deficits were evaluated by global score of neurological score (Wahl et al., 1992) as well as grip and string tests (Haddad et al., 2013), which were performed in a dedicated room at 20–22°C on 1st, 3rd, 7th, and 14th day after MSCs transplantation. The scores of the three tests were added as the global score, the maximum of which was 21. The lower global score indicated the more severe neurological deficits.

At 7 days after MSCs transplantation mice were sacrificed. The brain was removed and then coronally sliced into 1 mm slices. These slices were stained with 2% 2,3,5-triphenylterazolium (TTC) in normal saline for 30 min at 37°C. Normal brain tissue was stained red while the unstained area was considered to be the infarct area. The size of infarct areas were measured by Image J software (National Institutes of Health, Bethesda, USA) and presented as the percentage of the total brain volume.

Apoptotic cell death was detected in 6 μm frozen coronal sections by terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining (Roche Diagnostics, Mannheim, Germany) following the manufacturer's protocol. Sections were counterstained with DAPI. The number of TUNEL-positive cells in infarct zone was counted Image J software (National Institutes of Health, Bethesda, USA).

Two weeks after MSCs transplantation, coronal sections (6 μm) were processed for immunofluorescence staining. Sections were permeabilized with Triton X-100 and blocked with goat serum, then incubated with primary antibodies overnight at 4°C, either mouse anti-NeuN: (1:500; Millipore) or rabbit anti-GFAP: (1:1,000; Dako). Second antibodies including goat anti-rabbit or goat anti-mouse secondary antibody-conjugated to Alexa594 (1:500; Gibco) were used at 37°C for 1 h. Immunofluorescent signaling was observed with an Olympus fluorescence microscope (Olympus, Tokyo, Japan). The number of NeuN-positive cells were counted blindly using Image J software (National Institutes of Health, Bethesda, USA) and the integral optical density (IOD) of GFAP was obtained using Image-ProPlus 4.1 software (Media Cybernetics, Rockville, MD, USA).

Data were expressed as mean ± standard deviation (SD). Normality was determined using the Kolmogorov–Smirnov test. If normality was given and there were no significant differences in variance between groups (F-test), multiple groups were compared using one-way analysis of variance (ANOVA) and followed by LSD post hoc comparisons when appropriate. Otherwise, comparisons for non-normally distributed data were performed using non-parametric Kruskal–Wallis test followed by Mann–Whitney U-test. All statistical analyses were performed using SPSS 20.0 software (IBM, Armonk, NY) and GraphPad Prism 6.05 (GraphPad, La Jolla, CA). Values of P < 0.05 were considered statistically significant.

Before transplantation of MSCs for restoring the neurological function after ischemia, we first confirmed that the mouse fetus-derived MSCs had the features of stem cells. The MSCs showed typical fibroblast-like cell morphology in culture (Figure 1). Flow cytometric analysis confirmed that CD29, CD44 and SGA1 surface markers in MSCs were positive, whereas CD34, CD45, and CD11b surface markers were negative (Figure 2). The cell matrix exhibited nodus-like calcium deposition following the alizarin red staining and enlarged and fused fat drops in some cell bodies appeared salmon pink following oil red staining after 2 weeks induction, suggesting that these MSCs had the ability to differentiate into osteoblasts and adipocytes (Figure 1).

Compared with sham group, mice in MCAO group showed significant deficits in neurological score (P < 0.01; Figure 3A), string test (P < 0.01; Figure 3C) and global score (P < 0.01; Figure 3D), no matter at 1, 3, 7, or 14 d after ischemic stroke except for grip test (Figure 3B). With administration of borneol, mice in MCAO plus borneol group showed significant improvement in neurological score at 3 d (P < 0.01; Figure 3C; P < 0.05; Figure 3D), 7 d (P < 0.05; Figure 3A) and 14 d (P < 0.01; Figure 3A), in string test and global score at 3 d (P < 0.01, Figure 3C; P < 0.05; Figure 3D), 7 d (P < 0.01; Figures 3C,D) and 14 d (P < 0.01; Figures 3C,D) compared with MCAO group. Similarly, mice in MCAO plus MSCs group showed significant improvement in neurological score at 14 d (P < 0.05; Figure 3A), in string test and global score at 7 d (P < 0.05; Figures 3C,D), 14 d (P < 0.01; Figure 3C; P < 0.05; Figure 3D). Compared with MCAO group, mice in borneol plus MSCs group showed significant improvement in neurological score, string test and global score, no matter at 1 d (P < 0.05; Figures 3A,C,D), 3 d (P < 0.01, Figures 3A,C,D), 7 d (P < 0.01; Figures 3A,C,D) or 14 d (P < 0.01; Figures 3A,C,D). Moreover, combination of borneol administration and MSCs transplantation brought more benefit than MSCs transplantation alone on neurological score at 7 d (P < 0.05; Figure 3A) and 14 d (P < 0.05; Figure 3A), on string test at 7 d (P < 0.05; Figure 3C), and on global score at 3 d (P < 0.05; Figure 3D), 7 d (P < 0.05; Figure 3D) and 14 d (P < 0.05; Figure 3D).

