Cryopreserved adult hMSCs (Poietics^® normal human mesenchymal stem cells) were purchased from Cambrex Bio Science Walkersville (Walkersville, MD, USA). We cultured hMSCs (passages 4–12) in Dulbecco’s modified Eagle’s medium (DMEM)-low glucose (Hyclone) containing 10% fetal bovine serum (FBS; Gibco-BRL) at 37°C with 5% CO₂.

The lentiviral vector for expressing the mouse Olig2 gene was constructed by inserting the EcoR I – EcoR V gene fragment into the EcoR I – pme I site of the LentiM1.41 vector. The LentiM1.41 lentiviral vector was designed to produce interesting gene promotion from the murine cytomegalovirus (mCMV) promoter and to express eGFP from the PGK promoter. The nucleotide sequences of the constructs were verified by sequencing.

Macrogen, Inc. produced pseudotyped lentiviruses as described previously (19, 20). Briefly, three plasmids, a transfer vector, a VSV-G expression vector, and a gag-pol expression vector, were co-transfected into HEK-293T cells at a 1:1:1 molar ratio using Lipofectamine Plus (Invitrogen). The culture supernatant containing viral vector particles was harvested 48 hours later, clarified with a 0.45-μm membrane filter (Nalgene), concentrated using a Centricon Plus-20 (Millipore) and immediately stored at −70°C in a deep-freezer. Titers were determined by p24 ELISA and infection into HeLa cells. The eGFP expression of transduced cells was observed and photographed under a fluorescence microscope. The titer was approximately 10⁶~10⁷ transduction units (TU) per mL without further concentration.

All plasmid constructs were introduced into hMSCs by viral infection. The cells (2.4×10⁴ cells/well) were plated in six-well plates 2 days before transduction. hMSCs were infected with lentiviral vectors in the presence of 8 μg/mL polybrene. After 6 hours, the media were replaced with fresh growth media. Two days after viral infection, cells were selected using 1 μg/mL puromycin for 7 days.

Cells were fixed on coverslips with 4% paraformaldehyde at room temperature for 30 minutes and permeabilized for 10 minutes with 0.2% Triton X-100 and 1% bovine serum albumin (BSA) in PBS. Blocking was performed by incubating cells for 1 hour with 5% normal goat serum in PBS. Cells were then incubated overnight at 4°C with rabbit polyclonal anti-Olig2, 1:5,000 (gift from Dr. C. Stiles, Harvard Medical School). After washing with PBS, cells were incubated with goat anti-rabbit conjugated with Alexa Fluor 546, 1:500 (Invitrogen). All cells were mounted with fluorescent mounting medium (Dako). Immunofluorescence was examined using a laser-scanning confocal microscope (Fluoview FV300, Olympus).

Adult male Sprague-Dawley rats (Daehan Biolink, Chungbuk, Korea) weighing 260~280 g at the time of surgery were housed in groups of four and allowed free access to food and water. All animal procedures were carried out under the approval of the Institutional Animal Care and Use Committee of Seoul National University. Acute SCI was induced with an NYU Impactor. The rats were anesthetized by intraperitoneal injections of Zoletil (35 mg/kg) and Rompun (2 mg/kg) and a laminectomy was performed at the level of T9. The exposed dorsal surface of the spinal cord in each rat was then subjected to a weight-drop impact. To obtain moderately contused SCI models, a 10-g weight–impact rod was dropped from a 25-mm height. The contusion impact velocity and compression rate were monitored to guarantee consistency of injury between experimental animals. During recovery, the rats’ rectal temperatures were maintained at 37°C using a feedback- regulated heating pad. Postoperative nursing care included bladder expression twice per day. Gentamicin sulfate (1 mg/kg) was administered prophylactically to all animals for a week after SCI.

Open-field testing procedures, also known as Basso–Beattie–Bresnahan (BBB) procedures (21), were used to measure the functional recovery of rat hindlimbs. The scale used to measure hindlimb function with these procedures ranges from a score of 0, indicating no spontaneous movement, to a maximum score of 21, with an increasing score indicating the use of individual joints, coordinated joint movement, coordinated limb movement, weight-bearing and other functions. Rats were first gently adapted to the open field used for the test. After each rat had walked continuously in the open field, two investigators conducted 5-minute testing sessions on each leg. The open-field test was performed at least once per week from day 1 after SCI to 9 weeks after laminectomy on all animals in the study.

