Spinal Ligament Samples.

Spinal ligament samples of the ligamentum flavum (LF) were collected en bloc from 12 patients who underwent surgery at the thoracic vertebral level. Samples of OLF were obtained from six patients (four males and two females; mean age at surgery, 69 years; age range, 56–77 years) who had posterior decompression surgery for OLF. As a control, non-OLF samples were obtained from six patients (five males and one female; mean age, 50 years; age range, 22–81 years) who had undergone posterior surgery for spinal tumor, syringomyelia, or burst fracture. No patients had evidence of congenital osteoarthropathy or rheumatism. The study protocol was approved by the Human Ethics Review Committee of Hirosaki University Hospital, and sufficient informed consent was obtained from each patient.

Cell Isolation and Culture.

We isolated MSCs as described previously (Harada et al., 2014). Cells were identified as MSCs by fluorescence-activated flow cytometry. MSCs were stained using an anti-human CD34 mouse antibody coupled with phycoerythrin and an anti-human CD105 mouse antibody coupled with PerCP-Cy5.5, respectively (BD Biosciences, San Jose, CA). Cells at first passage (1 × 10⁶) were harvested and washed with PBS containing 20 µl/ml fetal calf serum (FCS). Cells suspended in PBS (50 µl) containing specific antibodies were incubated for 45 minutes on ice. The cells were washed with PBS, filtered through a 35-µm strainer (BD Biosciences, Bedford, MA), suspended in PBS (1 ml), and analyzed by using a FACSAria II Flow Cytometer (BD Biosciences). For cell sorting, gates were defined as negative for CD34 and positive for CD105 according to the intensity of the fluorescence from the isotype control. Data were analyzed by using BD FACSDiva software version 6.1.3 (BD Biosciences).

Preparation of Tissue.

Non-OLF and OLF were fixed with formaldehyde (100 mg/ml) at 4°C for 1 week. Samples of ossified LF were further decalcified with KC-X solution (Falma, Tokyo, Japan) for about 4 days. Then, serial 4-μm-thick sections were prepared using paraffin-embedded specimens and processed for H&E staining and immunohistochemistry.

Immunohistochemical Staining.

Immunohistochemical analysis was accomplished as described previously (Chin et al., 2013). Briefly, antigen retrieval and blocking of non-specific protein binding were performed after deparaffinization of sections. Then, the sections were incubated overnight at 4°C with a single primary antibody or a mixture of two primary antibodies diluted in 10 mg/ml BSA. Primary antibodies against MSC markers included a mouse monoclonal anti-CD73 antibody (1:50 dilution; Abcam, Cambridge, MA), rabbit polyclonal anti-CD90 antibody (1:100 dilution; Abcam), and mouse monoclonal anti-CD105 antibody (1:50 dilution; Dako, Tokyo, Japan). A mouse monoclonal anti-SDF-1 antibody (1:50 dilution; Abcam); rabbit polyclonal anti-SDF-1 antibody (1:50 dilution; Abcam); mouse monoclonal anti-CXCR4 antibody (1:100 dilution; Abcam), and rabbit polyclonal anti-CXCR4 antibody (1:500 dilution; Abcam) were used to detect the chemokine SDF-1 and its receptor, CXCR4.

For single staining, the sections were treated with secondary and tertiary agents, and then N, N′-diaminobenzidine was used to visualize peroxidase deposition at antigenic sites. Finally, the sections were lightly counterstained with H&E. And then, we calculated the prevalence of SDF-1–positive and CXCR4-positive cells in the collagenous matrix area. The ratio of positive cells to nucleated cells was the defined prevalence.

For double staining, sections were incubated with antibodies with green and red-fluorescent dyes, and finally the sections were lightly counterstained with DAPI (4′,6′-diamidino-2-phenylindole dihydrochloride). After staining with double immunofluorescence, the sections were observed under a confocal laser-scanning microscope (Digital Eclipse C1si/C1 plus; Nikon Instruments, Tokyo, Japan). Image data were saved as raw TIFF files without further thresholding or filtering (e.g., without background subtraction).

CXCR4 Expression Analysis by Using Real-Time Polymerase Chain Reaction.