TTC staining was conducted to detect brain tissue damage caused by ischemic reperfusion. Infarction area was showed in sham, MCAO plus saline, borneol, MSCs, and borneol plus MSCs group, respectively (Figure 4A). Compared with sham group, MCAO plus saline mice showed significant increase in cerebral infarct volume (27.8 ± 5.4; P < 0.01; Figure 4B). Either borneol administration alone (14.5 ± 3.7; P < 0.01), MSCs transplantation alone (17.4 ± 1.8; P < 0.05) or combination of these two (8.5 ± 1.0; P < 0.01) significantly decreased the infarct volume of ischemic mice, and borneol plus MSCs brought more benefit on infarct volume compared with borneol administration alone and MSCs transplantation alone (P < 0.05 or P < 0.01; Figure 4B).

Cell death was evaluated by detecting TUNEL positive cells at the infarct zone of stroke mice. Photomicrographs of apoptotic cells were showed in sham, MCAO plus saline, borneol, MSCs, and borneol plus MSCs group, respectively (Figure 5A). As show in Figure 5B, TUNEL positive cells was significantly increased in MCAO plus saline mice (89.6 ± 3.64) comparing with sham group (2.0 ± 0; P < 0.01). Borneol and MSCs treatment alone could slightly decreased TUNEL positive cells, with the amount of TUNEL positive cells reaching to 79.0 ± 3.87 and 87.6 ± 7.30, respectively. However, there was no significant statistical difference relative to MCAO group (P > 0.05; Figure 5B). The combination of these two (56.8 ± 9.15) could significantly attenuated TUNEL positive cells comparing with MCAO group (P < 0.01), borneol group (P < 0.05) and MSCs group (P < 0.01) (Figure 5B).

To investigate the effect of borneol and MSCs on neurogenesis in vivo, NeuN⁺ cells (Figure 6A) and GFAP⁺ cells (Figure 6B) in the infarct cortex were assessed by immunofluorescence in sham, MCAO plus saline, borneol, MSCs, and borneol plus MSCs group, respectively. Immunofluorescence study revealed that NeuN⁺ cells decreased significantly in ischemic mice (24.4 ± 4.9, P < 0.01; Figure 6C), while borneol administration alone (43.2 ± 3.9, P < 0.01; Figure 6C), MSCs transplantation alone (34.8 ± 3.3, P < 0.05; Figure 6C) and combination of these two (55.0 ± 6.9, P < 0.01; Figure 6C) all significantly increased the number of NeuN⁺ cells compared with MCAO group. Moreover, borneol plus MSCs had a better effect on NeuN⁺ cells improvement compared with either borneol administration alone (P < 0.05; Figure 6C) or MSCs transplantation alone (P < 0.01; Figure 6C).

The IOD of GFAP increased markedly in MCAO group, whereas after borneol administration (1,129,171 ± 436,009, P < 0.05; Figure 6D), MSCs transplantation (1,207,876 ± 191,142, P < 0.05; Figure 6D) and borneol plus MSCs (722,058 ± 111,587, P < 0.01; Figure 6D), IOD of GFAP decreased significantly compared with MCAO group (2,518,680 ± 885,565). Compared with either borneol administration alone (P < 0.05; Figure 6D) or MSCs transplantation alone (P < 0.05; Figure 6D), borneol plus MSCs group slightly reduce the number of GFAP⁺ cells.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.