The Genetics of Vitamin K Antagonists BF Gage and CS Eby 225

In a recent issue of Nature, Rost et al2 The genetics of vitamin K identified VKOR in this region and confirmed its function by transfecting antagonists the novel gene into a cell line. To identify the gene, they directly se- BF Gage1 and CS Eby2 quenced genomic DNA in the putative 4-Mb region on chromosome 16 from

1 two probands with FMFD and four Division of General Medical Sciences, Washington University School of Medicine, unrelated patients with warfarin resis- 2 St. Louis, MO, USA; Division of Laboratory Medicine at Washington University School tance. They found a 5126 base-pair of Medicine, St. Louis, MO, USA gene (GenBank id # gi:13124769; IM- AGE 3455200) of three exons coding a 163-amino-acid protein. They named The Pharmacogenomics Journal (2004) 4, example, Kohn and Pelz crossed a this gene vitamin K epoxide reductase 224–225. doi:10.1038/sj.tpj.6500258 coumarin-resistant rat with a suscep- complex subunit 1 (VKORC1), which Published online 18 May 2004 tible albino strain and studied the was identical to the VKOR gene iden- seventh generation of backcrossed off- tified by Li et al in the same issue spring.3 Using linkage analysis, they of Nature (see below). Northern blot- Coumarins such as warfarin (Couma- placed Rw, the gene for warfarin resis- ting showed high expression in fetal dint and others) are the most popular tance in the rat, near one of several liver and heart and in adult liver oral anticoagulant. They inhibit the microsatellite markers. Then, using and heart cells. Each resistant patient action of vitamin K epoxide reductase, comparative ortholog mapping, they had unique missense SNPs in the whose gene (VKOR) has been newly identified three candidate human new gene, while the FMFD patients discovered by two independent chromosome locations for Rw, one were homozygous for an identical teams.1,2 In the absence of coumarin, of which was on the short arm of C292T missense SNP that predicted the vitamin K cycle regenerates re- chromosome 16. an Arg98Trp substitution. Next, they duced vitamin K1 from its epoxide The second clue to the location of sequenced 26 coumarin-resistant rats (Figure 1). Reduced vitamin K is a VKOR came from a study of familial (16 caught in the wild) and found cofactor for post-translational g-car- multiple coagulation factor deficiency an A416C missense mutation in all boxylation of glutamic acid residues (FMFD). FMFD is an extremely rare, of them; 15 coumarin-sensitive rats on several proteins, including coagula- autosomal recessive, bleeding disorder lacked this SNP. By sequencing 192 tion factors II (prothrombin), VII, IX, characterized by inadequate g-carbox- control subjects, they detected only and X. Although coumarins also in- ylation of coagulation factors II, VII, two synonymous SNPs, indicating hibit the g-carboxylation of anticoagu- IX, and X. Fregin et al4 investigated that mutations in the gene are rare lant proteins C, S, and Z, inhibition of two kindreds with FMFD with sus- among unselected people (ethnic and clotting factor activity is their main pected VKOR complex mutations racial origins not stated). They estab- pharmacologic effect. g-Carboxylation due to biochemical evidence of ele- lished complementary DNA sequences allows for normal hemostasis by re- vated vitamin K epoxide/vitamin K in the mouse, rat, and puffer fish (F. sulting in negatively charged g-carbox- hydroquinone ratios. Genome-wide rubripes). A topology program pre- yglutamates on factors II, VII, IX, and scanning based on micro-satellite dicted an endoplasmic reticulum (ER) X, which bind to calcium cations and markers and linkage analysis identi- protein, which they confirmed with then to platelet phospholipid mem- fied a marker in the centromeric immunocytochemistry. Finally, they branes. g-Carboxylation is also re- region of chromosome 16 with a confirmed VKOR function by trans- quired for the development of other LOD score of approximately 3, and fecting human embryonic kidney cells tissues, and warfarin exposure in utero haplotype analysis confirmed homo- (HEK293) with the novel gene. The can cause mental retardation, nasal zygosity for chromosome region gene conferred VKOR activity, which hypoplasia, and limb or digit abnorm- 16p12–q21 in one of the families. was warfarin-sensitive, especially in alities in the fetus. Mapping studies of Rw in the rat higher warfarin concentrations. The By the same mechanism, coumarins and a similar gene in mice (war) gene from the coumarin-resistant are effective rodenticides, causing fatal identified a homologous region on rat (Rw) conferred VKOR activity bleeding in coumarin-naı¨ve rodents. human chromosome 16, and several that was minimally inhibited by war- Studies of wild-caught, coumarin-re- known genes that flank Rw and War farin. Surprisingly, genes from three sistant rats have provided important are located on the short arm of of four warfarin-resistant patients clues to the location of VKOR. For chromosome 16.3 conferred lower levels of VKOR The genetics of vitamin K antagonists BF Gage and CS Eby 225

