Autoactivation profiles of calcium-dependent matrix metalloproteinase-2 and -9 in inflammatory synovial fluid: effect of pyrophosphate and bisphosphonates.

BACKGROUND

The presence of matrix metalloproteinase-2 and -9 (MMP-2, MMP-9), gelatinase A and B, in synovial fluid is typical in inflammatory connective tissue diseases especially rheumatoid arthritis (RA). Because MMPs are synthesized as latent proforms, a pathophysiologic understanding of MMP regulation has focused on mechanisms of activation that remain to date largely unresolved.

METHODS

Synovial fluid was collected by aseptic aspiration from RA patients and incubated with and without physiologic levels of calcium and other modifiers (pyrophosphate, bisphosphonates, and the tissue inhibitors of MMPs (TIMPs), under conditions that activate MMPs. MMP-2 and -9 were then characterized by substrate gel electrophoresis (gelatin zymography) to resolve both latent and activated 'partially proteolyzed' forms.

RESULTS

Gelatin zymography revealed that RA synovial fluid contained latent neutrophil MMP-9 (92, 130, 225 kDa) and fibroblast MMP-2 (72 kDa). A small amount of activated MMP-2 (64 kDa) was also noted. Incubation of synovial fluid without calcium resulted in MMP-9 activation to 87, 116, and 209 kDa forms. MMP-9 activation was, however, substantially delayed in the presence of physiologic calcium (2.5 mmol/l). MMP-2 did not demonstrate any appreciable activation with or without physiologic calcium. MMP-9 activation likely occurred via an autoactivation mechanism since it was susceptible to inhibition by the tissue inhibitor of MMP-9 (TIMP-1). Pyrophosphate and bisphosphonates (alendronate and risedronate) were ineffective in blocking synovial fluid MMP-9 autoactivation. Some early MMP-9 activation was noted with alendronate despite the presence of physiologic calcium.

DISCUSSION

Although RA synovial fluid contained abundant MMP-2 and MMP-9, only MMP-9 underwent autoactivation to lower molecular weight forms. MMP-9 was transiently stable in the presence of physiologic calcium concentration, whereas autoactivation was more pronounced without exogenous calcium. The apparent lack of MMP-2 autoactivation with or without calcium, likely resulted from the coexistence of its bound endogenous inhibitor, TIMP-2. The role of differential autoactivation of MMPs activity in inflammatory arthritic disease is discussed.