Giant cell tumor stromal cells: osteoblast lineage-derived cells secrete IL-6 and IL-10 for M2 macrophages polarization

Clinical samples

Three GCT tissue samples were collected from patients who underwent surgical resection for primary GCT in Department of Orthopaedics, Xijing Hospital, the Fourth Military Medical University. The patient’s consent was obtained and informed consent form was signed. The use of all patient-derived materials was approved by the Ethics Committee of the Fourth Military Medical University (IRB-REV-2015005), and patient informed consent was obtained individually.

Cell culture

Primary GCT stromal cells (GCTSCs) were isolated from fresh GCT. The neoplasm was minced in small pieces and digested with 200 U/ml collagenase (Gibco BRL, Karlsruhe, Germany) for 1 h at 37 °C. Cells were collected by centrifugation, washed twice in PBS and cultured with alpha minimum essential medium (α-MEM, Gibco BRL, Gaithersburg, MD, USA) containing 10% FBS (Hangzhou Sijiqing Biological Engineering Materials Co., Ltd. Zhejiang, China) and 100 U/ml penicillin and 100 mg/ml streptomycin (Invitrogen Life Technology, Carlsbad, CA, USA). The cells were kept in culture at 37 °C, 5% CO₂, and the media was changed every 3 days thereafter until cells reached 70% confluency. Adherent GCTSCs were harvested using 0.25% Trypsin-EDTA (Invitrogen Life Technology, Carlsbad, CA, USA). Normal MSCs were obtained from the bone marrow of the normal donor patients as described previously (Soleimani and Nadri 2009). Human monocytes were isolated from the peripheral blood of normal human volunteers (Blood donors in Blood Transfusion Department of Xijing Hospital) using human monocyte isolation kit II (Miltenyi Biotech, Teterow, Germany). Purified CD14+ monocytes were plated into 6-well cell culture plates at a concentration of 1 × 106 per well in RPMI 1,640 media supplemented with 10% FBS.

Animal experiments

Immunocompromised mice (NMRI-nu/nu) were purchased from Beijing Vital River Laboratory, Beijing, China. All mice were housed five per cage under standard laboratory conditions at 12 h light /dark cycles, and 22 °C, with free access to food and water. Mice were anesthetized by intraperitoneal injection of pentobarbital (30 mg/kg) and implanted with cells. After 35 days of implantation, the mice were decapitated and killed, the de novo bone area was taken. Animal experiments in this study was authorized by the Animal Care Committee of Fourth Military Medical University (2016-kq-054), and all experimental protocols were performed with the approval of the Fourth Military Medical University.

Histology and immunohistochemistry staining

The neoplasm specimens were fixed in 4% paraformaldehyde overnight and then embedded in paraffin or in the optimal cutting temperature compound. Hematoxylin and eosin (HE) staining was used for general observation. Surface markers of GCTSCs and macrophages were stained with CD105 and CD14 antibodies (eBiosciences, San Diego, CA, USA) overnight at 4 °C, respectively. When they were washed 3 times, peroxidase conjugated goat anti-rabbit secondary antibody (Dako, Denmark) was added. Envision Detection System kit (DAKO, Denmark) was used after the nuclear stained by hematoxylin. Color development was triggered using Di-Amino Benzidine (DAB, Calbiochem, Germany).

Immunofluorescent staining

BMMSCs and GCTSCs were washed with PBS and fixed with 4% paraformaldehyde for 15 min, and then were incubated with antibodies CD146, CD90, CD73, STRO-1 and SSEA-4, as well as antibodies against hematopoietic cell markers of CD14 and CD34 and CD45 (All from BD Biosciences, San Diego, CA, USA). The positively stained cells were detected by a laser scanning confocal microscope (Olympus FluoView FV 1000, Tokyo, Japan).

