The FJ synovium samples (n=5) used in this study were provided by patients whose FJs were diagnosed as hyperplastic and hydrarthrotic during preoperative imaging examinations. The study population comprised two men and three women (age, 54–73 years; mean age, 65.6 years) who underwent lumbar surgery at the L4/5 level for LSS at Kumamoto University Hospital between September 2015 and March 2016. The study was approved by the Kumamoto University Ethics Committee (no., 1303), and written informed consent was obtained from all the patients. The trial complied with the principles in the Declaration of Helsinki in 1975.

FJ synovium samples were fixed in 4% paraformaldehyde (PFA), embedded in paraffin, and sliced into 4-μm-thick sections. After pretreatment of the sections with Target Retrieval Solution (pH 9; Dako, Glostrup, Denmark), endogenous peroxidases were blocked using 3% hydrogen peroxide in methanol for 20 minutes. The following primary antibodies were used: anti-human ANGPTL2 (rabbit polyclonal, 1:100, 12316-1-AP; Proteintech, Chicago, IL, USA), anti-vimentin (mouse monoclonal, 1:100, ab8069; Abcam, Cambridge, MA, USA), anti-CD68 (mouse monoclonal, 1:100, ab31630; Abcam), anti-CD15 (mouse monoclonal, 1:50, ab188610; Abcam), anti-CD45 (rabbit polyclonal, 1:100, ab10558; Abcam), anti-IL-6 (rabbit polyclonal, 1:400, ab6672; Abcam), and anti-monocyte chemoattractant protein-1 (MCP-1, rabbit polyclonal, 1:100, ab9669; Abcam). Following treatment with a peroxidase-labeled secondary antibody (Histofine Simple Stain Max PO; Nichirei, Tokyo, Japan), immunoreactivity was visualized using histofine 3,3’-diaminobenzine (Dojindo Molecular Technologies, Kumamoto, Japan). The sections were counterstained with hematoxylin and analyzed using an Olympus BX51 light microscope (Olympus, Tokyo, Japan). For double immunofluorescence staining, anti-ANGPTL2 (1:200), anti-vimentin (1:200), and anti-CD68 (1:200) were used as the primary antibodies, and Alexa Fluor 488-labeled anti-rabbit immunoglobulin G (IgG, 1:500, ab150077; Abcam) and Alexa Fluor 594-labeled anti-mouse IgG (1:500, ab150116; Abcam) were used as secondary antibodies. The nuclei were counterstained with 4’,6’-diamidino-2-phenylindole (DAPI).

FJ tissue samples that were harvested from LSS patients were washed in physiological saline, minced, and incubated for 1 hour at 37°C in Dulbecco’s modified Eagle’s medium (DMEM; Gibco Life Technologies, Carlsbad, CA, USA) that contained 0.2% type I collagenase (Gibco Life Technologies) and 1% penicillin-streptomycin (Gibco Life Technologies). The suspension was filtered using a 100-μm mesh cell strainer (Becton Dickinson and Co., Franklin Lakes, NJ, USA), and the cells were seeded into the wells of a 6-well plate (Becton Dickinson and Co.) containing DMEM with 10% fetal bovine serum (FBS, Gibco Life Technologies) and 1% penicillin-streptomycin. The subsequent experiments were conducted using cells from primary to the third passage.

FJ synoviocytes were seeded into a silicone chamber (STB-CH-10; Strex Inc., Osaka, Japan) at a density of 1×10⁵ cells/chamber. The chamber was attached to a stretching apparatus (STB-140, Strex Inc.), and a cyclic uniaxial stretch (10% elongation) was applied for 6 hours or 12 hours (10 cycles/min; 37°C, 5% CO₂) [17]. After stimulation via stretching, total RNA was isolated from the cells and reverse-transcribed for analyzing ANGPTL2 expression using polymerase chain reaction (PCR). The relative abundance of target transcripts was normalized to the expression of 18S ribosomal RNA (18S rRNA). The primers used for this analysis were as follows: ANGPTL2 forward (5'-GCCACCAAGTGTCAGCCTCA-3') and reverse (5'-TGGACAGTACCAAACATCCAACATC-3') and 18S rRNA forward (5'-TTTGCGAGTACTCAACACCAACATC-3') and reverse (5'-GAGCATATCTTCGGCCCACAC-3'). For the analysis of ANGPTL2 protein, subconfluent FJ synoviocytes cultured in a silicone chamber (STB-CH-10) were washed with phosphate-buffered saline (PBS), and the medium was changed to serum-free DMEM. After 24 hours of stimulation (10% elongation, 10 cycles/min; 37°C, 5% CO₂), the medium was harvested. The secreted ANGPTL2 protein was measured using an ANGPTL2 enzyme-linked immunosorbent assay (ELISA) kit (IBL, Fujioka, Japan) as per the manufacturer’s instructions.

For immunofluorescent staining, FJ synoviocytes were seeded onto a 4-well culture slide (Becton Dickinson and Co.), cultured subconfluently, treated with recombinant ANGPTL2 protein (5 μg/mL), and incubated for 1 hour (37°C, 5% CO₂). The cells were then fixed with 4% PFA and treated with anti-nuclear factor-κB (NF-κB) p65 antibody (rabbit polyclonal, 1:100, sc-372; Santa Cruz Biotechnology, Santa Cruz, CA, USA). Alexa Fluor 488-labeled anti-rabbit IgG (1:500, ab150077; Abcam) was applied as the secondary antibody, and DAPI was used for nuclear staining.