Total RNA was extracted from cultured MSCs using an RNeasy Mini Kit (QIAGEN, Valencia, CA) according to the manufacturer instructions. Genomic DNA contamination was eliminated by DNase treatment. First-strand cDNA was synthesized from total RNA (1 μg) using a standard random hexamer priming technique. Real-time polymerase chain reaction (PCR) was carried out with Power SYBR Green PCR Master Mix on an ABI Prism 7000 Sequence Detection System (Applied Biosystems, Foster City, CA). Specific primer pairs for CXCR4 (GenBank accession number NM_003467; forward primer, 5′-CTGCCCTCCTGCTGACTA-3′; reverse primer, 5′-GGGTAGAAGCGGTCACAGA-3′) were designed by using Primer-BLAST software (http://www.ncbi.nlm.nih.gov/tools/primer-blast/). Conditions for PCRs were as follows: 95°C for 1 minute, and then 50 cycles of 95°C for 15 seconds and 60°C for 1 minute. All samples were analyzed in parallel for glycerol 3-phosphate dehydrogenase (G3PDH) expression as an internal standard. The data of real-time PCR experiments were represented as Ct values. The Ct value is defined as the cycle number at which the amount of PCR product exceeds the threshold level. We used the comparative Ct method to compare mRNA expression in samples with that in the control in each experiment. Reactions were carried out in triplicate for each cell sample. Results are expressed as the mean ± S.D. (n = 6). Graphs depict relative expression levels compared with the control (non-OLF MSCs), unless otherwise indicated.

CXCR4 Expression Analysis by Western Blotting.

After MSCs from non-OLF and OLF patients (six cell samples each) reached confluency, the cells were lysed in RIPA buffer (250 mM Tris-HCL, pH 7.6, 150 mM NaCl, 10 mg/ml NP40, 10 mg/ml Na-deoxycholate, and 10 mg/ml SDS). After a Bradford protein assay, protein samples (3 μg each) were resolved by SDS-PAGE and then transferred to polyvinylidene difluoride membranes (Immobilon-FL; Millipore Corporation, Billerica, MA). The membranes were blocked with 1 mg/ml dry skim milk in PBS with 1 µl/ml Tween 20 and incubated with primary antibodies against CXCR4 (rabbit polyclonal antibody, 1:500 dilution; Abcam) or β-actin (rabbit polyclonal antibody, 1:1000 dilution; Abcam) and then an Alexa Fluor 680–labeled goat anti-rabbit secondary antibody (1:1000; Abcam). Proteins were detected using the Odyssey Imaging System (LI-COR Biosciences, Lincoln, NE).

Cell Migration Analysis.

Cell migration was assayed by the scratch wound healing method using an in vitro incisional wound model, as reported previously (Liang et al., 2007). MSCs were cultured in six-well plates (BD Biosciences, Franklin Lakes, NJ) with α–minimal essential medium containing 100 µl/ml FCS. At 80% confluence, FCS in the medium was reduced to 20 µl/ml to avoid cell proliferation during the assay period. After 1 day of incubation, scratch wounds of 1 mm in width were created in confluent monolayers by using an aseptic 1000-µl pipette tip. After detached cells were washed out, the wounded monolayers were cultured for 24 hours. Images of the wound area were obtained at 0 and 24 hours using a phase contrast microscope (Olympus, Tokyo, Japan). The area between the wound edges was analyzed using NIH ImageJ software (Bethesda, MD). The wound area was defined as follows: wound healing (%) = [1 − (wound area (t = 24 hours)/wound area (t = 0 hour))] × 100. Data were averaged over six independent experiments. Each experiment was repeated three times. MSCs were treated with CXCR4 inhibitor AMD3100 (Abcam) at a final concentration 300 nM to determine the dependency of migration on CXCR4.

Cell Chemotaxis Assay.

Chemotactic migration was evaluated, as described previously (Mori et al., 2012), in chemotaxis chambers (Chemotaxicell; Kurabo, Osaka, Japan) inserted into a 24-well culture plate. MSCs (2 × 10⁴) were spread into the upper chamber, and the lower chamber was filled with PBS with or without SDF-1 at various dosages (10, 30, and 100 ng/ml) in α–minimal essential medium containing 20 µl/ml FCS. The cells were incubated for 48 hours at 37°C with 5% CO₂. The chambers were washed two times with PBS and then fixed with 100 mg/ml formaldehyde for 5 minutes, followed by washing two times with water. The chamber was stained with methyl green for 20 minutes and washed three times with water, and then the number of stained cells in the filter, which had migrated from the upper chamber, was counted under a microscope.

Statistical Analysis.

All data analyses were performed by KyPlot 5.0 software (Kyenslab, Tokyo, Japan). Differences among groups were analyzed using one-way analysis of variance with the Student-Newman-Keuls post hoc correction procedure. P values less than 0.05 were regarded as statistically significant.