Flow cytometry analysis

To confirm mesenchymal stem cells character, we used flow cytometric analysis to show that GCTSCs were positive for CD90, CD105, and CD29, and negative for CD14, CD34 and CD45 (All from BD Biosciences, San Diego, CA, USA). Ki-67 staining was used to confirm the proliferation of macrophages after being co-cultured with GCTSCs. Apoptotic cells were examined using Annexin V-FITC/7AAD Apoptosis Detection Kit (Roche, Basel, Switzerland). Cells were stained with Annexin V-FITC and 7AAD for 15 min, both of which were in the dark at room temperature. After washing by PBS, immunofluorescent-stained cells were analyzed on FACS (Becton-Dickinson) can using Consort 30 software.

BrdU test

BMMSCs and GCTSCs were labeled with 5-bromodeoxyuridine (BrdU, sigma) 10 umol L⁻¹ for 24 h. After the cells were washed three times with PBS and fixed with 4% paraformaldehyde for 15 min, the incorporated BrdU was detected by mouse anti-BrdU monoclonal antibody conjugated with peroxidase (Abcam). The color reaction was parallel with the number of proliferating cells.

Osteogenic and adipogenic differentiation

BMMSCs and GCTSCs were incubated with osteogenic medium (100 nmol L⁻¹ dexamethasone, 50 mg mL⁻¹ ascorbic acid and one mmol L⁻¹ b-glycerophosphate, Sigma) for 21 d according to the manufacturer’s instructions. Cells were fixed with 60% isopropanol and stained with 1% Alizarin Red (Sigma) for osteogenic differentiation assay. BMMSCs and GCTSCs were also cultured with adipogenic medium (0.5 mmol L⁻¹ methylisobutylxanthine, 0.5 mmol L⁻¹ hydrocortisone and 60 mmol L⁻¹ indomethacin, Sigma) for 14 d. Intracellular lipid droplet was detected by staining with Oil Red O solution.

In vivo bone formation assay

5 × 10⁶ cells were combined with 40 mg hydroxyapatite/tricalcium phosphate (HA/TCP) and implanted at the back of immunocompromised mice (NMRI-nu/nu). After 35 days the mice were euthanized and the de novo bone area was stained by HE. The sections were analyzed by Image-Pro Plus software.

Co-cultures

The purified human blood monocytes were indirectly co-cultured with GCTSCs as well as BMMSCs for 14 d. The monocytes (2 ×10⁵ cells) were cultured in 24 well plates and GCTSCs or BMMSCs (2 ×10⁴ cells) were cultured in the cell-culture insert (pore size: 0.4 pm) of the same well (Falcon Inc.).

Antibody-based protein array

The assay was performed following the manufacturer’s instructions (RayBio). Briefly, samples with equal amounts of total proteins were incubated overnight at 4 °C. After washing, the membranes were incubated with Biotinylated Antibody Cocktail followed by second wash. Incubate with HRP-Streptavidin overnight at 4 °C and wash again. Chemiluminescence was used for signal detection, which was analyzed with Quantity One image analysis software (Molecular Imager FX, Bio-Rad Laboratories).

Quantitative RT-PCR

Total RNA was isolated from BMMSCs and GCTSCs by using RNAiso plus (TaKaRa, Tokyo, Japan) according to the manufacturer’s instructions. The mRNA was reversed to complementary DNA, and quantitative real time -PCR (qRT-PCR) detection was carried out by Prime Script TM RT master mix (TaKaRa, RR036A) and SYBR Premix Ex TaqTMII (TaKaRa). A CFX96 Trademark Real-time PCR detection system (Bio-Rad, Richmond, CA, USA) was used for the detection. The forward and reverse primers were used to measure IL-6 (F: AGTGAGGAACAAGCCAGAGC,R: AGCTGCGCAGAATGAGATGA) and IL-10 (F: ACCTGCCTAACATGCTTCGAG, R: TGGGTCTTGGTTCTCAGCTTG) gene expression, and target gene expression levels were normalized to that of GAPDH (F: TGACATCAAGAAGGTGGTGAAGC, R: GGAAGAATGGGAGTTGCTGTTG).

