Integrated approach of meta-analysis and bioinformatics towards elucidating disease mechanisms associated with age-related knee osteoarthritis

      Purpose: Though age-related knee osteoarthritis (KOA) is a highly prevalent and debilitating pathology, the development of disease-modifying treatments has been unacceptably slow. The lack of disease-modifying treatments for KOA is attributed, at least in part, to an incomplete understanding of the cellular and molecular mechanisms initiating and driving KOA. Although research aiming to identify molecular mechanisms of KOA is abundant, the majority of animal studies focus on individual proteins or pathways, thereby precluding a holistic understanding of disease pathogenesis. In addition, while transparent reporting with scientific rigor is an essential criterion for any research endeavor, it is frequently overlooked. With this in mind, the goals of this systematic review were (1) to summarize the state-of-the-science regarding age-related structural and molecular changes in articular cartilage and (2) to evaluate the methodological rigor of research using age-related KOA murine models. This study focused on age-related KOA because this accounts for 88% of KOA cases in the US, and it has a distinct disease trajectory as well as unique cellular and molecular mechanism when compared to post-traumatic KOA. Here, we used an integrated approach of meta-analysis and bioinformatics to provide a comprehensive and mechanistic view of aging-associated KOA pathogenesis.
      Methods: A literature search was conducted in PubMed, PEDro, CINAHL, and Cochrane CENTRAL through March 2020 to identify articles that evaluated cellular and molecular changes associated with cartilage aging in murine knee joint. Thie review excluded post-traumatic KOA murine model with the goal of following the development of age-related KOA from young to middle-aged and finally to aged mice. We then performed a meta-analysis on histological studies and characterized the course of cartilage degeneration, defined as progressive loss of cartilage extracellular matrix over time. To probe transcript level changes across aging, we accessed archived RNAseq data and performed gene-set enrichment analysis (GSEA) among groups of young, middle-aged, and aged. We used GO terms (GO_Biological_Process_2018) as a gene set downloaded from enrichr software. REVIGO software was applied to summarize redundant GO terms identified by the GSEA and visualize the summarized Results. To integrate protein data isolated from each included study, we constructed the protein-protein-interaction network using the STRINGdb software. To evaluate scientific rigor in pre-clinical models of age-related KOA, two independent reviewers assessed reporting quality (ARRIVE), risk of bias, and publication bias in a blinded manner.
      Results: A systematic search identified a total of 1,222 articles from electronic databases relating to KOA and aging, of which 40 met inclusion criteria for this systematic review. Meta-analysis of histological investigations showed that murine models recapitulate the clinical onset of cartilage degeneration starting in middle-age (Fig. blue panel). GSEA from RNAseq data revealed a positive relationship between aging and level of inflammatory cytokines (Fig. green panel) and an inverse relationship between aging and cartilage matrix synthesis. These findings are also supported by the protein data of elevated inflammation, impaired autophagy, and cellular senescence. Among these processes, Advanced Glycation End-Product (AGE)/Receptor for AGE (RAGE) signaling emerged as a strongly associated pathway with age-related KOA (Fig. brown panel). Whereas there is an encouraging trend towards increasing methodological rigor in recent years, we identified numerous opportunities to further enhance the impact of research in KOA. Such opportunities include recommendations for steps to handle missing information, inclusion of investigator blinding, and a priori sample size calculation. Together, these steps have the potential to significantly enhance the reproducibility and ultimately, translatability of findings. Furthermore, we found that no study to date has thoroughly considered whether and how sex alters the trajectory of KOA in pre-clinical models (Fig. red panel). This is an important line of investigation, especially given that women tend to have more severe KOA.
      Conclusions: According to both histological observation and transcriptomic analysis, our findings suggest that the pathogenesis of age-related KOA begins much earlier than is generally appreciated. The disease onset at middle-aged mice is consistent with clinical studies illustrating cartilage abnormalities in middle-aged people, which validates the naturally occurring model as being human-relevant. As such, when designing disease modifying treatments, future studies should bear in mind the trajectory of KOA. Our Results further revealed that the AGE-RAGE signaling pathway may be involved in age-related cartilage degeneration, and future studies into this pathway are worthwhile. Elucidating the role of AGE-RAGE axis in age-related KOA may contribute to accelerate the development of disease-modifying treatments for KOA. In light of a pervasive lack of methodological rigor, this systematic review presents opportunities and recommendations for improvements for transparent reporting and specifically, consideration of sex as a biological variable. Taken together, these steps will accelerate the translation of findings in age-related KOA murine models into effective interventions for an aging population.