R- and S-Warfarin Inhibition of Vitamin K and Vitamin K 2,3-Epoxide ...

THEJOURNALOF

BIOLOGICAL CHEMISTRY

Vol. 257, No.9. Issue of May 10, pp. 48944901, 1982 Printed In U.S.A.

R- and S-Warfarin Inhibitionof Vitamin K and Vitamin K 2,3-Epoxide Reductase Activities in the Rat* (Received for publication, September 3, 1981)

Michael J. FascoS and Louise M. Principeg From the Division of Laboratories a n d Research, New York State Department of Health, Albany, New York 12201

cofactor in vitro (the physiological reductant is not known) and is warfarin-sensitive (15, 16); and the other (DT-diaphorase EC 1.6.99.2) requires NADH as cofactor and isrelatively warfarin-insensitive (15, 17, 18). Thehydroquinonethen serves as a cofactor for a y-carboxylase, which produces ycarboxyglutamic acid residuesin postribosomal precursor protein, and as a substrate for an epoxidase, yielding vitamin K 2,3-epoxide. The requirements of the y-carboxylation reaction (temperature optimum, oxygen, etc.) are similar to those of as Chloro-K the epoxidation reaction and inhibitors such affect the two reactions similarly(19), thus suggesting an interrelationship. The epoxide is reduced back to vitamin K by vitamin K 2,3-epoxide reductase which also utilizes dithiothreitol as cofactor in vitro and is warfarin-sensitive. The mechanism(s) whereby warfarin antagonizes vitamin Kfunction hasnot beenunambiguously determined,but considerable evidence indicates vitamin K 2,3-epoxide reductase to be its primary site of action: (i) the reductase of warfarin-resistant ratsis less sensitiveto warfarin than thatof normal rats (20), and (ii)warfarin increases the normalepoxide/vitamin K ratio in vitro and in vivo which i s a consequence of reductase inhibition (4,21). Presumably epoxide reductase inhibition lowers the hepatic vitamin K concentration below that required to supportcoagulation factor synthesis (22). Other evidence, however, suggests that inhibition of the epoxide reductase may not be the solemode of warfarinantivitamin K activity. In particular, studiesby Bellet al. (23) indicated that warfarin inhibition of coagulation factor synthesis occurs at hepatic vitamin K concentrations which are The vitamin K-vitamin K 2,3-epoxide metabolic cycle has sufficient to restore coagulant activity to animals made hypoprothrombinemic fromvitamin K deficiency (24). Basedon beenlinked to the biosynthesis of calcium-bindingy-carboxyglutamic acid residuesin the coagulation Factors I1 (pro- these results we considered that thephysiological mechanism is of warfarin action also includes inhibition of vitamin K hydrothrombin), VII, IX,and X (1-4). Calciumionbinding required for activation of these coagulation factors to enzy- quinone formation. by We report hereour investigations to test this hypothesis matically active forms (5, 6). Vitamin K also promotes ycarboxyglutamic acid formation in two other blood proteins comparing the activities of the warfarin sensitive-vitamin K of unknown function (7-9) and in othercalcium-binding pro- and vitamin K 2,3-epoxide reductases with thecorresponding teins of developing bone (lo), kidney ( l l ) ,lung (12), spleen coagulant activitiesa t various times following warfarin administration to rats.R- and S-Warfarin enantiomers were used to (13),and placenta (14). The first stepin vitamin K metabolism is itsconversion to determine whether S-warfarin preferentially inhibits one or vitamin K hydroquinone catalyzed byeither of two reductases both reductases which would be consistent with its greater which differ in their cofactor requirement and sensitivity to anticoagulant activity demonstrated in both man (25) and the inhibition by warfarin. Onereductase utilizes dithiothreitol as rat (26). * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Supported by Grant R 0 1 HL-19772 from the National Institutes of Health, United States Public Health Service, Department of Health and Human Services. To whom correspondence should be addressed. 5 T o be submitted as part of a thesis in partial fulfillment of the requirements for the degree of doctor of philosophy a t Albany Medical College of Union University, Albany, NY.

+

EXPERIMENTAL PROCEDURES

Materials-Racemicwarfarin was purchased from CalbiochemBehring, Simplastin from General Diagnostics (Morris Plains, NJ), activated rabbit brain thromboplastin from Dade Diagnostics (Aquada,PuertoRico), inosithinfrom Associated Concentrates (Woodside, NY), Russell’s viper venom from Miami Serpentarium Labs (Miami, FL), vitamin K’ and dithiothreitol from Sigma, and Emulgen 911 from Kao Atlas (Tokyo, Japan). The high performance

4894

’ Throughout this work vitamin K1 was used.

Downloaded from http://www.jbc.org/ by guest on October 1, 2017

Reduction of vitamin K 2,3-epoxide and vitamin K catalyzed b y hepatic microsomal enzymes is required for normal,postribosomal, y-carboxyglutamate formation in the prothrombin complex Factors 11, VII, and IX, X. The R- and S-warfarin enantiomers differentially inhibit (S-warfarin is 2 to 5 times more active) vitamin K function by mechanisms which have not been unambiguously determined. As a step toward determining the physiologically relevant site(s) of warfarin-antivitamin K activity we investigated in Wistar rats the effects of R- and S-warfarin on vitamin K 2, 3-epoxide and vitamin K reductase activities and correlated t h e m with effects on plasma concentrations of the Factors 11, VII, and X. Based on the results of these studies we conclude that: 1) warfarin inhibition of the vitamin K 2,3-epoxide and vitamin K reductases is essentially irreversible; 2) S-warfarin stereoselectively inhibits both reductases in vivobut not in vitro; 3) the vitamin K reductase which utilizes dithiothreitol as cofactor in vitro is primarily responsible f o r vitamin K reduction to vitamin K hydroquinone under physiological conditions; 4) warfarin initially inhibits y-carboxyglutamate formation by inhibiting simultaneously the vitamin K 2,3-epoxide and vitamin K reductases; and 5 ) following enantiomer administration there is an apparent lack of correlation between the restoration of the reductase activities and the reinitiation of coagulation factor synthesis.

