In a previous study of structure-activity relationships of selected phenazines against Mycobacterium leprae in vitro, compounds containing a 2,2,6,6-tetramethylpiperidine substitution at the imino nitrogen were most active. Therefore, the effect of substitution at the para positions of the phenyl and anilino groups in tetramethylpiperidine-substituted phenazines was assessed. As determined by radiorespirometry, activity in ascending order was observed in compounds substituted with hydrogens or fluorines, ethoxy groups, methyl groups, chlorines, and bromines and correlated with partition coefficients in octanol-water.Despite its dose-limiting gastrointestinal toxicity and doserelated skin pigmentation, clofazimine (B663) remains a firstline drug in leprosy chemotherapy (1,14). Although a wide variety of structural analogs of B663 has been synthesized (11), only a limited number of compounds have been evaluated in the 6-to 12-month mouse footpad model (13), which until recently was the only available screening system for new antileprotics. An examination of six structural analogs of clofazimine in a Mycobacterium leprae-macrophage in vitro system failed to identify compounds with superior activity (10).Using a radiorespirometric metabolic assay, we have been rapidly and inexpensively evaluating a large number of antimicrobial agents for antileprosy activity in vitro (3-8). Recently we assessed 12 phenazines which (besides not being expected to accumulate in body fat) differ from clofazimine in substitution at the imino nitrogen and in some cases in chlorination of the phenazine nucleus or anilino and phenyl rings (6). In general, activity increased with the degree of chlorination. Phenazines containing a 2,2,6,6-tetramethylpiperidine substitution (TMP) at the imino nitrogen were clearly the most active compounds in the series, with the monochloro (B4019) and dichloro (B3786) analogs displaying essentially identical activity.We describe here the in vitro assessment of substitution at the para positions of the anilino and phenyl groups of TMP, with a view toward identifying nonpigmenting compounds with high activity against M. leprae.Susceptibility testing was performed essentially as previously described (6). Briefly, footpads of M. leprae-infected, athymic nude mice were surface decontaminated, and bacilli were obtained by homogenization of the footpads in Dubosalbumin medium (pH 6.5) followed by differential centrifugation. Samples were removed for contamination checks on a variety of culture media. The remaining M. leprae suspension was held at 4°C for 2 days, during which time the culture medium was observed to confirm the absence of contamination. The suspension was then diluted in Dubos-albumin medium to a final density of 107/ml and distributed in 1-ml aliquots to 6-ml glass screw-cap vials.