new agent would inhibit but not entirely suppress g-carboxylation. One novel strategy to prevent complete suppression of g-carboxylation would be to include a low-dose of vitamin K1 with any marketed product so that g- carboxylation would never be entirely suppressed. In summary, the results of rigorous and creative experiments performed by these investigators confirm the discovery of a novel reductase with VKOR activity. In vivo, VKOR likely functions in conjunction with other proteins to form a complex in the endoplasmic reticulum. Knowledge of the new gene should help in under- Figure 1 The liver takes up free warfarin where it is biologically active and metabolized by standing the mechanism of g-carbox- the cytochrome P450 complex (CYP2C9). Commercially available warfarin is a racemic ylation of clotting factors and other mixture with each of the two enantiomers having its own route of metabolism. The S- enantiomer more strongly blocks vitamin K epoxide reductase (VKOR), thereby preventing proteins and may aid in the develop- regeneration of reduced vitamin K. Reduced vitamin K is needed for c-carboxylation of ment of novel drugs. glutamic acid residues on coagulation factors II (prothrombin), VII, IX, and X. DUALITY OF INTEREST The authors have no financial or other conflict of interest. activity, especially in the presence The discovery of VKOR and recent of warfarin. This finding suggests studies of SNPs in cytochrome P450 that VKOR is only one subunit of a (CYP2C9) suggest that pharmacoge- ACKNOWLEDGEMENTS complex that confers VKOR activity netics-based warfarin doing has the This review was supported by NIH grant R01 HL71083. in vivo. potential to improve the safety of oral In the same issue of Nature, T Li and anticoagulant therapy. S-warfarin, the Correspondence should be sent to: colleagues used positional cloning to more potent half of commercial war- Dr BF Gage, Division of General Medical identify VKOR in the chromosome farin, is metabolized by CYP2C9. Each Sciences, Washington University School of region 16p12–q21. First, they did a CYP 2C9*2 allele is associated with a Medicine, Campus Box 8005, 660 S. Euclid BLASTX search of this region and 19% decrease in warfarin requirements Ave., St. Louis, MO, 63110, USA. eliminated genes with known function. and each CYP 2C9*3 allele is asso- Tel: 314 454-8369 Then, because VKOR was expected to ciated with a 30% decrease in warfarin E-mail: [email protected] be a transmembrane protein,5 they requirements. As approximately 30% focused on 13 candidate genes of un- of Americans carry at least one of the known function with sequence motifs variants, and because the variants REFERENCES predictive of transmembrane proteins. double or triple the risk of hemorrhage 1LiTet al. Nature 2004; 427: 541–544. By cleverly using double-stranded, in patients initiating warfarin,6,7 in the 2RostSet al. Nature 2004; 427: 537–541. short-interfering RNA (siRNA), they future clinicians may initiate coumar- 3 Kohn MH, Pelz HJ. Blood 2000; 96: 1996– 1998. were able to degrade each gene sequen- ins with a pharmacogenetic-based 4 Fregin A et al. Blood 2002; 100: 3229– tially in a lung carcinoma cell line dose.8 Such an approach could also 3232. (A549) that had VKOR activity. They incorporate clinical factors and genetic 5 Carlisle TL, Suttie JW. Biochemistry 1980; 19: found that MGC11276 mRNA, which variability in the coagulation factors 1161–1167. 9 6 Higashi MK et al. JAMA 2002; 287: 1690– mapped to 16p11.2, was required for themselves. 1698. VKOR activity. They found that VKOR An alternative is to develop newer 7 Margaglione M et al. Thromb Haemost 2000; consisted of 5126 base pairs and three anticoagulants that overcome the pro- 84: 775–778. exons, and it encoded a protein, weigh- blems with coumarin therapy. Already, 8 Gage BF et al. Thromb Haemost 2004; 91: 87–94. ing 18.2 kDa, and having no homology one oral thrombin inhibitor (ximela- 9 Shikata E et al. Blood 2004; 103: 2630– to known proteins. They confirmed the gatran) has had promising results in 2635. activity of VKOR by transfecting insect clinical trials,10,11 and newer antic- 10 Olsson SB, for the Executive Steering Com- cells (S. frugiperda) with the gene and oagulants are under development. mittee on behalf of the SPORTIF III Investi- gators. Lancet 2003; 362: 1691–1698. demonstrating that they gained warfar- VKOR is a potential pharmacological 11 Schulman S et al. N Engl J Med 2003; 349: in-sensitive VKOR activity. target for a novel drug. Ideally, the 1713–1721.