Antivitamin K Activity of R- and S- Warfarin

____.

The abbreviations used are: HPLC,high performance liquid chromatography; PEG, polyethylene glyvi.

water), 1 ml of inosithin (1 mg/ml of 0.15 M NaCl), and 8.5 ml of imidazole-PEG buffer. The Factor X-deficient plasma, diluted test plasma, thromboplastin, and CaCL buffer were added as described for the assay of Factor VII. Clot times were recorded from the time of CaC12 addition. WarfarinBinding to Microsomal Reductases-All operations were performed a t 1 to 5 "C. Two ultracentrifuge polyallomer tubes (2.5 X 8.9 cm) each contained 9.8 ml of 200 mM Tris-HC1,0.15 M KC1 buffer, pH 7.4, and 4.2 ml of microsomes (10 mgof protein/ml). Warfarin sodium salt (28 nmol) was included in the Tris-KC1 buffer added to one of the tubes. After 1 h, two 2-ml aliquots were removed from each tube and were assayed for vitamin K and vitamin K 2,3epoxide reductase activities as described below. The remaining 10-ml portions were diluted with sufficient 20 mM Tris-HC1, 0.15 M KC1 buffer, pH 7.4, to fiil thetubes (approximately 25 m l ) andthe microsomes were pelleted by centrifugation a t 105,000 X g for 75 min. The supernatants were decanted, and the pellets carefully washed 3 times with 30 to 35 mi of the 20 mM Tris-KC1 buffer, pH 7.4. Each pellet was resuspended by homogenization with a Teflon pestle in sufficient 200 mM Tris-HC1, 0.15 M KC1 buffer, pH 7.4, to restore the volume to 10 ml. Aliquots (2 ml) were assayed for vitamin K and vitamin K 2,3-epoxide reductase activities in the same way as the uncentrifuged samples. Vitamin K a n d Vitamin K 2,3-Epoxide Reductase Assays-Microsomal vitamin K and vitamin K 2,3-epoxide reductase activities in normal rats and rats administered the warfarin enantiomers were determined as previously described(16)withslight modification. Reaction mixtures were 2 ml and contained in order of addition: 1.35 ml of 200 mM Tris-HCI, 0.15 M KC1 buffer, pH 7.4, with or without Rand S-warfarin sodium salt added; 40 nmol of vitamin K or vitamin K 2,3-epoxide substrate in 0.01 ml of 2% (v/v) Emulgen 911; and 0.6 ml of microsomes (10 mg of protein/&). Mixtures were incubated a t 25 "C for 1 min and 0.04 ml of dithiothreitol (100 mM) was added with mixing. After 5 additional min a t 25 "C, reaction was terminated with isopropanol/hexane and concentrations of vitamin metabolites were determined by HPLC as described (16). Variations in this procedure are cited in the text. RESULTS

Addition of 2 PM warfarin to microsomes inhibited vitamin 5% of normal and vitamin K 2,3epoxide reductase activity to25% of normal (Fig. 1). Dilution of the microsomes with buffer,followed by ultracentrifugation and repeated washing of the resultant pellet did not even partially restore the activity of either enzyme. Control microsomes containing no warfarin but otherwise identicallytreated lost only 28 and 4%, respectively, of their vitamin K and epoxide reductase activities as a consequence of the washing treatment.Warfarininhibition of bothreductases is thus irreversible by these criteria. These results canbe compared with those of Lorusso and Suttie (35) which demonstrated that rat hepatic microsomes bind warfarin in an essentially irreversible manner. Since the warfarin concentrations required to saturate the microsomal binding site(s) (0.1 to 5 PM) (35) were similar to the concentrations which we previK and epoxide ously demonstratedinhibitedthevitamin reductases (16), it is probable that the irreversible binding occurred at the enzymes involved in vitaminK and vitaminK 2,3-epoxide reduction. The effects of R- or S-warfarin (0.75 p ~ on) quantities of vitamin K hydroquinone formed from vitamin K by microsomes with dithiothreitol as electron donor as a function of time are illustrated in Fig. 2. The rate of hydroquinone formation, in the absenceof warfarin enantiomers, was linear for 5 min at 0.22 nmol/mg of protein/min. R- and S-Warfarin were equally effective as inhibitors of hydroquinone formation, diminishing the linear rate to 0.11 nmol/mg of protein/ min. Initial hydroquinone formation rates were linear for at least 20 min under conditions of warfarin inhibition by 75% or greater. R- and S-Warfarinwere also equivalently effective as inhibitors of the reduction of vitamin K 2,3-epoxide t o vitamin K