Gene ablation by siRNA

2 × 10⁵ cells/well were seeded in a 6-well culture plate with 2 ml antibiotic- free normal growth medium supplemented with FBS. Incubated the cells at 37 °C in a CO2 incubator until the cells were 60–80% confluent. Then added IL-6 (Santa) and IL-10 (Santa) siRNA duplex and siRNA transfection reagent (Santa) mixture after cells were washed with two mL siRNA transfection medium, then incubate the cells 5–7 h at 37 °C in a CO₂ incubator. one mL of medium containing 2 times FBS and antibiotics was added to cells and incubated for additional 18–24 h before used.

Statistical analysis

GraphPad Prism software (Graph-Pad Software, Inc., USA) was used for statistical analysis. All data are presented as mean ± s.d and the measurements were performed on three different GCT patient samples (n = 3). Unpaired, two-tailed Student’s t tests for comparisons between two groups and one-way analysis of variance (ANOVA) with Bonferroni for multiple comparisons were applied. Each experiment was repeated at least three times. P value < 0.05 was considered to be statistically significant.

CD14 and CD105 cells exist in Gaint cell tumor

We harvested Giant cell tumor (GCT) samples from three patients with two females (46 and 41 years old) and one male (21 years old). The lesions in which were tibia, ilium and both terminal of the tibia and femur. All of these three patients presented a common radiologic feature: a lytic and well-defined lesion without sclerotic margin (Figs. 1A– 1C). HE staining showed GCT consisted of three kinds of cells: multinucleated giant cells, mononuclear histocytic cells and giant cell tumor stromal cells (GCTSCs) with a spindle morphology. Immunohistochemistry staining verified that GCT cells contained CD14 positive and CD105 positive cells, which represented a macrophage-like and mesenchymal-like features respectively (Fig. 1D). Immunofluorescence staining of the GCT indicated STRO-1 positive and CD14 positive cells in the samples and these two kinds of cells were not overlapped (Figs. 1E, 1F).

GCTSCs showed increased proliferation and decreased differentiation characteristics

It had been widely proved that giant cell tumor stromal cells (GCTSCs) were the neoplastic component of GCT and had similar functions with mesenchymal stem cells. We isolated the cells from neoplasm and compared different bio-markers (CD90, CD105, CD29, CD14, CD34, CD45) between three patients’ GCTSCs and normal bone marrow MSCs. Flow cytometer analysis indicated that both GCTSCs and BMMSCs had a similar specific expression of CD90, CD105 and CD29 and negative expression of CD14, CD34 or CD45 (Fig. 2A). To characterize the proliferation property of GCTSCs, we performed serial implantation of GCTSCs. Brdu measurement indicated that all of three GCTSCs had a significantly higher proportion of Brdu-positive cells than BMMSCs (P < 0.01). After subcutaneously implantation of GCTSCs into immunocompromised mice, we found that re-GCTSCs collected from implants still showed higher proliferation than re-BMMSCs. (Figs. 2B, 2C). These results indicated GCTSCs had mesenchymal stem cell like properties and had an obviously higher proliferative activity than BMMSCs.

In order to examine the differentiative capacity of GCTSCs into osteoblast and adipocyte, we induced the GCTSCs with osteogenic and adipogenic induction medium respectively. Alizarin Red staining verified that the total mineralized area in GCTSCs was 0.11 compared with 0.25 in BMMSCs (2.27-fold higher, P = 0.056) (Fig. 2D). Oil-Red O stained cells represented the adipocytes, the proportion of which in GCTSCs was 2.22% compared with 32.44% in BMMSCs (14.61-fold higher, P < 0.01) (Fig. 2E). GCTSCs or BMMSCs combined HA/TCP were subcutaneously implanted into immunocompromised mice, and a clearly bone formation area of BMMSC implantation was founded, the total de novo bone area was about 0.12%, compared with much smaller spots about 0.04% among GCTSC implantation (P < 0.001) (Fig. 2F). These results indicated that GCTSCs indeed had bone and adipose differentiation capacities to some extent in vitro as well as in vivo, but which were much weaker than BMMSCs.

