The synthesis of collagenase, gelatinase-A (72 kDa) and -B (95 kDa), and TIMP-1 and -2 by human osteoblasts from normal and arthritic bone

Bone resorption is a complex multistep process that involves removal of both the organic and mineral constituents of bone matrix by proteolytic enzymes synthesized by osteoblasts and osteoclasts. To further understand the role of matrix metalloproteinases (MMPs) and their specific inhibitors TIMPs (tissue inhibitor of metalloproteinases) in this process, human osteoblasts were obtained by sequential enzymatic digestion from samples of bone from normal donors and patients with various forms of arthritis; first passage cells were used in all experiments and cultured on a type I collagen substratum. Collagenase was detected by an ELISA in supernatants from unstimulated osteoblasts (range 12–730 ng/mL), although the levels did not appear to bear any relationship to the age or clinical status of the patient; treatment with parathyroid hormone (PTH; 2 units/mL) and 1,25-dihydroxyvitamin D3[1,25(OH)2D3, 10 ng/mL] had no added effect, but mononuclear cell conditioned medium (MCM; 5% v/v) and interleukin-1α (IL-1α; 1 ng/mL) both stimulated collagenase synthesis, in the case of MCM by two orders of magnitude. TIMP-1 was detected in unstimulated cultures by an ELISA (range 320–590 ng/mL), the mean level being three-fold greater than for collagenase and was stimulated by 1,25(OH)2D3, and MCM treatment. Degradation studies showed that, over a 120 h culture period, one third of the collagen substratum was degraded by unstimulated cells. PTH and 1,25(OH)2D3, had no effect on this endogenous level of lysis, but addition of MCM and IL-1α resulted in a significant increase in collagen degradation. Zymography using SDS/PAGE containing gelatin provided evidence for the presence of progelatinase-A (72 kDa) and progelatinase-B (95 kD) in both treated and untreated cultures, with the active species of gelatinase-A generated in all cases. Inhibitor gel zymography confirmed that TIMP-1 was constitutively synthesized, but we could find no evidence for TIMP-2 using this method; immunocytochemistry was subsequently used to detect the presence of TIMP-2 in osteoblast monolayers. These findings provide additional evidence that human osteoblasts can synthesize collagenase and other MMPs, and support the hypothesis that osteoblast-derived MMPs play an important role in bone resorption, particularly in removal of the surface osteoid that precedes osteoclast attachment.