K reductase activity to


liquid chromatograph was a Waters Associates (Milford, MA) model 244 equipped with a Spectra Physics (Santa Clara, CA) model 4000 recording integrator and WISP automatic injector (Waters Associates). pBondapak CIH columns and HPLC2 solvents were obtained from Waters Associates. Water was deionized, glass-distilled, and fiitered through a 0.22 pm membrane (Millipore Corp., Bedford, MA) prior to use in the HPLCstudies. Coagulation assays were performed with a fibrometer from BBL (Cockeysville, MD). = 149")- and Preparation of Compounds-Optically pure R ([a],, S ( [ C U ]=~ ,-149.7")-warfarin and their sodium salts were prepared by the method of West et al. (27). Vitamin K 2,3-epoxide was prepared by the hydrogen peroxide oxidation of vitamin K by the method of Fieser et al. (28). Both vitamins were purified to chromatographic homogeneity with detection a t 254 nm on a pBondapak CIR preparative column (7.8 mm, inner diameter, X 30 cm) using acetonitrile as the mobile phase a t a 3 ml/min flow rate. The solvent was removed in uacuo from collected fractions containing vitamin K and its 2,3epoxide. Residues of each vitamin weredissolved in aqueous Emulgen 911 (10%v/v) to a final concentration of 20 mg/ml, using the method described by Lowenthal and Jaegar (29). Solutions of vitamin K and epoxide a t 2 mg/ml were prepared by dilution of the concentrated solutions withwater and were used in metabolic studies. Animal Studies-The experimental animals were male Wistar rats (250 & 10 g) from a colony maintained in this division. An aqueous solution of the sodium salt of R-warfarin was administered orally to rats atdoses of 20 or 10 mg/kg. S-Warfarin sodium salt was similarly administered a t doses of 10 or 1 mg/kg. Atvarious timesafter administration rats (4/group)were rendered unconscious withN2 and 4.5 ml of blood was withdrawn by cardiac puncture with a 5-ml plastic syringecontaining 0.5 ml of 3.8% trisodium citrate.Plasma was obtained by centrifugation of the citrated blood a t 2000 X g for 15 min a t 5 "C. R- andS-Warfarinconcentrations in plasmawere determined by HPLC using the rapid chromatographic method, which permits quantitation of warfarin free of its metabolites (30). T h e liver of each ratwas perfused in situfor 90 s with physiological saline,removed, and stored in cold 20mM Tris-HCI, 0.15 M KC1 buffer, pH 7.4. All further operations were performed a t 5 "C. Each liver was minced and homogenized in 3volumes of the Tris-KC1 buffer, pH 7.4. The suspensions were centrifuged a t 10,OOO X g for 20 min and microsomes pelleted by centrifugation of the supernatant at 105,000 X g for 75 min. The microsomes were resuspended in 3 to 5 ml of the Tris-KC1 buffer, pH 7.4, and the protein concentrations were determined by the method of Bradford (31) using commercially available reagents (Bio-Rad). T h e suspensions were diluted to 10 mg of protein/ml with the same buffer and were stored a t -80 "C. No change in the activities of the vitamin K or vitamin K 2,3-epoxide reductases occurred for a t least 1 month. Control microsomes were prepared identically from a pool of 10 rats. Plasma Coagulation FactorAssays-All clot formation times are the average of triplicate determinations. Factors11, VII, and X plasma concentrations in rats administered the warfarin enantiomers are expressed as per cent normal concentration. Pooled control plasma was obtained from 10 to 15 untreated rats and per cent normal curves were produced by dilution of the control plasma. Factor-deficient plasmas were prepared from species describedin the assay references (32-34). One-stage prothrombin time testswere performed with Simplastin according to manufacturer's direction. Factor I1 concentrations in plasma were determined by the method of Magnin and Lawson (32). Factor VI1 assays were performed by a modification of the method of Nemerson and Clyne (33). T h e imidazole-saline buffer used was 0.01 M imidazole-HCI, 0.15 M NaC1, 0.88% (w/v) PEG 6000 buffer, pH 7.4, and the thromboplastinused was Dade rabbit brain diluted with 3 parts of the imidazole-PEG buffer. All reagents were stored a t 0 "C prior to use. Factor VII-deficient plasma (0.1 ml) and test plasma (0.1 ml) (diluted with the imidazole-PEG buffer) were warmed in a fibrometer cup at37 "C for 2 to 3 min. To this were added sequentially 0.1 r d of thromboplastin (37 "C) and 0.1 ml of25 mM CaCI2 in the imidazole-PEG buffer (37 "C). Clot times were recorded from the time of CaClz addition. Factor X-deficient plasma was prepared by the method of Hougie (34). The imidazole-PEG and imidazole-PEG-CaCI2 buffers were the same as those used in the Factor VI1 assay. Thromboplastin was prepared by mixing 0.5 ml of Russell's viper venom (0.1 mg/ml of

4895

Antivitamin KActivity of R- and S- Warfarin

4896

and its further reduction to vitamin K hydroquinone (Fig. 3). Epoxide consumption catalyzed by control microsomes was linear for 5 min at a rate of 0.16 nmol/mg of protein/min. Vitamin K was the only product detected during the fist minute of reaction and its rate of formation equalled that of epoxideconsumption. Attimes exceeding 1 min, rates of vitamin K formation became nonlinear and attaineda steady state rateduring the period of 5 t o 15 min. As rates of vitamin K formation decreased, vitamin K hydroquinone formation increased linearly at a rate of 0.08 nmol/mg of protein/min for approximately 9 min. R- or S-Warfarin,at a concentration , the initial rate of reduction of the of 0.75 p ~ diminished epoxide to vitamin K to 0.1 nmol/mg of protein/min (63%of control) and the rate of hydroquinone formation 0.01 to nmol/ mg of protein/min (16% of control). The net effect of this of vitamin K formation inhibition is that the steady state rate