GCTSCs play no role in proliferation or apoptosis of CD14 cells

Many researches had pronounced that GCTSCs played a significant role in giant cells formation, osteolysis as well as aggression, but there were rarely researches to explore the influence of GCTSCs on the proliferation or apoptosis of CD14 cell. Co-culturing macrophage cells with GCTSCs and BMMSCs using transwell for 7 days, 14 days and 21 days respectively and the amount of macrophage cells was counted. The results of different co-culturing period with GCTSCs or BMMSCs exhibited no significance correlation (Fig. 3A). The proportion of Ki67 cells, as an indication of proliferative ability, among normal macrophages and co-cultured with BMMSCs as well as GCTSCs are 3.83%, 2.57% and 4.72% (P > 0.05) (Figs. 3B, 3D), which indicated both GCTSCs and BMMSCs didn’t alter the proliferative ability of CD 14 positive macrophages.

Moreover, we detected the Annexin V+7-ADD+ macrophages after co-cultured with BMMSCs or GCTSCs. BMMSCs had a prominent tendency in inducing macrophage apoptosis, the proportion of which was approximately 3.65% (P = 0.01), however, co-culture with GCTSCs hadn’t resulted in any enhancement of macrophage apoptosis, which were 1.25%, 3.65% and 1% (BMMSCs vs Macrophage, P < 0.05. GCTSCs vs Macrophage, P > 0.05), respectively (Figs. 3C, 3D). These data confirmed that BMMSCs could induce the apoptosis of macrophage but not the GCTSCs. However, the proportion of apoptosis cells induced by BMMSCs was very low.

GCTSCs promote the polarization into M2 phenotype of macrophage via increasing IL-6 and IL-10 release

It had been wildly accepted the polarization of macrophage into M2 phenotype was a contribution for tumor propagation and angiogenesis. To investigate the polarization of macrophages in GCT, we detected the proportion of M2 phenotype macrophages in GCT and peripheral blood respectively, we found a numerous proportion of M2 phenotype macrophages among the neoplasm (85.05%) compared with normal macrophages (61.15%) (P < 0.05). This suggests that a high number of M2 macrophages exist in giant cell tumor of bone. (Fig. 4A). By co-culture the macrophages with GCTSCs as well as BMMSCs, it aroused a significant (11.0%, 11.6%, 19.1%) polarization of M2 macrophages in GCTSCs and 8.6% in BMMSCs (Fig. 4B). These results verified that, compared with BMMSCs, GCTSCs could promote the macrophages into M2 phenotype.

Human cytokine antibody array can semi-quantitative detected 42 human proteins expression using 2-D densitometry. By analyzing the culture supernatant of BMMSCs and GCTSCs, it was clarified that GCTSCs had a distinctive extraction in the spot of IL-6 and IL-10 (Fig. 4C). q-PCR analysis verified that GCTSCs had a much higher expression of IL-6 and IL-10 compared with BMMSCs (P < 0.05) (Fig. 4D). In order to verify the promotion effect of increased GCTSCs IL-6/10 expression on GCT macrophages M2 activation, we used siRNA to reduce the expression of IL-6/10 in GCTSCs, and then co-cultured with THP-1 cells for 48 h. The results of cell flow cytometry showed that knock down the expression of IL-6/10 in GCTSCs could reduce the proportion of THP-1 CD206 positive cells in the co-culture system (P < 0.05) (Fig. 4E). These results illustrated that GCTSCs expressed and secreted a large amount of IL-6 and IL-10, which might induce the polarization into M2 phenotype macrophages.