Stimulation of microsomal production of reactive oxygen intermediates by rifamycin SV: effect of ferric complexes and comparisons between NADPH and NADH

Rifamycins--obstacles and opportunities

With nearly one-third of the global population infected by Mycobacterium tuberculosis, TB remains a major cause of death (1.7 million in 2006). TB is particularly severe in parts of Asia and Africa where it is often present in AIDS patients. Difficulties in treatment are exacerbated by the 6-9 month treatment times and numerous side effects. There is significant concern about the multi-drug-resistant (MDR) strains of TB (0.5 million MDR-TB cases worldwide in 2006). The rifamycins, long considered a mainstay of TB treatment, were a tremendous breakthrough when they were developed in the 1960's. While the rifamycins display many admirable qualities, they still have a number of shortfalls including: rapid selection of resistant mutants, hepatotoxicity, a flu-like syndrome (especially at higher doses), potent induction of cytochromes P450 (CYP) and inhibition of hepatic transporters. This review of the state-of-the-art regarding rifamycins suggests that it is quite possible to devise improved rifamycin analogs. Studies showing the potential of shortening the duration of treatment if higher doses could be tolerated, also suggest that more potent (or less toxic) rifamycin analogs might accomplish the same end. The improved activity against rifampin-resistant strains by some analogs promises that further work in this area, especially if the information from co-crystal structures with RNA polymerase is applied, should lead to even better analogs. The extensive drug-drug interactions seen with rifampin have already been somewhat ameliorated with rifabutin and rifalazil, and the use of a CYP-induction screening assay should serve to efficiently identify even better analogs. The toxicity due to the flu-like syndrome is an issue that needs effective resolution, particularly for analogs in the rifalazil class. It would be of interest to profile rifalazil and analogs in relation to rifampin, rifapentine, and rifabutin in a variety of screens, particularly those that might relate to hypersensitivity or immunomodulatory processes.

A comparative study of the redox-cycling of a quinone (rifamycin S) and a quinonimine (rifabutin) antibiotic by rat liver microsomes

Rifamycin S and rifabutin are clinical drugs used to treat tuberculosis and leprosy. The formation of reactive oxygen species during the redox-cycling of rifamycin S (quinone) and rifabutin (quinonimine) was evaluated. The semiquinone (or semiquinonimine) and hydroquinone (or hydroquinonimine) formed during the reduction of the parent molecules by microsomal electron transfer in the presence of nicotinamide-adenine dinucleotide phosphate, reduced (NADPH) or nicotinamide-adenine dinucleotide, reduced (NADH) reoxidizes in air to generate superoxide radical and hydrogen peroxide. In the presence of added iron, hydroxyl radicals, formed by the Fenton reaction, were detected using 5,5'-dimethyl-1-pyroline-N-oxide as the spin-trap. Rifamycin S, a quinone, redox cycles more efficiently than rifabutin, a quinonimine, as approximately five times the concentration of hydroxyl radical adduct of 5,5'-dimethyl-1-pyroline-N-oxide (DMPO) was detected, when compared with rifabutin. The NADPH-dependent microsomal production of hydroxyl radical in the presence of rifamycin S was somewhat higher than the NADH-rifamycin S system with most iron chelators. However, with rifabutin, NADH-dependent microsomal production of hydroxyl radical was higher than that found with the NADPH-rifabutin system. An exception was the iron chelator, diethylene-triamine-pentacetic acid (DTPA), in which NADPH-dependent rates exceeded the rates with NADH with both antibiotics. Rat liver sub-mitochondrial particles also generated hydroxyl radical in the presence of NADH and either rifamycin S or rifabutin. The electron transport chain inhibitors such as rotenone and antimycin A enhanced the signal intensity of DMPO-OH, suggesting NADH dehydrogenase (complex I) as the major component involved in the reduction of rifamycin S. Rifamycin S was shown to be readily reduced to rifamycin SV, the corresponding hydroquinone by Fe(II); under similar conditions Fe(II) did not reduce rifabutin. Using optical spectroscopy, we determined that rifamycin S forms a complex with Fe(II). The stoichiometry of the complex was Fe(rifamycin S)3 in phosphate buffer at pH 7.4. Rifabutin did not form a detectable complex with Fe(II). The redox cycling of rifamycin S and rifabutin did not cause microsomal lipid peroxidation. In fact, the Fe:ATP induced lipid peroxidation was completely inhibited by these two molecules. These results indicate that rifamycin S and rifabutin can interact with rat liver microsomes to undergo redox-cycling, with the subsequent production of hydroxyl radicals when iron complexes are present. Compared to NADPH, NADH is almost as effective (rifamycin S) or even more effective (rifabutin) in promoting these interactions. These interactions may play a role in the hepatotoxicity associated with the use of these antibiotics.