2um NARMFfIN

NO 2 Am WARFARIN WARFARIN

WARFARIN

(woshed)

FIG. 1. Effects of warfarin addition to microsomes on per centnormalvitamin K 0 andvitamin K 2.3-epoxide (m) reductase activities before and after centrifugation and washing of microsomes. Control microsomes were treated identically except that warfarin was omitted. Details were as described under “Experimental Procedures.” Total vitamin K 2,3-epoxidemetabolism was calculated from the sum of the concentrations of vitamin K and vitamin K hydroquinone formed during 5 min of incubation at 25 “C.

0

12

8

4

I(

MINUTES FIG. 3. Microsomal metabolism of vitamin K 2,3-epoxide to vitamin K and vitamin K hydroquinone in the presence and absence of warfarin (0.75 p ~ as) a function of time.Vitamin K (M and )vitamin K hydroquinone ( O ” 0 ) concentrations formed during incubation at 25 “C in the absence of the warfarin enantiomers. Concentrations of vitamin K formed in the presence of R (e - 4)-or S (A- - -A)-warfarin; and of vitamin K hydroquinone formed in the presence of R (EL - a)or S (A- - -A)-warfarin during incubation at 25 “C. Reaction mixtures were prepared and the products were assayed as described under “Experimental Procedures.” TABLE I Plasma concentrations of R- or S-warfarin at various times after administration The sodium salt of R-warfarin dissolved in water was administered as a single oral dose of 20 mg/kg. The sodium salt of S-warfarin was similarly administered at a dose of 10 mg/kg. R- or S-Warfarin was extracted from plasma by Sep-Pak C,8 chromatography. R- or SWarfarin concentrations in the eluates were determined by HPLC. Plasma levels are the average & S.D. for four rats. 0

4

8

12

16

MINUTES FIG. 2. Effects of R 0-or S (A)-warfarin (sodiumsalts, 0.75 PM) on normal(M quantities ) of vitamin K hydroquinone formed during microsomal metabolism of vitamin K as a function of time. Reaction mixtures were prepared and the product was assayed as described under “Experimental Procedures” except the time of incubation at 25 “ C was varied.

Time after administration S-Warfarin

h 1

8 16 24 32

Concentration in citrated plasma R-Warfarin

pg/ml

16.2 & 6.1 6.4 & 1.1 2.0 & 1.3 0 3 -+ 0.2 0.2 ‘Ll ~~

13.4 -C 2.7 7.4 -e 0.8 2.2 0.2 1.2 k 0.2

*

0.2 -e 0.1


(washed)

4897

Antivitamin KActivityof R - and S- Warfarin

FIG. 4. Effects of the warfarin enantiomers on per cent normal Factor W (A-A), Factor X (-), and Factor 11 (0- -0)concentrations in rat plasma at various times after administration of a single oral dose. A, R-warfarin at 20 mg/kg; B; S-warfarin at 10 mg/kg. Factor concentrations were determined as described under “Experimental Procedures” and are theaverage f S.D. value obtained from four rats.

-

J

20

HOURS is reached after longer times and atslightly higher concentrations of the vitamin (Fig. 3). At a R- or S-warfarin concentraof epoxide to vitamin tion of 0.25 IJM (data not shown), the rate K conversion was not detectably inhibited while the rate of hydroquinone formation was inhibited by 37%. Warfarin conM complete inhibition centrations greater than1.0 ~ L produced of hydroquinone formation and withincreasing warfarin concentrations a progressive decrease in the rate of vitamin K formation. The plasma concentration of R- and S-warfarin administered separately at doses of 20 and 10 mg/kg, respectively, and determined a t various times thereafter are presented in Table I. These doses producedessentially equivalent plasma concentrations of the enantiomersat most of the timesinvestigated. The effects of these doses on the per cent normal levels of Factors 11, VII,and X in plasma as a function of time are illustrated in Fig. 4. Both R- and S-warfarin diminished Factor VI1 below detectable levels at 16 h after administration, but at 24 h Factor VI1 levels were partially restored only in In contrast to Factor VII,Factors rats treated with R-warfarin. I1 and X reached minimumlevels a t 24 h after administration of either enantiomer. At this time FactorI1 levels were equivX alent in rats administered either enantiomer, but Factor levels were slightly more depressed in S- than in R-warfarinI I I I I 1 “1 treated rats. Factors 11, VII, and X each initially decreased 3 20 40 60 from plasma at a rate which was independent of treatment HOURS with either R- or S-warfarin. At 48 h after dosing with the warfarin enantiomers the average plasma concentrations of FIG. 5. Effects of a single oral dose of R-warfarin (sodium Factors 11, VII, and X had extensively recovered and were salt, 20 mg/kg, C “ 0 ) or S-warfarin(sodium salt, 10 mg/kg, [1“0) on vitamin K reductioncatalyzedbyhepaticmihigher in R- thanin S-warfarin-treated rats. A single oral dose of R-warfarin (20 mg/kg) or S-warfarin crosomes and on one-stage prothrombin times (R-warfarin, (10mg/kg) inhibited the hepaticmicrosome catalzyed reduc- 0- - -0;S-warfarin, H- - -.) at various times after administration. Coagulation and reductase assays were performed as dein vitro to undetectable scribed under “Experimental Procedures.” Normal values were estabtion of vitamin K to the hydroquinone a period of lished from plasma and hepatic microsomes obtained from 10 rats. levels (lessthan 2% of normal concentrations) over 1 to 24 h after administration (Fig. 5 ) . No difference in the R - andS-warfarin-treatedratsand extent of R- and S-warfarin inhibition of the reductase was essentiallyparallelin normal values were not attained even at timesof 96 and 144 observed during this period. At 48 h after administration, however, microsomes from S-warfarin-treated rats exhibited h after administration, respectively. The reductase thus resubstantially less reductase activity than did microsomesfrom covers from the warfarin inhibition very slowly even duringa R-warfarin-treated rats, thus demonstratinga stereoselective period when plasma levels of warfarin are no longer detectainhibition at equivalent plasma concentrations of the enan- ble. A comparison of the time courses of reductase inhibition tiomers. The rates of recovery of reductase activitywere and anticoagulation measured by one-stage prothrombin as-