After rats had been subjected to behavioral tests for 1 week, they were randomly allocated into groups that received cell-culture medium, GFP-expressing hMSC transplants, or Olig2-GFP-expressing hMSC transplants. At 7 days after SCI, the rats were anesthetized by intraperitoneal injections of Zoletil (35 mg/kg) and Rompun (2 mg/kg), and received either 5 μL of GFP-hMSCs (1×10⁵/μL) or Olig2-GFP-hMSCs (1×10⁵/μL) transplanted through a capillary glass tube into the epicenter of the injury, or received a 5-μL injection of culture medium into the epicenter of the injury (control). To induce immunosuppression, cyclosporine A (1 mg/100 g) was injected daily from 2 days before injury until the day of transplantation or injection of culture medium.

At 8 weeks after the induction of SCI, all rats were sacrificed by perfusion with 4% paraformaldehyde in phosphate-buffered saline (PBS) under anesthesia. The spinal cords were removed and fixed for 18 hours in 4% paraformaldehyde, followed by overnight immersion in 30% sucrose in PBS. Serial longitudinal 12-μm-thick sections of the spinal cord were obtained with a cryostat. The specimens were stored at −80°C before further processing. Staining with hematoxylin and eosin (H&E) was performed for histopathological analysis.

For immunohistochemical analysis, the spinal cord sections were rinsed with PBS for 3 minutes and fixed for 30 minutes with 4% paraformaldehyde. The sections were then treated with blocking solution to prevent nonspecific staining and were incubated with anti-neurofilament-M (1:5,000; rabbit polyclonal; Chemicon), anti-glial fibrillary acidic protein (GFAP) (1:1,000; mouse monoclonal; Chemicon), anti-Iba1 (1:1,000; rabbit polyclonal; Wako), anti-Olig1 (1:20,000; rabbit polyclonal; kindly provided by Dr. C. Stiles, Harvard Medical School). After washing with PBS, the spinal cord sections were incubated with the following secondary antibodies: Alexa Fluor 488 anti-mouse IgG (Invitrogen) and Alexa Fluor 546 anti-rabbit IgG (Invitrogen). The sections were mounted on glass slides with a fluorescent mounting medium (Vectashield; Vector Laboratories) and observed under a laser-scanning confocal microscope (Fluoview FV300, Olympus).

Histological analysis was carried out on 12-μm-thick longitudinal sections of spinal cord obtained from culture medium (n=5), GFP-hMSCs (n=5), or Olig2-GFP-hMSCs (n=6) transplanted rats that were stained with H&E. Lesioned areas were outlined using Image Pro software and lesion sizes were calculated for statistical analysis. Serial sections were also immunohistochemically processed for each marker protein.

Data were analyzed by two-tailed Student’s t-test or one-way analysis of variance (ANOVA) followed by post hoc comparisons with Fisher’s protected least significant difference (PLSD) test. A probability value of <0.05 was considered statistically significant. All values were expressed as the mean±standard error of the mean (SEM).

In this study, we ectopically expressed Olig2 in hMSCs using Olig2-GFP-containing lentiviral vectors (Olig2-GFP-hMSCs). Control hMSCs were infected with GFP-containing lentiviral vectors (GFP-hMSCs). Both GFP-hMSCs and Olig2-GFP-hMSCs expressed green fluorescence under a fluorescence microscope (Fig. 1A). Immunostaining for Olig2 demonstrated that Olig2-GFP-hMSCs ectopically expressed Olig2, whereas GFP-hMSCs did not. No significant differences in morphology or viability were observed between these cells. We next sought to determine survival and distribution of GFP-hMSCs and Olig2-GFP-hMSCs in spinal cord lesions at 7 weeks after transplantation. Contusion injuries of the rat thoracic spinal cord was performed according to a standard protocol (21). At 7 days after SCI, animals received transplantation of GFP-hMSCs or Olig2-GFP-hMSCs (Fig. 1B). Transplanted GFP-hMSCs and Olig2-GFP-hMSCs were found by GFP immunohistochemistry in the spinal cord lesion at 7 weeks after cell transplantation (Fig. 1C), indicating survival of engrafted cells.

