Exposure Of Chondrocytes In Vitro With Microbial Dna Amplified From Human Osteoarthritis Cartilage Induces Widespread Oa-Associated Dna Methylation Changes

Purpose: In recent years, strong links between epigenetic changes, particularly alterations in DNA methylation, have been linked with the onset and progression of knee osteoarthritis (OA); however, the environmental factors driving epigenetic changes within articular tissues have not yet been fully elucidated. We have previously demonstrated that the microbial DNA signatures exist within human cartilage, and that changes in microbial signatures correlate with primary knee OA progression. In this study, we hypothesized that alterations in microbial DNA may induce epigenetic changes within articular cartilage. To test this, we exposed chondrocyte cell lines in vitro to amplified microbial DNA from human OA patients and controls. Methods: Matched discarded cartilage samples from macroscopically eroded (n=4) and intact (n=4) OA specimens were obtained from patients undergoing total knee arthroplasty. Control cartilage samples were obtained from cadaveric control patients without a history of arthritis via the National Disease Research Interchange (NDRI). Cartilage was flash-frozen in nitrogen, cryogenically ground, and DNA was isolated and pooled within the sample groups. Microbial DNA was isolated from human DNA using an anti-MBD2 magnetic bead approach and microbial DNA subjected to whole-genome amplification. Residual LPS was removed from all samples using LPS cleanup columns, and an absence of LPS was confirmed via ELISA assays in amplified DNA. Tc28a2 human chondrocyte cells were distributed in a 24-well plate (8 wells per group), and 20μg of amplified microbial DNA was added to each well. After a 3-day incubation period, DNA was extracted, treated with sodium bisulfite, and loaded onto Illumina Infinium Methylation EPIC chips which quantify genome-wide DNA methylation at >850,000 CpG sites across the genome. DNA methylation beta values were extracted from raw chip imaging and processed in R using the ChAMP package. CpG sites within known SNPs and those on sex chromosomes were excluded from the analysis. CpG sites were considered significantly differentially methylated if FDR-corrected (Benjamini-Hochberg) group p≤0.01 with absolute DNA methylation difference between groups (delta-beta) ≥10% criteria were met. Genes associated with differentially methylated CpG sites were evaluated ontologically using the Ingenuity Pathway Analysis (IPA) software package. Results: Microbial DNA treatment induced substantial changes in DNA methylation levels. Comparing eroded-OA microbial DNA-treated cells with control microbial DNA-treated cells, we found 1638 differentially methylated CpG sites (DMPs), 1637 hypomethylated in the eroded group, corresponding to 1109 unique genes. No DNA methylation differences were induced when comparing non-OA with intact-OA microbial DNA-treated cells. However, substantial differences were seen in the eroded-OA analyses. Comparing eroded-OA to intact-OA, we identified 1275 DMPs, 1274 hypomethylated in the eroded group, corresponding to 881 unique genes. Roughly one-half (630) of DMPs were shared among these two groups, corresponding to 464 unique genes. Ontology analysis of differentially methylated genes associated with eroded-OA vs. control microbial DNA treatment revealed numerous OA-related pathways and upstream regulators (Figure 1). These include axonal guidance signaling, Rho family GTPases, ERK/MAPK signaling, stem cell pluripotency, and ErbB signaling, among others. Upstream analysis included enrichment in ERG, miR-137 (which regulates ADAMTS5), miR-21 (which targets Gdf5), miR-16 (which targets Smad3), among others. Of note, the only hypermethylated gene we identified with eroded-OA microbial DNA treatment was KLF3, a cartilage repair, and proliferation factor. Conclusions: Herein, we demonstrate that exposure of human chondrocytes in vitro to microbial DNA is sufficient to induce widespread epigenetic changes. OA-related epigenetic modifications are identified when cells are treated with microbial DNA extracted from eroded-OA human cartilage samples. These data represent a novel potential mechanism of microbiome-epigenome interaction in the development of OA. Future work should focus on testing whether these findings are replicable in vivo and explore the mechanism(s) through which this epigenetic alteration is induced.