0

Antivitamin KActiuity of

4898

say following warfarin administration (Fig. 5) demonstrates that maximal inhibition persists beyond times when anticoagulation levels have reached theirmaximum and havebegun to return to normalvalues.

I\

'\

I I I

' \\ I

\ \

\ I

\

\

I$8'

\ \

-

.-

At these doses R- and S-warfarin were also very potent inhibitors of vitamin K ZJ-epoxide reduction and the profiie of the inhibition a t various times after enantiomer administration is illustrated in Fig. 6. The corresponding one-stage prothrombin coagulation times are againincluded for the purpose of comparison. In general epoxide reduction was not as extensively inhibited as was vitamin K reduction (cf. Fig. 5). Further, S-warfarin wassignificantly more effective an inhibitor than was R-warfarin at all the time points investigated. Maximum inhibition of epoxide reduction occurred at 6 h after administration of either enantiomer with average values of 7.5 and 2.5% normal, respectively, for R- and Swarfarin-treated rats. During the ensuing period to 24 h there was a slow recovery phase followed by a more rapid phase with essentially normal values being attained at 96 h in R144 h in S-warfarin-treated rats. warfarin-treated rats and Reinitiation of coagulation factor synthesis occurred at 16 h in the former group withapproximately 9 to 11% normal epoxide reductase levels, and at 24 h in the latter groupwith approximately 3 to 8%normal epoxide reductase levels. A dose of 1 mg/kg of S-warfarin or 10 mg/kg of R-warfarin produced essentially equal changes in the per cent normal plasma concentrations of Factors 11, VII, and X (Fig. 7) thus producing an equivalent anticoagulant response. Factor VI1 was again most affected followed by Factors X and 11. Synthesis of Factor VI1 was initiated sometime before 16 h after enantiomeradministration since atthistimeitspercent normal plasma concentration was greater than at the higher doses (Fig. 5). The effects of these doses on sensitive vitamin K reductase activity are illustrated in Fig. 8. Included in the figure are the profiies of the one-stage prothrombin times. S-Warfarin was a better inhibitor of reductase activity than was R-warfarin and levels of activity during thefist 6 h after administration were approximately 1 and 3.5% of normal, respectively. Inhibition produced by R-warfarin persisted for a much shorter time than that by S-warfarin resulting inmore rapidrecoveries to normal activity. At16 h after administration, when coagulation factor synthesis was occurring,reductase activities were approximately 10 and 1%normal in R- and S-warfarin-treated rats, respectively.

FIG. 7. Effects of the warfarin enantiomerson per centnormal Factor VI1 (A-A), Factor X 1 -(, and Factor I1 (0- -0)concentrationsin rat plasmaat various timesafter administration of a single oral dose. A, R-warfarin at 10 mg/kg; B, S-warfarin a t 1 mg/kg. Conditions were as described for Fig. 4.

-

HOURS


HOURS FIG. 6. Effects of a single oral dose of R-warfarin (sodium or S-warfarin (sodiumsalt, 10 mg/kg, salt, 20 mg/kg, M) (l3-U)onvitamin K 2,t-epoxide reductioncatalyzed by hepaticmicrosomesandon one-stage prothrombintimes -.) at various times (R-warfarin, 0- -0 S-warfarin, after administration.Conditions were as described for Fig. 5 except vitamin K 2,3-epoxidewas the substrate. Total vitamin K 2,3-epoxide metabolism was calculated from the sum of the concentrations of vitamin K and vitamin K hydroquinone formed during 5 min of incubation at 25 "C.

