The paracrine effect of adipose-derived stem cells inhibits osteoarthritis progression

Skeletally mature female Japanese white rabbits, weighing between 2.5 and 3.0 kg, were used for this experiment.

ADSCs were isolated by modifying a previously reported method [14, 20]. Adipose tissue (1.5 g) was harvested from the posterior cervical subcutaneous adipose tissue of a rabbit and washed with phosphate-buffered saline (PBS) (Wako, Osaka, Japan). The tissue was cut into strips over a period of 5 min. Collagenese (Wako) was dissolved in PBS so that its concentration would be 0.12 % in 25 ml, and was used to digest adipose tissue at 37 °C for 45 min in a water bath. The mixture was shaken every 15 min during the digestion period. Immediately after the reaction was completed, 25 ml of Dulbecco’s modified Eagle’s medium, DMEM (Wako), was added to neutralize collagenase activity. The resulting solution was filtered. The filtrate was centrifuged at 1300 rpm for 6 min at 25 °C, and the supernatant was removed. Next, a pallet of ADSCs was seeded at 5 × 10⁴ cells/cm² in 60.1-cm² tissue culture dishes and cultured with complete medium, DMEM containing 10 % fetal bovine serum (FBS), 100 units/ml penicillin, and 100 μg/ml streptomycin. After being cultured for 1 week, the cells were harvested with 0.25 % trypsin-EDTA, centrifuged at 1300 rpm for 6 min, and washed twice with PBS. The obtained cells were used for intra-articular injections in vivo and co-culture assays in vitro.

Traumatic degeneration was induced as previously described for the anterior cruciate ligament model [21, 22]. This model is characterized by OA-like damage. Hayashi et al. noted that rabbits show considerable individual variability in osteoarthritis progression at 4 weeks after ACLT [23]. As they suggested, we used matched-pair analysis to examine the paracrine effect of ADSCs on osteoarthritis progression in a stricter manner.

Both knees were shaved and disinfected with iodine. ACLT was performed as described by Yoshioka et al. [22]. Briefly, a medial parapatellar incision was made and an arthrotomy was performed. The patella was dislocated laterally and the knee placed in full flexion. The ACL was visualized and transected with a No.11 blade. An anterior drawing test was performed gently to confirm that the ACL was transected completely. The joint was irrigated with sterile saline and closed. A same operation was performed in the contralateral knee. The knee was opened and the patella was dislocated. After gently performing the anterior drawing test, the joint was irrigated and closed. After the operation, free activity was allowed in the cage without immobilization.

ADSCs were isolated from the subcutaneous adipose tissue 3 weeks after ACLT and cultured for 1 week as described above.

For matched –pair analysis of ADSCs, 12 rabbits (24 knees) were used. At 4, 5, 6 weeks after ACLT, ADSCs (2.0 × 10⁶ cells) in 100 μl PBS and 0.1 mL hyaluronic acid (Synvisc®; kindly provided by Teijin Limited, Japan) were intra-articularly injected into the left knee, using a 23 gauge needle. In right knees, the same volume of plain PBS and and hyaluronic acid were injected as a control. Immediately after injection, the rabbit was kept down for 5 min so that injected ADSCs could be attached around the synovium.

Femoral condyles from both knees were harvested at 8 and 12 weeks after surgery following the intravenous injection of 6 ml sodium pentobarbital.

The femoral condyles were dissected and stained with India ink (American MasterTech, CA, USA). Macroscopic pictures were taken using a Panasonic LUMIX digital camera (Panasonic, Tokyo, Japan). Gross findings were classified and scored as described in our previous reports [4, 23, 24]. The medial and lateral femoral condyles were individually scored from grades 0 to 5 and the two scores were summed to obtain a cumulative macroscopic osteoarthritis score. A blind assessment was then independently conducted by two individual examiners (KK, TK) and the scores from the two examiners were averaged to obtain an overall score.

The dissected distal femurs were fixed in a 4 % paraformaldehyde solution after gross morphological examination. The specimens were decalcified in 4 % EDTA solution, dehydrated with a gradient ethanol series, and embedded in paraffin blocks.

Ten coronal sections were prepared in the coronal plane through the middle of the femoral condyles, and one section from each sample, which included the most severely degenerated area, was used for each of the histological analyses.

The specimens were stained with safranin O or H&E using standard procedures.

Histological sections were visualized using a fluorescence microscope (Keyence Japan, Osaka, Japan), assessed in a blind manner by two individual examiners (TK, KK), and quantified according to the cartilage OA histopathology grading system methodology of the Osteoarthritis Research Society International (OARSI) [25].

