Functional evaluation of ex vivo polarized monocytes/macrophages in a human joint model of osteoarthritis

Chan, Mable Wing Yan 1, 2; Gómez-Aristizábal, Alejandro 2; Gandhi, Rajiv 2; Marshall, Wayne 2; Mahomed, Nizar 2; Viswanathan, Sowmya 1, 2, 3

1. Institute of Biomaterials and Biomedical Engineering, University of Toronto; 
2. Arthritis Program, Krembil Research Institute, University Health Network; 
3. Cell Therapy Program, University Health Network

Background: Osteoarthritis (OA) is a degenerative joint disease of increasing incidence and no standard therapies that impede worsening. Chronic joint inflammation plays an central role in OA: synovial inflammation precedes radiographic OA and correlates with progression.  Our lab has identified inflammatory monocytes as the major immune infiltrators of OA joints. Monocytes/macrophages (Mφs, as a heterogeneously differentiated population) mediate both inflammation and homeostatic tissue repair; their dysregulation is implicated in many inflammatory and degenerative diseases. We are investigating the role of polarized inflammatory (M1) versus homeostatic (M2) Mφs within OA and propose that Mφ population modulation via ex vivo polarized Mφ-based cell therapy will serve as a multimodal treatment that reduces both inflammation and matrix degradation. 

Hypothesis: Addition of ex vivo polarized homeostatic Mφs into a cartilage and synovium human OA explant model will reduce synovial inflammation and cartilage degradation. 

Methods: Our ex vivo human OA joint explant model was developed using cartilage and synovial tissue from end-stage OA total knee replacements. Explant culture changes were evaluated by: gene expression (qRT-PCR, fold-change by ΔΔCT), protein expression (immunoassay), extracellular matrix (ECM) histology (Safranin-O), and ECM loss (DMMB proteoglycan content in medium). CD14+ Mφs were isolated from peripheral blood and ex vivo polarized using standard 48h cytokine protocols into M1/M(IFN-γ+LPS) and M2/M(IL-10+TGF-β). After polarization, Mφs, which are stable even when challenged, were added to OA explants for up to 7 days. 

Results: Our human OA joint explant model was validated through negative and positive inflammatory (OSM+IL-1β for cartilage; IFN-γ for synovium) controls (N=4) that demonstrated reproducible changes in gene expression and ECM loss that mimicked OA inflammation and degradation. Polarized M1 Mφs in OA explants upregulated cartilage catabolic (MMP1), inflammatory (INOS, IL6) and chemotactic (CCL2, CXCL8) genes while downregulating anabolic ECM genes (COL2A1, ACAN, PRG4, TIMP1), whereas M2 Mφs upregulated anabolic ECM genes (PRG4, TIMP1, COL1A1, COLX) only (N=3). Synovium gene expression is variable, but M1 Mφ treatment trended to upregulate inflammatory and chemotactic genes (N=3). Protease inhibitors TIMP-1 and TIMP-2 levels in medium are higher in the M2 Mφ treatment group and lower in the M1 Mφ group (N=3). 

Conclusion: Ex vivo polarized inflammatory versus homeostatic Mφs differentially modulate the OA joint inflammatory and degradative environment; inflammatory Mφs drive inflammatory cytokine upregulation and ECM downregulation, whereas homeostatic Mφs upregulate expression of matrix components. Our human OA joint explant is a clinically relevant tool to reproducibly demonstrate this differential role of Mφs. We have proof-of-concept data that supports further investigation of ex vivo polarized Mφ treatment for OA.