R- and S- Warfarin

Antivitamin KActiuity

of R- and S- Warfarin

4899

16 h after enantiomer administration were 11 or 5% of normal in R- and S-warfarin-treated rats,respectively. DISCUSSION

.-

w

-e

-

w

.-


The effects of R- and S-warfarin administration on hepatic microsomal vitamin K and vitamin K 2,3-epoxide reductase activities were compared with effects on coagulant activity to probe the physiologically relevant sites of warfarin-antivitamin K function. Since the extent of R- and S-warfarin inhibition of reductase activities in vivo could only be determinedby an in vitro assay in hepatic microsomes, it was necessary to assess the effects of microsome isolation on the extent of inhibition. Experiments designed to sirnulate these effects demonstrated that warfarin inhibition of vitamin K and epoxide reductase activitiesis essentially irreversible and is probably not altered during microsome isolation. Results obtained from the in vitro assay thereforereflect the situation in the intact animal. The coagulation data presented here agree well with results reported by Vainieri and Wingard (36) for Factors 11, VII, and X following acute dosage with warfarin. Factor VI1 has the shortest plasma half-life and therefore its loss is the most pronounced. FactorVI1 is also the mostsensitive to variations of its in plasma warfarinlevels(36),presumablybecause HOURS FIG. 8. Effects of a single oral dose of R-warfarin (sodium relatively greater depression and only small changes in concentration are required to overcome its decline. The profile of salt, 10 mg/kg (M or ) S-warfarin (sodium salt, 1 mg/kg, on vitamin K reductioncatalyzed by hepatic mi- the one-stage prothrombin assay per cent normal activity crosomes and on one-stage prothrombin times (R-warfarin, uersus clot time curveis hyperbolic and large changes in clot 0- S-warfarin, - -.) at various times after adminis- formation times accompany relatively small changes in low tration. Conditions were as described for Fig. 5. per cent normal coagulation factor concentrations. Based on coagulation results described here,theassay is primarily 1 reflecting changes in Factor VI1 concentrations. Indeed, the IO0 earliest times detectedfor restoration of Factor VI1 activity in R- and S-warfarin-treated ratsall agree with those detected by one-stage assay. Of the doses employed in these investigations, S-warfarin 80 at 10 mg/kg produced the greatest anticoagulant response (determined by one-stage prothrombin assay) attaining the peak of maximum effect at 24 h after administration. This dose is in large excess of that required to completely block 60 vitamin K dependent-coagulation factor synthesisin rats (36) and thus rates of change of one-stage prothrombin times prior to the peak reflect only coagulation factor loss from plasma of change were due to degradation. Since the same rates 40 initially exhibited following administration of the other R- or S-warfarin doses it canbe concludedthat theyalso completely blocked coagulation factor synthesis.For R-warfarin at 20 mg/kg complete blockage persisted for 16 h after administra20 tion andfor R- and S-warfarinat 10 and 1mg/kg, respectively, 6 h after administration. for We have previously demonstrated (16)that large quantities of vitamin K hydroquinone areformed by Wistar rat hepatic 0, microsomes in the presence of dithiothreitol with either vi20 40 60 tamin K or the epoxide as substrate, and thatonly relatively HOURS insignificant quantities of hydroquinoneare formed when FIG. 9. Effects of a single oral dose of R-warfarin (sodium vitamin K and NADH (the cofactor for DT-diaphorase) are salt, 10 mg/kg (M or) S-warfarin (sodium salt, 1 mg/kg, on vitamin K 2,3-epoxidereductioncatalyzed by added. The time course of dithiothreitol-supported epoxide hepatic microsomes and on one-stage prothrombin times (R- metabolism to vitaminK and hydroquinone (Fig. 3 ) is typical - -.) at various times after of a coupled enzyme system indicating that thetwo warfarin warfarin, 0- - -0. S-warfarin, administration. Conditions were as described for Fig. 6. sensitive-reductases are closely associated in the microsomal membrane. Since the vitamin K concentrations normally presFig. 9 illustrates theeffects of these R- and S-warfarin doses entinhepatic tissue are very low, highlyactivecoupled on vitamin K 2,3-epoxide reductase activity and the relation- enzyme systems would best be able toperform the multistep ship of reductase activities to one-stage prothrombin times. conversions of the vitamin K-vitamin K 2,3-.epoxidemetabolic As with the higher doses, epoxide reduction was not as com- cycle. It thus follows that hydroquinone formation in vivo pletely inhibited as was vitamin K reduction. S-Warfarinwas arises preferentially via the sensitive reductase pathway and the moreeffective inhibitor andlevels of reductase activitya t not via DT-diaphorase as has been suggested (18,37).Hydro-

