All procedures related to the use of animals complied with the guide published by NIH [14]. Adult male Sprague-Dawley rats (80–100 g) were purchased from the Animal Center of Zhejiang Chinese Medical University, Hangzhou, China (Laboratory Animal Certificate: scxk 2008-0016).

MSCs were obtained from rat femoral and tibia bone marrow as previously described [15]. Briefly, muscles, connective tissue and epiphyses were removed from femur and tibia. Marrow was harvested by 1.0 mL syringe with cell medium that consist of Dulbecco's modified Eagle's medium/Ham's F-12 (DMEM/F-12; Invitrogen, Grand Island, NY, USA), 10% fetal bovine serum (FBS; Life Technologies, Carlsbad, CA, USA) and 1% penicillin-streptomycin (HyClone Laboratories, Inc., South Logan, Utah, USA). Cells were then centrifuged for 10 min at 1000 rpm, re-suspended in the cell medium and cultured at 37 °C, 5% CO₂. After 24 h, all non-adherent cells were removed by medium exchange; the fresh medium was subsequently replaced every 2 days. The adherent cells were trypsinized at 80% confluence, and passaged at 1:2. MSCs from passage 3 were used in the subsequent experiments.

MSCs had received a hypoxia pretreatment during 24 h to mimic the microenvironment in the damaged area in order to activate the endogenous gene expression, which is the natural effect of cells in response to hypoxia. Briefly, the cells were placed in a hypoxia incubator and balanced with a gas mixture (1% O₂, 5% CO₂, 94% N₂) for 10 min at a flow rate of 1 L/min. Then, the hypoxia incubator was sealed to maintain a hypoxic environment before being put into a cell incubator under 5% CO₂ at 37 °C.

MSCs with HIF-1α overexpression were constructed by a lentivirus system following the manufacturer's instructions (Invitrogen, Carlsbad, CA, USA). The cells were infected with a lentivirus encoding either GFP alone or both GFP and HIF-1α. Then the MSCs were assigned into three experiment groups: HIF-1α-MSCs (transfection with HIF-1α and GFP), GFP-MSCs (transfection with GFP), NM-MSCs (undisturbed normal MSCs).

The transfection efficiency of the cells was detected by fluorescence microscopy analysis after transfecting for 48 h. The protein expression level of HIF-1α in MSCs was detected by Western blotting.

The expression level of HIF-1α was evaluated by western blotting after hypoxia pretreatment. Equal amounts of protein from MSCs were separated by 10% SDS-PAGE and transferred to polyvinylidene difluoride membranes (Millipore, Boston, MA). After being blocked with 5% nonfat dry milk for 2 h, the membranes were incubated with rabbit polyclonal antibodies against either HIF-1α (1:1000 dilution, Abcam Technology, Cambridge, MA, USA) or GAPDH (1:2000 dilution, Abcam Technology, Cambridge, MA, USA) overnight at 4 °C. The membranes were then incubated with a second antibody (anti-rabbit IgG HRP, 1:3000; Abcam Technology, Cambridge, MA, USA) and ECL detection systems (Thermo Fisher Scientific, Waltham, MA, USA) were used for detection, GAPDH was used as a loading control.

An assay with 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) was performed to determine the influence of HIF-1α overexpression on MSC viability. Briefly, pretreated MSCs were seeded in 96-well plates at a density of 2000 cells per well, and cultured for 48 h, and NM-MSCs were used as a control. 10 μL (5 g/L in PBS) of MTT (Sigma Chemical Co., St. Louis, MO, USA) were added to each well and incubated at 37 °C for 4 h. The supernatants were carefully discarded and 150 μL of dimethyl sulfoxide (Sigma Chemical Co., St. Louis, MO, USA) was added. Absorbance was measured at 490 nm by microplate reader (Bio-RAD Model 680, Cal, USA).

In order to make it easily observable, MSCs were stained with chloromethylbenzamido dialkylcarbocyanine (CM-DiI, Molecular Probes, USA) at the dose of 1 μg CM-DiI per million cells, and washed twice after incubation at 37 °C for 20 min. Labeled MSCs were used in the following experiments.

A cell migration assay was performed in Transwell chambers (pore size: 8 μm, polycarbonate membrane; Corning Incorporated, NY, USA) to evaluate the migratory capacity of MSCs. Transwell chambers were put into six well plates. CM-DiI labeled cells were harvested in a serum-free medium; the density was adjusted to 2 × 10⁵ cells/ml. 200 μL of cell suspension were plated in the upper chamber and 500 μL of a fresh medium containing 10% FBS were added to the lower chamber. After incubation at 37 °C under 5% CO₂ for 24 h, the cells on upper membrane layer were gently wiped away using a cotton swab and the cells migrating to the lower membrane were fixed with 4% paraformaldehyde. The number of cells in the lower side of the membranes was counted microscopically.

As HI model was established as in the previous method of Rice et al. [16]. The rats were anesthetized with a 10% chloral hydrate solution (0.3 mL/100 g), and the left carotid artery was isolated and ligated doubly. After 2–3 h, the waking rats were put into a transparent container which was ventilated with a constant flow of mixed gas containing 8% oxygen and 92% nitrogen for 2 h.