ANGPTL2 protein (5 μg/mL) was added to the wells of a 12-well plate (Becton Dickinson and Co.) containing subconfluent FJ synoviocytes with or without the NF-κB inhibitor BAY 11-7082 (10 μM; Wako Pure Chemical Industries, Osaka, Japan). DMEM, containing 10% FBS and 1% penicillin-streptomycin, was added to each well, and the cells were incubated for 6 hours before RNA extraction. The IL6 mRNA expression was evaluated using real-time quantitative reverse-transcription (qRT)-PCR. The relative abundance of target transcripts was normalized to the expression of 18S rRNA. The IL6 primers were as follows: forward, (5'-AAGCCAGAGCTGTGCAGATGAGTA-3') and reverse, (5'-TGTCCTGCAGCCACTGGTTC-3'). In order to analyze the IL-6 protein expression after ANGPTL2 administration, subconfluent FJ synoviocytes cultured in a 6-well plate were washed with PBS, and the medium was changed to serum-free DMEM. ANGPTL2 (5 μg/mL) was added to each well, plates were incubated for 24 hours (37°C, 5% CO₂), and the amount of secreted IL-6 protein was measured using IL-6 ELISA kit (R&D Systems, Minneapolis, MN, USA).

To evaluate the MCP-1 response to inflammatory stimuli, recombinant IL-6 protein (200 ng/mL; Wako Pure Chemical Industries, Osaka, Japan) with the same amount of soluble IL-6 receptor α (sIL-6Rα; Wako Pure Chemical Industries) was first added to a 12-well plate (Becton Dickinson and Co.) to culture the synoviocytes. Thereafter, ANGPTL2 protein (5 μg/mL) was added to the same kind of plate (Becton Dickinson and Co.) containing subconfluent FJ synoviocytes with or without sIL-6Rα (200 ng/mL). Each plate was incubated for 6 hours before RNA extraction. MCP-1 mRNA expression was evaluated with qRT–PCR. The relative abundance of target transcripts was normalized to the expression of 18S rRNA. The MCP-1 primers were as follows: forward, (5'-CAAACTGAAGCTCGCACTCTCGCC-3') and reverse, (5'-CTCCTAATGTCACGCACGATTT-3').

Data are expressed as the mean±standard error of the mean values. Student t-tests were used for performing comparisons between the two groups in in vitro studies. All p-values <0.05 were considered to indicate significance. For statistical analyses, we performed post-hoc analysis using G*Power (ver. 3.1; Heinrich-Heine-Universität Düsseldorf, Düsseldorf, Germany; http://www.gpower.hhu.de/) and confirmed the statistical validity (power>0.8) for all tests [19].

A large number of ANGPTL2-expressing cells were identified in the hyperplastic FJ tissues in all LSS patients (Fig. 1A). Thereafter, the cell types present in the degenerative FJ tissues were examined with immunohistochemical analysis using antibodies against vimentin (a mesenchymal cell marker), CD68 (a macrophage marker), CD15 (a granulocyte marker), and CD45 (a lymphocyte marker). Patient cells expressed vimentin and CD68 at high levels, while the expression of CD15 and CD45 was very low, suggesting that hyperplastic FJ tissue mainly includes fibroblast-like synoviocytes and macrophage-like synoviocytes (Fig. 1B). Immunofluorescent double staining with anti-ANGPTL2 and anti-vimentin or anti-CD68 antibodies showed that ANGPTL2 was partially expressed by vimentin-positive cells and CD68-positive cells. This suggests that ANGPTL2 is produced by both fibroblast-like and macrophage-like synoviocytes in the FJ tissues (Fig. 1C). IL-6 was also expressed in the cells of the hyperplastic FJ tissues (Fig. 1D), suggesting that IL-6 and ANGPTL2 are highly expressed in the degenerative FJ synovium.

In order to determine whether mechanical stress directly induces ANGPTL2 expression, we subjected FJ synoviocytes isolated from LSS patients to repeated cyclic mechanical stretching, as described above. Stretching stimulation increased ANGPTL2 mRNA expression in the FJ synoviocytes and ANGPTL2 protein secretion in the culture medium (Fig. 2A, B). These findings suggest that ANGPTL2 production via mechanical stress could contribute to the pathological process underlying FJ degeneration.

ANGPTL2 promotes the nuclear translocation of NF-κB and the subsequent induction of IL-6 [16,18]; therefore, we investigated whether ANGPTL2 activates the nuclear translocation of NF-κB and induces IL6 expression in FJ synoviocytes. We confirmed that NF-κB nuclear translocation in FJ synoviocytes was induced by ANGPTL2 treatment (Fig. 3). Further, ANGPTL2 upregulated the IL6 mRNA expression and IL-6 protein secretion in FJ synoviocytes. In contrast, NF-κB inhibition via treatment with BAY11-7085 abrogated the elevation of IL-6 at the mRNA and protein levels (Fig. 4A, B). These results demonstrate that ANGPTL2 induces IL-6 expression and secretion via the nuclear translocation of NF-κB in FJ synoviocytes.

As the combination of IL-6 with sIL-6Rα increases the MCP-1 secretion of the stromal cells, favoring a transition from neutrophil to monocyte/macrophage accumulation and formulating chronic inflammation at the site of injury [13], we attempted to determine whether MCP-1 expresses in the LSS-FJ tissues and whether ANGPTL2-induced IL-6 upregulates the MCP-1 expression in FJ synoviocytes. As expected, several MCP-1-expressing cells were detected in the hyperplastic FJ tissues (Fig. 5A), and the expression of MCP-1 in FJ synoviocytes was elevated due to IL-6/sIL-6Rα administration (Fig. 5B). Only Angptl2 protein with sIL-6Rα increased the MCP-1 mRNA expression (Fig. 5C). These findings suggest that ANGPTL2-induced IL-6 combined with sIL-6Rα can raise the MCP-1 secretion, leading to macrophage recruitment in FJ synovium and subsequent chronic inflammation.