4900

Antivitamin KActivity

observation (22) that hepatic vitamin K concentrations in rats treated with anticoagulants are sufficient to restorecoagulant activity to rats made hypoprothrombinemic from vitaminK deficiency. Following each dose of R - or S-warfarin,levels of vitamin K 2,3-epoxide reductase activitywere depressed maximally during the period from 1 to 6 h and recovered slightly before restoration of coagulation factor synthesis was detected. Assuming that hepatic reductase concentrations are in excess of those needed to maintain normal coagulant function,only partial recovery of these activities would permit coagulation factor synthesis and thus recovery of coagulation. There is, however, no apparent correlation between extents of reductase inhibition in R- and S-warfarin-treated rats andcoagulation factor synthesisactivity. From the dataof Fig. 6, coagulation factor synthesiswas detected in S-warfarin-treated rats a t 24 h and at epoxide reductase levels which were equal to orless than the 1 to 6 h levels in R-warfarin-treated rats. Sinceat 1 to 6 h no coagulation factor synthesis can be detected, these reductase levels at 24 h should also produce complete inhibition. From Fig. 9, reductaseactivity levels in R-warfarintreated rats were substantially higher than in S-warfarintreated rats, yet the anticoagulant responses were indistinguishable. Since the anticoagulant response at 16 h was less than that producedby the higher R- and S-warfarin doses (cf Fig. 6), thencoagulation factor synthesis mustbe occurring at this time. Assuming that theepoxide reductase is normally an essential part of this process, then in R-warfarin-treated rats (Fig. 9), the levels of epoxide reductase activity at 16 h must be at least partially functional. If this is in fact thecase, then the lower levels of epoxide reductase activity in S-warfarintreated rats (Fig. 9) mustbe less functional and should therefore produce a greater anticoagulantresponse. Finally, even if vitamin K formation via the epoxide reductase is sufficient, how does hydroquinone formationoccur rapidly enoughwhen the sensitive-vitamin K reductase is also inhibited to levels which are apparently below those required to sustaincoaguof these investigations do lation factor synthesis? The results not provide a satisfactory explanation, but do suggest that currently proposed relationships between the vitamin K-vitamin K 2, 3-epoxide metabolic cycle, y-carboxyglutamic acid formation in precursor protein and coagulantactivity may be oversimplified with respect to in vivo events. In summary we have provided evidence that warfarin initially inhibits vitamin K-dependent coagulation factor synthesis by blocking simultaneously vitamin K and vitamin K 2,3epoxide metabolism. We have proposed that DT-diaphorase has a minor role in the reductionof vitamin K under physiological conditions and haveshown that the results of a number of other investigations are consistent with this conclusion. Lastly, we have demonstrated thatfollowing warfarin administration thereis a lack of correlation between the restoration of vitamin K-dependentcoagulation factor synthesis and the recovery of the sensitive reductases. Coagulant activity recovers as plasma warfarin levels diminish, but reductase activities only fully recover very much later. Experiments are currently in progress to elucidate more completely the roles of the enzymesinvolvedin the vitamin K-vitamin K 2,3epoxide metabolic cycle andthemechanisms of warfarin antivitamin K activity. REFERENCES 1. Bell, R. G., and Matschiner, J. T.(1972) Nature 237, 32-33 2. Maschiner, J. T., Zimmermann,A., and Bell, R. G . (1974) Thromb. Diath, Haemorrh. Suppl. 57,45-52 3. Willingham, A. K., and Matschiner, J. T.(1974) Biochem. J.140, 435-441 4. Sadowski, J. A,, and Suttie, J. W. (1974) Biochemistry 13,


quinone formed by cytosolic DT-diaphorase would reach the carboxylase/epoxidase enzymes only by diffusion and would be expected to play only a minor role in coagulation factor synthesis. Support for these conclusions is derived from the experiments of Lowenthal and Birnbaum (38) who studied warfarin and vitaminK effectson FactorVI1 synthesis inliver slices isolated from vitamin K-deficient rats. They concluded that coumarin anticoagulants irreversibly block vitamin K transport to its site of action and that at high vitamin K concentrations the inhibition can be overcome because viof action by an alternate route.We tamin K can reach its site have demonstrated here that warfarin inhibition of the sensitive vitamin K reductase is essentially complete and irreversible at the warfarin doses used. Further, no detectable increase in sensitive vitamin K reductase activity occurs in warfarinized rats during the period when coagulation factor synthesis is proceeding at a rapid rate(Fig. 5). Inview of these data and recent findings demonstrating that hydroquinone formation isnecessary for y-carboxyglutamic acid synthesis in hepatic precursor protein (17, 39), it is highly probable that the irreversible warfarin block observed by Lowenthal and Birnbaum was actually at the site of sensitive hydroquinone formation. At higher vitamin concentrationssufficient hydroquinone was supplied via the alternate pathwaycatalyzed by DT-diaphorase to overcome the warfarin inhibition. Whitlon et al. (15) also reached similar conclusions based on in vitro studies of y-carboxyglutamic acid formation in hepatic microsomes drivenby vitamin K in thepresence of dithiothreitol or NADH. The data from numerous investigations have demonstrated that after warfarin administration there is an increase in the normal vitamin K 2,3-epoxide/vitamin K ratio. This increase has only been observed after administration of vitamin K, however, and the extent of change under physiological conditions is not known. Vitamin K administration also restores coagulant activity to warfarin-treated in ratsadose-dependent K reductase is manner (23). Sincethesensitivevitamin blocked by warfarin, the exogenous vitamin K is probably metabolized to hydroquinone (and subsequently epoxide) by the normally alternate route catalyzed by DT-diaphorase. Warfarin inhibition of the epoxide reductase causes an accumulation of epoxide with a resultant lowered vitamin K concentration. At low concentrations of the vitamin, it cannot sustainDT-diaphorase-catalyzedhydroquinoneformation sufficiently to overcome the coagulation factor degradation is apparent aswas the case inthe studies rate and no synthesis of Bell et al. (23). Athigher concentrationsof vitamin K, DTdiaphorase-catalyzed hydroquinone formation should be more effective thus producing a measurable net increase in coagulation factor synthesis. Since R- and S-warfarin differentially inhibit vitamin K function, it follows that if the sensitive vitamin K and/or vitamin K 2,3-epoxide reductases are the primary site(s) of anticoagulant action, enantiomer inhibition of them shouldbe correspondingly differentiated. While R- and S-warfarinwere equivalently potent inhibitorsof the sensitive vitamin K and epoxide reductases in vitro, S-warfarin wasa much better inhibitor of them when administered in uiuo both at equivalent plasma concentrations and at equivalent anticoagulant effect. The in vivo stereoselectivity of the inhibition coupled with its magnitude strongly suggests that warfarin antagonism of these reductases is an essential part of its antivitamin K activity. Assuming that the DT-diaphorase has a minor role in the physiological metabolism of vitamin K for the reasons discussed above,blockage of vitamin K and epoxide reduction by warfarin essentially prevents vitamin K metabolism in the liver. Thissituationthus offers anexplanation for Bell’s

