Vitamin K-dependent proteins undergo post-translational modification in which the γ-carbon on selected glutamic acid (Glu) residues is carboxylated, producing γ-carboxyglutamic acid (Gla). The vitamin K-dependent proteins include the coagulation proteins prothrombin, factor VII, factor IX, factor X, protein C, protein S, and protein Z, as well as proteins involved in bone formation and other processes. The γ-carboxylation reaction is catalyzed by vitamin K γ-glutamyl carboxylase. Vitamin K hydroquinone is a co-substrate for the carboxylase and undergoes oxidation to vitamin K epoxide during the carboxylation reaction. To cycle back and participate in another reaction, vitamin K epoxide is reduced to vitamin K hydroquinone by the vitamin K epoxide reductase (VKOR).1,2 The anticoagulant activity of warfarin is due to its ability to inhibit VKOR.
Homologs of VKOR recently have been identified in bacteria and participate in a pathway leading to disulfide bond formation, which is necessary for the stability of many secreted bacterial proteins. Although bacterial oxidation of cysteine that leads to disulfide bond formation is a different reaction than γ-glutamyl carboxylation, both bacterial and human VKOR participate in the transfer of electrons to a quinone. Additionally, human VKOR and bacterial homologs both contain an active site CXXC sequence, which cycles between reduced and disulfide-bonded states.
Now, Dutton et al., in the laboratory of Dana Boyd at Harvard, show that warfarin inhibits the VKOR homolog from Mycobacterium tuberculosis, the causative agent of human tuberculosis. Additionally, using a random mutagenesis strategy they identified three warfarin-resistant M. tuberculosis VKOR mutations at sites homologous to those of human warfarin-resistant VKOR mutants. Deletion of VKOR in another mycobacterial species, M. smegmatis, produced a growth defect that was rescued by expression of M. tuberculosis VKOR. Additionally, the growth of M. tuberculosis was inhibited by warfarin.
Does this mean that warfarin should be added to the armamentarium of drugs used to treat tuberculosis? No. M. tuberculosis VKOR shares only 18 percent amino acid identity with its human homolog. Millimolar concentrations of warfarin are required to inhibit mycobacterial VKOR, in contrast to the micromolar concentrations required to inhibit human VKOR. Additionally, mutants of M. tuberculosis VKOR that were warfarin-resistant when expressed in E. coli were not warfarin-resistant when expressed in M. smegmatis, indicating that the effect of warfarin on mycobacteria may involve additional targets other than VKOR.
In Brief
Although further work is necessary to understand the actions of warfarin and the structure and function of VKOR in mycobacteria and possibly other pathogenic bacteria, this study suggests that bacterial VKOR may represent a therapeutic target. Additionally, because human VKOR is a membrane-bound enzyme and difficult to study, bacterial VKOR may represent a useful model for structure-function studies of human VKOR, as witnessed by the recent description of an X-ray structure of Synechococcus VKOR.3
References
Competing Interests
Dr. Lollar indicated no relevant conflicts of interest.