After the induction of SCI, all experimental rats showed flaccid paralysis of hindlimbs during the first few days after surgery. To determine the effect of Olig2-GFP-hMSC transplantation on functional recovery in contusion SCI rats, BBB scores were evaluated one day after injury. As shown in Fig. 2, one week after SCI (the day before hMSC transplantation), all rats received a score of 5 or less. There were no significant differences among the three groups. At the second week post-treatment, animals transplanted with Olig2-GFP-hMSCs showed significantly faster functional recovery. From 3 to 7 weeks following treatment, rats transplanted with Olig2-GFP-hMSCs showed significant increases in BBB scores compared to rats transplanted with GFP-hMSCs or sham-injected with culture medium (vehicle). At 6 and 7 weeks following treatment, the BBB scores of rats transplanted with GFP-hMSCs were greater than those of sham-injected rats. These results indicate that rats receiving transplantation of Olig2-GFP-hMSCs exhibited significantly enhanced functional recovery compared to rats transplanted with GFP-hMSCs or sham-injected with vehicle.

To elucidate the mechanisms of enhanced locomotor recovery in rats transplanted with Olig2-GFP-hMSCs, we performed morphological analysis of spinal cord lesions using H&E staining. H&E staining showed that various cavity morphologies were formed at the spinal cord lesions of all SCI animals, including those that received transplantations of GFP-hMSCs or Olig2-GFP-hMSCs as well as vehicle (Fig. 3A–C). The spinal cord cavity was largest in the vehicle transplant group. The spinal cord cavity of the GFP-hMSC transplant group and the Olig2-GFP-hMSC transplant group was smaller than that of the vehicle transplant group. H&E stained images were manually outlined and analyzed to quantify cavity area (Fig. 3A–C). Cavity size in the Olig2-GFP-hMSC transplant group (1.18±0.06 mm²) was significantly reduced compared to the GFP-hMSC transplant group (1.46±0.09 mm²) as well as the vehicle transplant group (2.57±0.37 mm²) (p<0.05, Fig. 3D). These results indicate that Olig2-GFP-hMSCs attenuate glial scar formation in spinal cord lesions. Correlation analysis showed a significant correlation (r=−0.967, p<0.01) between the BBB score at 7 weeks post-spinal cord injury and the size of the spinal cord injury site (Fig. 3E).

At 7 weeks after SCI, we performed immunohistochemical analysis in longitudinal sections of injured spinal cord for all three treatment groups: vehicle transplantation, GFP-hMSC transplantation, and Olig2-GFP-hMSC transplantation. Oligodendroglial lineage-associated marker Olig1 immunohistochemistry showed that expression of GFP and Olig1 was higher in the Olig2-GFP-hMSC transplant group compared to the GFP-hMSC transplant group (Fig. 4). This suggests that transplanted Olig2-GFP-hMSCs partially differentiated into Olig1-positive oligodendrocyte-like cells in spinal cords. Next, we investigated the expression of NF-M in transplanted GFP-hMSCs and host cells in the injured spinal cord. Transplanted Olig2-GFP-hMSCs, but not GFP-hMSCs, seemed to be directly associated with NF-M positive axons (Fig. 5). Furthermore, NF-M positive axons were more abundant in the Olig2-GFP-hMSC transplant group than in the GFP-hMSC transplant group. As expected, there were few NF-M positive axons in the epicenters of injured spinal cords in the vehicle transplant group. Microglial marker Iba1 immunohistochemistry revealed increased density of Iba1-positive microglial cells in the host tissue of the vehicle transplant group when compared to the Olig2-GFP-hMSC transplant or GFP-hMSC transplant groups (Fig. 6). Next, we investigated the expression of astrocyte marker GFAP in transplanted hMSCs and host cells at the injured spinal cord. Quite a few cells expressed both GFP and GFAP in the spinal cord in the Olig2-GFP-hMSC transplant group (Fig. 7). This suggests that transplanted Olig2-GFP-hMSCs partially differentiated into Olig2- lineage astrocytes (22).