Background/Purpose: Osteoarthritis (OA) is both the most common form of arthritis and a leading cause of chronic disability. Bacterial products, including lipopolysaccharide (LPS), have been suggested as drivers of OA; however, an examination of the microbiota signature of human articular cartilage has not yet been performed.

Methods: Using sterile technique, OA-eroded and OA-intact cartilage was obtained from primary OA knee (intact=21, eroded=21) and hip (intact=33, eroded=34) replacements, OA-free cadaveric control tissue was obtained from the National Disease Research Interchange (n=30). Age, race, sex, and BMI were not statistically significantly different among groups. 16s rRNA gene deep sequencing was done on an Illumina HiSeq. Operational taxonomic units (OTUs) were assigned using QIIME 1.9 and alpha and beta diversity were calculated. Group OTU differences were identified using Linear Discriminant Analysis Effect Size (LEfSe); LDA log-scores ≥ 2 or ≤ -2 (p≤0.01) significant, Benjamini-Hochberg FDR q-values were calculated (q≤0.1 significant). Differences were confirmed with clade-specific qPCR in a separate confirmation cohort (10 eroded, 10 intact, 10 control samples). The Gram status of DNA sequences for each sample was determined and compared.

Results: We found reduced alpha diversity in OA tissue vs. control (control vs. eroded p=0.009, control vs. intact p=0.006, eroded vs. intact p=1.0) and in hip vs. knee (knee-control vs. hip-control p<0.0001, knee-OA vs. hip-OA p<0.0001). Beta diversity was also different between groups (OA vs. control p=0.001, knee-OA vs. hip-OA p=0.001, knee-control vs. hip-control p=0.002). In group analysis, there were 63 clade differences between OA and control, 35 passed FDR correction; OA was dominated by Betaproteobacteria whereas controls characterized by Actinobacteria and Clostridia. We found 36 differences in knee-OA vs. hip-OA, 8 passing FDR correction. Hip-OA was dominated by Proteobacteria including Rhodocyclaceae, whereas knees were characterized by Actinobacteria. 46 differences were seen in knee-control vs. hip-control. Deep sequencing findings were confirmed by clade-specific qPCR analysis of Alphaproteobacteria, Bacteroides, Betaproteobacteria, Pseudomonas, and Burkholderiales in a separate confirmation cohort. We noted a shift in OA towards Gram-negative constituent DNA (OA-intact: 37%±3% vs. control: 27%±2%, p=0.03; OA-eroded 38%±3% vs. control 27%±2%, p=0.02, OA-intact vs. OA-eroded N/S).

Conclusion: Our results suggest that bacterial nucleic acid is present in human cartilage samples, and significant diversity and clade differences exist between OA and control tissue, and between hip and knee samples. We identified an increase in nucleic acid from Gram-negative organisms in OA samples, complementary to recent reports of elevated local LPS levels in OA. Further work should be done to confirm our findings and examine the pathogenic role of cartilage microbiota changes in OA.

ACR Meeting Abstracts - https://acrabstracts.org/abstract/identification-of-a-human-cartilage-microbial-dna-signature-and-characterization-of-distinct-microbiome-profiles-associated-with-osteoarthritis/