Induction of c-jun protooncogene expression by hydrogen peroxide through hydroxyl radical generation and p60SRC tyrosine kinase activation

The mechanisms of signal transduction of c-jun induction by hydrogen peroxide are elucidated in NIH3T3 cells by using trapping agents of hydroxyl free radical or inhibitors of various protein kinases. Pre-treatment of the cell with hydroxyl radical scavenger dimethyl sulfoxide (DMSO) abolishes the H2O2-induced c-jun expression. Hydroxyl radical generation can be detected and quantified in cells treated sequentially with DMSO and H2O2 for 30 min respectively by methane sulfinic acid (MSA) production, especially that from particulate fraction. Induction of c-jun by H2O2 is also dramatically reduced by pretreating the cells with biological antioxidant vit. E. Protein tyrosine kinase activity of membrane fraction is induced by H2O2 within 5 to 10 min, which can be prevented by DMSO pre-treatment. Inhibitor of non-receptor type tyrosine kinase, herbimycin A, has inhibitory effect on H2O2-induced c-jun expression while the inhibitor of receptor type tyrosine kinase, tyrphostin 23 or inhibitor of cyclic AMP dependent protein kinase, KT 5720, has not. TPA pre-treatment that depletes protein kinase C (PKC) has no influence on the c-jun induction by H2O2. Our results suggest that the highly reactive species HO is generated after H2O2 enter cells and mediate the signal responses of H2O2 including c-jun induction and the activation of p60src tyrosine kinase might be one of the molecular events associated with the c-jun induction pathway.

Stimulation of NADH-dependent microsomal DNA strand cleavage by rifamycin SV

Rifamycin SV is an antibiotic anti-bacterial agent used in the treatment of tuberculosis. This drug can autoxidize, especially in the presence of metals, and generate reactive oxygen species. A previous study indicated that rifamycin SV can increase NADH-dependent microsomal production of reactive oxygen species. The current study evaluated the ability of rifamycin SV to interact with iron and increase microsomal production of hydroxyl radical, as detected by conversion of supercoiled plasmid DNA into the relaxed open circular state. The plasmid used was pBluescript II KS(-), and the forms of DNA were separated by agarose-gel electrophoresis. Incubation of rat liver microsomes with plasmid plus NADH plus ferric-ATP caused DNA strand cleavage. The addition of rifamycin SV produced a time- and concentration-dependent increase in DNA-strand cleavage. No stimulation by rifamycin SV occurred in the absence of microsomes, NADH or ferric-ATP. Stimulation occurred with other ferric complexes besides ferric-ATP, e.g. ferric-histidine, ferric-citrate, ferric-EDTA, and ferric-(NH4)2SO4. Rifamycin SV did not significantly increase the high rates of DNA strand cleavage found with NADPH as the microsomal reductant. The stimulation of NADH-dependent microsomal DNA strand cleavage was completely blocked by catalase, superoxide dismutase, GSH and a variety of hydroxyl-radical-scavenging agents, but not by anti-oxidants that prevent microsomal lipid peroxidation. Redox cycling agents, such as menadione and paraquat, in contrast with rifamycin SV, stimulated the NADPH-dependent reaction; menadione and rifamycin SV were superior to paraquat in stimulating the NADH-dependent reaction. These results indicate that rifamycin SV can, in the presence of an iron catalyst, increase microsomal production of reactive oxygen species which can cause DNA-strand cleavage. In contrast with other redox cycling agents, the stimulation by rifamycin SV is more pronounced with NADH than with NADPH as the microsomal reductant. Interactions between rifamycin SV, iron and NADH generating hydroxyl-radical-like species may play a role in some of the hepatotoxic effects associated with the use of this antibacterial antibiotic.