(J-400) Developing an in vitro miniature joint model to study the biochemical correlations between osteoarthritis and obesity

Osteoarthritis (OA) is a mentally and physically debilitating disease and is the most common chronic joint disorder in the United States¹. Though causes are still unknown, clinical studies indicate a relationship between OA and obesity^2,3. Although originally theorized as due to additional mechanical load on the joint, correlations between obesity and hand OA are also shown, indicating other biochemical contributions⁴. For example, it is suggested that the hypertrophic adipocytes–a phenotype characteristic of obesity–releases numerous pro-inflammatory adipokines⁵ that potentially initiate and progress OA.

Along with articular cartilage destruction, OA progression affects all other joint elements⁶. Each of these pathological changes contributes to OA, suggesting OA is a “joint failure” disease⁷. Our lab has recently developed a three-dimensional (3D), multicomponent microphysiological joint chip (miniJoint) that includes 4 human mesenchymal stem cell (MSC)-derived tissue units: osteo-chondral, adipose, and synovial tissues⁸. Each component can successfully recapitulate the unique phenotype and physiology seen in vivo. The goal of the current study was to adapt the miniJoint model to study how obese-like dysfunctional adipose tissue contributes to pathological changes in the entire joint.

Our lab recently found that sodium palmitate, the salt of palmitic acid, can be introduced into the adipogenic medium (AM) to induce obese-like changes in mesenchymal stem cell-derived adipose tissues. In this study, we added sodium palmitate into AM that fed adipose tissues, and tested whether the “obese” fat caused OA-like phenotypes in the miniJoint.

The separate miniJoint units were generated by ultra-violet photo‐crosslinking of human bone mesenchymal stem cells (hBMSC)‐laden gelatin methacryloyl (GelMA) with 15% lithium phenyl-2,4,6-trimethylbenzoylphosphinate (LAP) in 3D printed resin inserts.The hBMSCs were differentiated into osteochondral, synovial, and adipose tissue units using protocols previously reported by our lab⁵. After 28 days of differentiation separately, the The miniJoint chip was established by integrating differentiated tissues in a 3D printed chamber, with respective mediums flowing through the correct chambers, with a universal, synovium-like medium connecting the different tissue compartments to allow tissue-tissue crosstalk similar to in vivo tissues.

Once combined, 350μM sodium palmitate (PA)-the salt of palmitic acid-was supplemented only into the adipogenic medium to induce obesity-like changes. After 5 weeks of PA exposure, all tissue inserts were collected and analyzed via real time polymerase chain reaction (RT-PCR) and enzyme linked immunosorbent assays (ELISA). Each tissue was assessed for tissue specific genes and inflammatory cytokine expression.

RT-PCR was run to analyze the samples, utilizing Ribosomal Protein L13 alpha (RPL13α). Results displayed that obesity-like changes occurred in the adipose tissue with PA, with the fat displaying a decreased level of serine protease adipsin. Additionally, adiponectin, an anti-inflammatory adipokine was downregulated. Pro-inflammatory cytokine interleukin 6 (IL-6) was increased, as well as leptin, a critical obesity marker.

ELISA results showed that the “synovial fluid-like medium” that connect the tissues in the miniJoint displayed high levels of pro-inflammatory cytokine IL-8, indicating that the obese-like adipose tissue could have effects on the surrounding tissues in a crosstalk-dependent manner.

Further PCR results showed high levels of inflammatory cytokines IL-6 and IL-8 in the synovial fibroblasts. The osteochondral tissue unit also showed changes, with the cartilage showing increased matrix metalloproteinase 13 (MMP13) and a decrease in aggrecan, indicative of cartilage turnover and resembles OA pathophysiology. Moreover, a disintegrin and metalloproteinase with thrombospondin motifs 5 (ATS5) was increased, indicating increased cartilage destruction, phenotypic of OA.

These results indicate a correlation between obese-like adipose tissue and OA-like joint changes in other tissue units.

Introducing PA into the adipogenic medium affected all other units, with an increase in cartilage degradation markers, bone remodeling markers, and pro-inflammatory cytokines. In future studies, different concentrations of PA should be tested and analyzed for their subsequent effects on the rest of the joint tissues. In particular, the medium allowing tissue-tissue crosstalk should be examined for the specific factors causing the OA-like phenotype, bringing us closer to finding a biochemical correlation between OA and obesity.

These findings are significant due to the mounting obesity rates. It is projected that by 2030, 1 in 2 United States citizens will be obese [9]. Therefore, it is critical to understand the role that obesity has within OA, so that obesity-associated OA’s pathogenesis can be understood. Further, the miniJoint provides a unique ability to accurately capture in vivo physiology, elucidating the intricate connections between obesity and OA. Further, the miniJoint can serve as a clinically relevant drug testing platform for obesity-associated OA, bringing us one step closer to alleviating its chronic symptoms for millions worldwide.

This work was carried out with the support and resources at the Lin Laboratory, within the Department of Orthopaedic Surgery at the University of Pittsburgh. This work was supported by the University of Pittsburgh David C. Fredrick Honors College. This work was partially supported by the NIH (UH3TR002136).