Background/Purpose: Given the complexity and heterogeneous nature of rheumatoid arthritis (RA), it is unlikely that a single biomarker may provide sufficient discrimination; therefore biomarker signatures may represent more realistic approach for the future of personalized medicine in RA. These signatures may represent different subsets in this heterogeneous disease. The study of metabolomics represents a new approach to develop biomarker signatures. Given that metabolism and metabolome are similar between species, biomarkers field might benefit from animal models studies in which pathogenic mechanisms can be controlled, and correlations with joint metabolome can be evaluated.

Methods: Ankles from 10 weeks old K/BxN mice (an IL-1b-driven mouse model of arthritis; n=6) and from 10 weeks old TNFARE mice (a TNF-driven mouse model of arthritis; n=6) and its corresponding littermates, were snap frozen, pulverized and then homogenized using a biphasic chloroform/methanol/water extraction method for polar and lipid metabolite analysis.  The phases were separated, dried, and then reconstituted before injection and analysis on a 5600 Triple-TOF LC-MS (MS; AB Sciex).  Peak integration, alignment, and statistical analysis were performed using Markerview software (AB Sciex).

Results: Multivariate statistical analysis of the MS spectra successfully discriminated between inflamed joints samples and controls. Several metabolites were upregulated in both mouse models compared to their respective littermates.  Among them, metabolites in purine (AMP, GMP, xanthosine and deoxyinosine), pyrimidine (uridine and thymidine), TCA cycle (malate, fumarate and citrate), and phospholipid (lysophosphatidylcholine) pathways were statistically increased in joints from both murine models, which suggest that there may be common metabolic perturbations in these disease models.  However, some metabolites were upregulated uniquely in the K/BxN model such as metabolites in phenylalanine/tyrosine metabolism, branched chain amino acid metabolism (leucine, isoleucine and valine) and eicosanoid metabolism (15-HEPE, 13-HOTrE, 12-oxo-ETE).

Conclusion: The differences between up-regulated metabolites in both murine models suggest different metabolic signatures secondary to different pathogenic pathways. More metabolic profiles studies using controlled animal models are needed to determine if metabolic signatures can distinguish subsets of RA patients and help in diagnoses, prognosis and predicting drug choice in this disease. 

« Back to 2015 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/metabolomic-profiling-of-joints-in-murine-models-of-inflammatory-arthritis/