The animals were randomly assigned into four experimental groups after HI treatment: sham group, HI-vehicle group, HI-GFP-MSC group, HI-HIF-1α-MSC group. For the rats in the sham group, the carotid artery was isolated but not ligated, and no hypoxia exposure was performed. The rats in the HI-GFP-MSC group and the HI-HIF-1α-MSC group had received 5 × 10⁵ MSCs re-suspended in 0.5 mL of a physiological saline solution by intravenous injection in the tail 24 h after HI treatment, and the other rats received the same volume of physiological saline solution.

Rats (n = 6 per group) have been submitted to the Morris water maze test [17] on day 14 after HI to evaluate their spatial learning and memory abilities. The maze consisted of a black circular tank (160 cm diameter × 50 cm height) filled with water (temperature: 22–24 °C). An automatic tracking system (San Diego Instruments, San Diego, CA, USA) was used to record the swimming path and time. Each rat was tested for five training days and one test day. In the task, the rat was put in every quadrant of the pool sequentially, and swam for 60 s to locate the submerged platform; if it failed, the experimenter would guide it to the platform. Then the rat was allowed to remain on the platform for 10 s. On the last day, the platform was removed before the test, and the swimming time in the former platform quadrant was recorded.

Rats brains were obtained on days 7, 14 and 21 after HI (each time point n = 6 per group), and each hippocampus was embedded in paraffin. The paraffin sections (4 μm thick) were prepared for detection of the migration of MSCs and histological examination in the hippocampus. To measure the migrated cells’ number, the sections were observed with a fluorescence microscope to count the CM-DiI positive cells. The histological examination was determined by hematoxylin and eosin (HE) staining.

The data obtained were presented as mean ± standard deviation (SD). Differences between groups were evaluate by analysis of variance for multiple comparisons using SPSS 19.0. A value of P < 0.05 was considered to be statistically significant.

After transfection for two days, GFP expression observed under the fluorescence microscope reached up to 90% of the cultured MSCs (Fig. 1A). In the meantime, western blotting was performed to investigate the expression of HIF-1α in MSCs after hypoxia pretreatment. In the present study, the expression of HIF-1α in HIF-1α-MSCs was approximately two times that in NM-MSCs and GFP-MSCs, while the expression level in GFP-MSCs was in accordance with NM-MSCs (Fig. 1B and C). The results indicated that HIF-1α overexpressed MSCs were successfully constructed by a recombinant lentivirus system.

MSC viability was assessed by an MTT assay to determine the effect of HIF-1α overexpression. As shown by our results, HIF-1α overexpression immensely improved the viability of MSCs, while there was no significant difference between NM-MSC and GFP-MSC groups (Fig. 2A). The result implied that MSCs with overexpression of HIF-1α had better viability.

The migration of MSCs was analyzed both in in vitro and in vivo experiments. To observe the migration of MSCs directly, the cells were labeled with CM-DiI, a fluorescence marker. The results showed that more than 95% of MSCs were CM-DiI positive after having been labeled (Fig. 3A).

According to the Transwell assay, the number of cells migrating to the lower compartment in HIF-1α-MSC group was more than that in NM-MSC and GFP-MSC groups after hypoxia pretreatment, while no significant difference was found between NM-MSC and GFP-MSC groups after culturing for 24 h (Fig. 3B).

One day after HI, rats received an intravenous transplantation of MSCs, and the presence of migrated MSCs in the hippocampus after transplantation was investigated by observing CM-DiI positive cells. CM-DiI positive cells were clearly observed in the hippocampus on day 7 after HI. And from day 7 to day 21 subsequent to HI, the number of CM-DiI positive cells in the hippocampus increased gradually in a time-dependent manner. A remarkable increase in the recruitment of HIF-1α transduced MSCs was observed in the hippocampus at every time point compared with GFP transduced MSCs (Fig. 3C). Statistical analysis showed that HIF-1α overexpression improved the migration capacity of MSCs toward the sites of injury.

The quantitation of migrated cells demonstrated significant increases in MSCs motility in HIF-1α overexpressed MSCs both in in vitro and in vivo experiments in comparison to GFP infected cells.

In order to investigate the therapeutic effect of HIF-1α transduced MSCs, the Morris water maze test and HE staining in the hippocampus were performed on HI rats. On day 14 after HI, the spatial performance of four groups was evaluated by the Morris water maze test. In the spatial probe test, spatial learning ability was significantly affected by HI, as HI-treated rats spent less time in the target quadrant compared to the sham group. The results suggested that the rats suffered from cognitive dysfunction due to HI treatment. There was a significant increase of time in the target quadrant in both MSC-transplanted groups compared with HI-vehicle group, but a more increase in the amount of time have been found in HI-HIF-1α-MSC group (Fig. 4A). These results indicated that the HI-induced rats achieved partial remission on memorial and cognitive functions after MSCs transplantation, while HIF-1α overexpression could promote this effect.

Some visualized pathological changes were found in HE staining. The data supported that the morphological structures of the hippocampus in the sham group were normal, with uniform dyeing (Fig. 4B1). In contrast with the sham group, the pathological changes in the hippocampus in the HI-vehicle group revealed serious injury, the cells were disordered, and cavitation was visible, suggesting that the HI model had been successfully established. Meanwhile, the rats in HI-GFP-MSC and HI-HIF-1α-MSC groups showed similar but slight pathological changes at every time point after having been subjected to HI, and the changes in the MSCs with HIF-1α transfection group were less marked than those of the GFP control group (Fig. 4B2). The results illustrated that the transplantation of MSCs ameliorated pathological changes after HI, while HIF-1α overexpression could enhance therapeutic efficiency.