Background Long noncoding RNAs (lncRNA, >200nt) influence gene expression through their interaction with transcription factors and/or epigenetic enzymes. We have shown that the histone methyltransferase EZH2 is upregulated in rheumatoid arthritis (RA) synovial fibroblasts (SF) and is induced by TNFα. EZH2 mediates epigenetic repression by the trimethylation of histone 3 lysine 27 (H3K27me3) at target gene promoters. The lncRNA HOX transcript antisense RNA (HOTAIR) is overexpressed in various human malignancies, interacts with EZH2 and other epigenetic enzymes, and was shown to regulate a subset of EZH2 target genes specifically involved in cellular invasiveness (Wang KC and Chang HY, Mol Cell, 2011).
Objectives To study expression, regulation and function of HOTAIR in SF.
Methods SYBR Green or TaqMan real-time PCR and enzyme-linked immunosorbent assays (ELISA) were used to measure gene expression. For chromatin immunoprecipitation (ChIP), antibodies against histone 3 (H3), H3K4me3, H3K27me3, and H3 acetylation (Ac) were applied. Osteoarthritis (OA)SF were transfected with an EZH2-encoding vector or siRNA targeting HOTAIR and/or were stimulated with TNFα (10ng/ml; n=11) and IL-1β (1ng/ml; n=5) for 24h or 48h. NF-κB activity was measured by reporter gene assay.
Results In RASF (n=9), the expression of HOTAIR was strongly reduced (by 92%) in comparison to OASF (n=13) (ΔCt 14.3±3.3 vs. 10.6±1.7, p=0.005). In OASF transfected with EZH2 (n=4), HOTAIR expression was decreased by 30±17% (p<0.05) suggesting that HOTAIR may be regulated by EZH2-mediated methylation of H3K27 in SF. ChIP analysis indeed revealed elevated levels of this repressive mark at the promoter of HOTAIR in RASF (H3K27me3:H3 ratio 0.51±0.35, n=16) as compared with OASF (H3K27me3:H3 ratio 0.25±0.23, n=12; p=0.0309). Thus, HOTAIR expression levels in SF were inversely correlated with the presence of H3K27me3 at its promoter (Spearman R=-0.8725, p<0.0001). Of interest, HOTAIR promoter-associated H3K4me3 or H3-Ac was neither detected in RASF nor in OASF. Stimulation of OASF with TNFα or IL-1β reduced the expression of HOTAIR by 54% up to 63% (p<0.0005). Silencing of HOTAIR in OASF significantly increased mRNA levels of the matrix metalloproteases (MMP) 1, 3, 13 and 14 (by 2.4±0.9-fold, 1.8±0.7-fold, 4.2±2.9-fold and 2±0.7-fold, p<0.005). The TNFα-induced expression of MMP1 (51.5±23.6-fold), MMP3 (49.9±19-fold) and MMP13 (10.6±5.5-fold) was further increased by silencing of HOTAIR (87.2±37.4-fold, 67.7±23.9-fold and 22±9.2-fold, p<0.005) whereas the IL-1β-stimulated expression was not significantly changed (n=8 each). Data for MMP1 were confirmed at the protein level measuring secreted MMP1 in cell culture supernatants (p<0.05). Furthermore, silencing of HOTAIR increased the constitutive activity of NF-κB signalling in OASF by 26±21% (n=5, p<0.05).
Conclusions We demonstrate that the expression of the lncRNA HOTAIR is repressed in RASF via epigenetic mechanisms. Silencing of HOTAIR was associated with the upregulation of matrix-degrading enzymes and NF-κB activity thus implying the involvement of HOTAIR repression in the activated phenotype of RASF.
Acknowledgements Masterswitch FP7, IMI BTCure, IAR Epalinges
Disclosure of Interest None Declared