Quantitative comparison of the cytocidal effect of seven macrolide antibiotics on human periodontal ligament fibroblasts.

The cytocidal effect of seven macrolide antibiotics on human periodontal ligament fibroblasts (Pel cells) was studied. Pel cells were exposed for 48 h to erythromycin (EM), clarithromycin (CAM), roxithromycin (RXM), azithromycin (AZM), josamycin (JM), midecamycin (MDM), and rokitamycin (RKM), and allowed to form colonies. The cytocidal effect of the macrolides was measured as a decrease in colony-forming efficiency and was found to increase with the concentration. To obtain a quantitative measure of the cytocidal effect, the LD50, i.e. the concentration that decreases colony-forming efficiency 50% relative to control cells, was extrapolated from the concentration-response curves. The rank of the macrolides according to their cytocidal effect (LD50) was RKM > RXM > CAM > AZM > JM > MDM approximately EM. RKM, RXM, CAM, AZM, and JM were at least 1.7-12.2 times more cytocidal than MDM or EM. When extrapolated from the concentration-response curves, the relative survival of the Pel cells exposed to each of the macrolides at the MIC90 concentrations for periodontopathic bacteria was estimated to be: > or = 53.8% for RKM, > or = 92.7% for RXM, > or = 94.6% for CAM, > or = 97.1% for AZM, and > or = 86.2% for EM. The effect of the antibiotics on the mRNA expression of alkaline phosphatase (ALP) and type I procollagen (COL) was examined in Pel cells exposed for 48 h to RXM, CAM, AZM, and EM, which exhibited strong, moderate, and weak cytocidal activity. The constitutive levels of both ALP and COL mRNA were retained in cells exposed to RXM at < or = 3 microM, CAM at < or = 10 microM, and AZM or EM at < or = 3 microM. The MIC90 against periodontopathic bacteria is < or = 4.8 microM for RXM, 5.3 microM for CAM, 2.7 microM for AZM, and 21.8 microM for EM. These results suggest that topical administration of CAM or AZM to the gingival crevice at their MIC90 concentration for periodontopathic bacteria would have little adverse effect on the growth and differentiation of the periodontal ligament. It is important to note, however, that these findings have yet to be extrapolated to in vivo conditions.