of R- and S- Warfarin

Antivitamin K Actiuity of R - and S-Warfarin

(1970) Biochim. Biophys. Acta. 201, 309-315 22. Bell, R. G. (1978) Fed. Proc. 37,2599-2604 23. Bell, R. G., Sadowski, J . A., and Matschiner, J . T. (1972) Biochemistry 11, 1959-1961 24. Matschiner, J . T. (1970) in Fat-soluble Vitamins (DeLuca, H. F., and Suttie, J. W., eds) pp. 377-397, University of Wisconsin Press, Madison 25. Hewick, D. S., and McEwen, J. (1973) Pharm. Pharmacol. 25, 458-465 26. Breckenridge, A., and Orme, M. (1972) Life Sci. 11,337-345 27. West, B. D., Preis, S., Schroeder, C. H., and Link, K. P. (1961) J. Chem. SOC. 83,2676-2679 28. Fieser, L. F., Campbell, W. P., Fry, E. M., and Gates, M. D., Jr. (1939) J. Am. Chem. SOC. 61, 3216-3223 29. Lowenthal, J., and Jaeger, V. (1977) Biochem. Biophys. Res. Commun. 74,25-32 30. Fasco, M. J., Cashin, M. J., and Kaminsky, L. S. (1979)J. Liquid Chromatogr. 2, 565-575 31. Bradford, M. M. (1976) Anal. Biochem. 72, 248-254 32. Magnin, A. A., and Lawson, W. B. (1975) Throm. Res. 7,555-566 33. Nemerson, Y., and Clyne, L. P. (1974) J. Lab. Clin. Med. 83, 301-303 34. Hougie, C. (1962) Proc. SOC.Enp. Biol. Med. 109, 754-756 35. Lorusso, D. J., and Suttie, J. W.(1972) Mol.Pharmacol. 8, 197-203 36. Vainieri, H., and Wingard, L. B., Jr. (1977) J . Pharmacol. Exp. Ther. 201,507-517 37. Bell, R. G., and Ren, P. (1981) Biochem. Pharmacol. 30, 1953-1958 38. Lowenthal, J., and Birnbaum, H. (1969)Science (Wash. DC) 164, 181-183 39. Friedman, P. A,, andShia, M. (1976) Biochem. Biophys. Res. Commun. 70,647-654


3696-3699 5. Stenflo, J., Fernlund,P., Egan, W., and Reopstorff, P. (1974) Proc. Natl. Acad. Sci. U. S. A . 71,2730-2733 6. Nelsestuen, G. L., Zytkovicz, T. H., and Howard, J. B. (1974) J. Biol. Chem. 249, 6347-6350 7. Stenflo, J . (1976) J. Biol. Chem. 251,355-363 8. Di Scipio, R. G., Hermodson, M. A., Yates, S. G., and Davie, E. W. (1977) Biochemistry 16,698-706 9. DiScipio, R. G.,and Davie, E. W. (1979)Biochemistry 18,899-904 10. Hauschka, P. V., Lian, J . B., and Gallop, P. M. (1975) Proc. Natl. Acad. Sci. U. S. A. 72, 3925-3929 11. Hauschka, P. V., Friedman, P. A., Traverso, H. P., and Gallop, P. M. (1976) Biochem. Biophys. Res. Commun. 71, 1207-1213 12. Bell, R. G. (1980) in Vitamin K Metabolism and Vitamin Kdependent Proteins (Suttie, J. W., ed) pp. 286-293, University Park Press, Baltimore 13. Buchthal, S. D., and Bell, R. G. (1980) in Vitamin K Metabolism and VitaminK-dependentproteins (Suttie, J . W., ed) pp. 299-302, University Park Press, Baltimore 14. Friedman, P. A., Hauschka, P. V., Shia, M. A,, and Wallace, J. K. (1979) Biochim. Biophys. Acta 583,261-265 15. Whitlon, D. S., Sadowski, J . A., and Suttie, J. W. (1978) Biochemistry 17, 1371-1377 16. Fasco, M. J., and Principe, L. M. (1980) Biochem. Biophys. Res. Commun. 97, 1487-1492 17. Sadowski, J. A. Esmon, C. T., and Suttie, J. W. (1976) J . Biol. Chem. 251,2770-2776 18. Wallin, R., Gebhardt, O., and Prydz, H. (1978) Biochem. J. 169, 95-101 19. Sadowski, J. A., Schnoes, H. K., and Suttie, J. W. (1977) Biochemistry 16, 3856-3863 20. Bell, R.G., and Caldwell, P. T. (1973)Biochemistry 12, 1759-1762 21. Matschiner, J . T., Bell, R. G., Amelotti, J . M., and Knauer, T. E.

4901

R- and S-Warfarin inhibition of vitamin K and vitamin K 2,3-epoxide reductase activities in the rat. M J Fasco and L M Principe J. Biol. Chem. 1982, 257:4894-4901.

Access the most updated version of this article at http://www.jbc.org/content/257/9/4894 Alerts: • When this article is cited • When a correction for this article is posted Click here to choose from all of JBC's e-mail alerts Downloaded from http://www.jbc.org/ by guest on October 1, 2017

This article cites 0 references, 0 of which can be accessed free at http://www.jbc.org/content/257/9/4894.full.html#ref-list-1