Tetracyclines have recently been shown to inhibit the activity of some but not all mammalian matrix metalloproteinases believed to mediate periodontal destruction. However, the specificity of this effect, which could have significant therapeutic implications for different periodontal diseases, has not been examined in detail. Doxycycline and 4-de-dimethylaminotetracycline (CMT-1) have been tested in vitro for their ability to inhibit human neutrophil and fibroblast interstitial collagenases and collagenase in human gingival crevicular fluid (GCF). The GCF samples were obtained from systemically healthy and insulin-dependent diabetic adult periodontitis patients and from localized juvenile periodontitis (LJP) patients. The concentrations of these 2 tetracyclines required to inhibit 50% of the collagenase activity (IC50) were found to be 15 to 30 microM for human neutrophil collagenase and for collagenase in GCF of systemically healthy and diabetic adult periodontitis patients. These concentrations approximate the tetracycline levels observed in vivo during treatment with these drugs. In contrast, human fibroblast collagenase and GCF collagenase from LJP patients were both relatively resistant to tetracycline inhibition; the IC50 for doxycycline and CMT-1 for these 2 sources of collagenase were 280 and 500 microM, respectively. Based on these and other findings, we propose the following: 1) that systemic levels of tetracycline may inhibit connective tissue breakdown by inhibiting neutrophil collagenase; 2) that tetracyclines do not inhibit fibroblast-type collagenase, which may help explain their lack of effect on normal connective tissue remodeling; 3) that tetracycline inhibition of collagenases may serve to identify the cellular origin of the enzyme; and 4) that tetracyclines can also prevent the oxidative activation of latent human procollagenases.(ABSTRACT TRUNCATED AT 250 WORDS)