Summary
Pharmacologic inhibition of fibroblast activation protein (FAP) rescues the transcriptional and functional profile of pathologic synovial fibroblasts in OA.
Abstract
Background
Osteoarthritis (OA) is a major source of pain and disability. Current treatments are limited to symptom management rather than disease modification. Recently, synovial fibrosis has been recognized as a potential driver of disease. Pro-inflammatory cytokines and extracellular matrix (ECM) stiffening in the synovium activate fibroblast-like synoviocytes (FLS), priming them to differentiate into myofibroblasts. These activated cells secrete additional ECM and inflammatory mediators, creating a vicious cycle of stiffening, inflammation, and degeneration. Fibroblast activation protein (FAP) is expressed in the synovium of patients with OA and contributes to disease progression, but the mechanisms by which it does so are undefined. This study sought to determine whether FAP inhibition can reverse the activated phenotype of FLS in OA.
Methods
Minipigs underwent destabilization of the medial meniscus (DMM). Synovium was harvested at 6 weeks or 6 months and analyzed for FAP and ⍺SMA expression. An in vitro model of FLS activation was established using bovine FLS on 15kPa polyacrylamide (PA) gels or glass. Media was supplemented with TGFβ-1 to induce activation. After 2 days, cells were stained for actin, ⍺SMA, paxillin, and FAP. Cell area and focal adhesions were quantified. Expression of FAP, ACTA2, FN1, CCN2, and PRG4 were assessed by RT-qPCR. FAPi and Fasudil, a Rho/ROCK inhibitor that inhibits actin and other cytoskeletal elements, were added to appropriate groups. After 6 days, immunofluorescence, morphometric analysis, and RT-qPCR were conducted as above. To assess contractility, FLS were seeded in type I collagen gels and percent initial area was quantified over time. FAP and ⍺SMA staining correlation was determined by linear regression. Other outcomes were compared by one- or two-way ANOVA.
Results
FAP is elevated in the synovium after DMM. FAP and ⍺SMA staining intensity were significantly increased in DMM synovium at both 6 weeks (p=0.0024, p=0.0004 respectively) and 6 months (p=0.0298, p<0.0001 respectively) compared to control, with a positive correlation between FAP and ⍺SMA intensity for individual donors (R2= 0.4294, p=0.0002). Stiff substrates and TGFβ activate FLS. Cell spreading increased in FLS on glass relative to 15kPa PA gels (p<0.0001) and trended upward with TGFβ treatment. Expression of ⍺SMA and CTGF were significantly higher on glass (p=0.0004, p=0.0013 respectively) and with TGFβ (p=0.0033, p<0.0001 respectively). FAP, FN, and PRG4 expression increased with TGFβ treatment. FAP inhibition partially reverses activation of FLS: Treatment with FAPi reduced cell spreading and expression of FAP, ⍺SMA, FN, and CTGF, while PRG4 expression remained unchanged. FAP inhibition also reduced the number and area of focal adhesions and slowed the contraction of FLS-embedded collagen gels.
Conclusion
Our findings show that FAP inhibition can rescue the transcriptional and functional profile of activated FLS. In vivo, correlation of FAP and ⍺SMA suggests that more mechano-activated cells express higher levels of FAP. In vitro, FAPi reversed the activated FLS phenotype, and did so to an equivalent or greater extent than Fasudil. These data reveal a role for FAP in FLS mechanobiology, offering a potential mechanism by which this protein contributes to OA, and indicate that FAPi is a promising therapeutic.