Immunohistochemical examinations were performed as follows. In brief, after deparaffinization, sections were incubated with 0.3 % hydrogen peroxide for 30 min. Then, sections were treated with hyaluronidase for 60 min after which they incubated with mouse anti-human MMP-13 monoclonal antibody (1:20; AnaSpec Inc., San Jose, Ca, USA), or mouse anti-rabbit MMP-3 monoclonal antibody (1:50; Daiichi Fine Chemical Co. Toyama, Japan) [26]. All antibody dilutions were made in PBS. After an overnight reaction with the primary antibody at 4 °C, sections were incubated with labeled polymer-HRP anti-mouse IgG (Dako, Tokyo, Japan) at room temperature for 30 min. Signals were visualized with 3, 3’-diaminobenzidine tetrahydrochloride, and nuclei were counterstained with hematoxylin. A semi-quantitative method that assigns immunohistochemistry values as a percentage of positive cells (MMP-13, MMP-13) was provided for a complete assessment of protein expression, with a maximum score of 100 % [12].

Results of the histological and immunohistochemical analyses were subjected to a blind evaluation by two observers (TK, KK).

On the day of intra-articular injection, ADSCs were labeled for cell tracking with a fluorescent lipophilic tracer 1,10-dioctadecyl-3,3,30,30-tetramethylindocarbocyanine perchlorate (DiI; MolecularProbes ). For labeling, the cells were re-suspended at 1 × 10⁶ cells/ml in aMEM, and DiI was added at 5 ml/ml in a serum-free DMEM. After incubation for 20 min at 37 °C with 5 % humidified CO2, the cells were centrifuged at 1300 rpm for 6 min and washed twice with PBS, then re-suspended in PBS for the injection.

An examination was performed to determine where the ADSCs existed after the intra-articular injection.

A fresh frozen coronal section containing a central portion of the knee joint was prepared using Kawamoto’s method [27].

Primary culture of chondrocytes was performed using articular cartilage tissues harvested from a rabbit. Briefly, thinly sliced cartilage tissues were incubated with 0.15 trypsin with 0.1 % collagenase for one hour. Subsequently, cartilage tissues were incubated in DMEM with 0.15 % collagenase for 2 h 30 min. The cells were then collected by centrifugation, seeded into collagen I coated 24-well plates (collagenI multiwell plates, BD Falcon™), and cultured with DMEM containing 10 % FBS supplemented with 100 units/ml penicillin and 100 mg/ml streptomycin at 37 °C in a humidified atmosphere of 5 % CO2/95 % air. Chondrocytes were grown in monolayer cultures, and were passaged when reaching confluence. Cells at the second passage were used for the assay.

To confirm the paracrine effect of ADSCs for chondrocyte, we used the co-culture system (Fig. 1).

The co-culture groups were established between ADSCs (5 × 10³cells/well) in Cell culture inserts (Cell Culture inserts, 0.4 μm pores, BD Falcon™) and chondrocytes (5 × 10³cells/well) in twelve-well plates (Cell Culture Insert Companion Plates, BD Falcon™) in DMEM. The control group was established between no cells in Cell culture inserts and chondrocytes only in twelve-well plates (5 × 10³cells/well) in DMEM.

After 48 h, cell viability of the chondrocyte group was determined using the MTT assay and compared to that of the co-culture group.

Subsequently, to confirm the paracrine effect of ADSCs for chondrocyte under TNF-α, the culture experiments were divided into 3 groups: chondrocytes (control); chondrocytes with TNF-α (10 ng/mL) (TNF-α); and co-culture of ADSCs and chondrocytes with TNF-α (10 ng/mL) (ADSCs). After 48 h, cell viability of the chondrocytes was determined using the MTT assay and compared among the three groups.

Next, 72 h after TNF-α stimulation, the concentration of MMP-13 in the three groups was measured by ELIZA. The Human Biotrak MMP-13 ELIZA kit (GE Healthcare) was used because these kits were previously shown to be suitable for assaying rabbit MMPs [28, 29].

We believe that quantifying the cellular survival rate enabled us to evaluate the chondrocyte proliferative effects of the paracrine effect of ADSCs, while the MMP-13 quantification allowed us to evaluate the chondrocyte matrix protection effects. TNF-α stimulation induced chondrocyte apoptosis and reproduced the clinical condition of OA.

Statistical analyses were performed using SPSS ver.19.0 (SPSS, Inc, Chicago, Ill). The results are shown as the mean ± standard deviation (SD). The unpaired t-test was used. In all analyses, P < .05 indicates statistical significance.