Background/Purpose: Rheumatoid arthritis (RA) fibroblast-like synoviocytes (FLS) have joint-specific epigenetic and transcriptome profiles. This is particularly notable for homeobox (HOX) genes, which contribute to joint formation during embryogenesis. Furthermore, HOX gene expression patterns are often maintained in adulthood. In this study we analyzed joint-specific HOX gene regulation in hip- and knee-derived FLS. We then focused on HOXD10, which can regulate cell growth and migration, to understand the mechanism of its joint-specific expression and its potential role in FLS function.

Methods: RA FLS lines were obtained from hip and knee synovial tissues at arthroplasty and used from passages 5 through 8. qPCR was performed to assess gene expression, regulation, and mRNA half-life. HOXD10 was silenced by siRNA (~70% inhibition), and non-targeting siRNA was used as control. mRNA half-life was measured using actinomycin D to stop gene transcription. Cell migration was evaluated using an in vitro wound healing assay. Analysis of HOXD10 epigenetic marks was performed using our public ChIPseq data.

Results: Several HOX genes were differentially expressed in hip and knee join FLS. For example, HOXD10, HOXA13 and HOXD9 mRNA were more highly expressed in knee than hip FLS (85±20-fold, P=0.01; 3.8±1.8-fold, P=0.023; and 7.4±3.2-fold higher in knee, P=0.01, respectively). HOXA11 and HOXD3 mRNA expression were similar in FLS from the two joints. Knee FLS also showed higher expression of noncoding RNAs in HOX loci than hip FLS (HOXA11-AS 4.0±0.6-fold, P=0.003; and HOTAIR 1.7±0.1-fold higher in knee, P=0.02). Because differential HOXD10 expression was the greatest, we focused on this gene for subsequent studies. First, we confirmed that HOXD10 regulates cell migration in PDGF-stimulated FLS using siRNA knockdown (22±4.4% increase in hip FLS with HOXD10 deficiency, P=0.0006, and 15±6.3% increase in knee, P<0.0.05). To study HOXD10 regulation, we stimulated FLS with TNF and noted increased HOXD10 transcripts in both hip and knee FLS. The fold increase in gene expression was greater in hip than knee (7.2±2.2-fold induction in hip and 2.2±0.3-fold induction in knee; P=0.03 for hip vs. knee). However, knee HOXD10 expression remained higher than hip after TNF stimulation. To understand the mechanism of increased mRNA levels in knee FLS, we evaluated the transcript half-life. HOXD10 mRNA half-lives were similar in hip and knee FLS (t_1/2 of 11 and 9 h, respectively, P=0.74). Evaluation of HOXD10 epigenetic marks in the promoter region showed increased H3K27ac and H3K4me3 peaks in knee FLS compared with hip FLS consistent with increased transcription in the former.

Conclusion: HOXD10 expression is constitutively higher in knee FLS than hip FLS and is highly induced by TNF. The mechanism of differential expression is mostly likely related gene transcription rather than mRNA degradation. Embryonic epigenetic imprinting might contribute to differences in joint-specific FLS behavior. HOX genes expression profiles might contribute to the patterns of joint involvement in arthritis and joint-specific responses to therapy.

ACR Meeting Abstracts - https://acrabstracts.org/abstract/mechanism-of-joint-specific-homeobox-d10-regulation-in-rheumatoid-arthritis-fibroblast-like-synoviocytes/