Vitamin K-dependent carboxylase: effect of endogenous microsomal protein precursors on the rate of exogenous substrate carboxylation

Rat liver microsomes contain a Triton X-100 solubilizable vitamin K-dependent carboxylase activity that converts specific glutamyl residues of a microsomal prothrombin precursor to gamma-carboxyglutamyl residues. This activity has been studied in partially (0.25% Triton X-100) and completely (1.0% Triton X-100) solubilized rat liver microsomal preparations. The rate of vitamin K-dependent carboxylation of endogenous microsomal protein precursors was very rapid in the completely solubilized liver microsomal preparation, and carboxylation of an exogenous peptide substrate (Phe-Leu-Glu-Glu-Leu) proceeded at the same time. In the partially solubilized liver microsomal preparation, the rate of protein carboxylation was greatly reduced, and a lag in carboxylation of the exogenous substrate was observed. When microsomal preparations which were depleted of endogenous precursors were used, this lag was eliminated. These data suggest that both substrates utilize the same microsomal pool of carboxylase and that the fraction of the carboxylase bound to the endogenous precursors is not immediately available to exogenous substrates.     

Kinetics of carboxylation of endogenous and exogenous substrates by the vitamin K-dependent carboxylase

The vitamin K-dependent carboxylation of the exogenous pentapeptide, Phe-Leu-Glu-Glu-Ile, and endogenous liver microsomal protein was studied in solubilized rat liver microsomes. The MnCl2 stimulation of the vitamin K-dependent pentapeptide carboxylation rate, which is conducted at subsaturating concentrations of pentapeptide, is due to the cation's ability to lower the Km of the substrate. Although there are clear kinetic differences observed between the carboxylation rates for the pentapeptide and the endogenous protein substrates, several lines of evidence suggest that the same carboxylase system is responsible for both. These points of evidence are (i) the initial velocity of endogenous protein carboxylation is lowered in the presence of 3 mM pentapeptide; (ii) the presence of endogenous microsomal protein substrate causes an initial lag in pentapeptide carboxylation; and (iii) this initial lag phase is not observed when the total endogenous substrate pool is carboxylated by a preincubation reaction prior to the addition of pentapeptide.     

Characteristics of the vitamin K-dependent carboxylating system in human placenta.

Gamma-carboxyglutamic acid, formed during the post-translational vitamin K-dependent carboxylation of glutamic acid residues in polypeptides has been identified not only in coagulation factors II (prothrombin),, VII, IX and X [1--4], but also in several other plasma proteins [3,5,6] and in protein of bone [7,8] and kidney [9]. In rat liver, carboxylation is mediated through an enzyme system located in the microsomal membrane [10]. The enzyme system requires CO2, O2 and the reduced (hydroquinone) form of the vitamin, as well as a suitable substrate [10,11]. Rat liver microsomes also convert vitamin K1 (phylloquinone) to its stable 2,3-epoxide [12]. Several studies suggest a link between carboxylation and the formation of the epoxide [12--14]. In one of these [14], a survey of rat tissues for vitamin K1 epoxidation revealed that, in addition to liver, this activity was also possessed by kidney, bone, spleen and placenta. In preliminary experiments, vitamin K-dependent carboxylating systems have been found in rat and chick kidney [9], in chick bone [15] and in rat spleen and placenta (unpublished observations). In this communication, we describe some of the basic characteristics of the vitamin K-dependent carboxylating system as found in human placental microsomes.     

Characterization of vitamin K-dependent carboxylase from the livers from the adult ox and dicoumarol-treated calf.

The properties of the microsomal vitamin K-dependent carboxylase from the livers of the adult ox and dicoumarol-treated calf were investigated. The enzymes from both sources utilized glutamic residues of synthetic peptides as substrates and could be solubilized with Triton X-100 similarly to the enzyme from vitamin K-deficient rat liver. Under the optimal assay conditions, the microsomes from calf liver had peptide carboxylase activity comparable with that of the rat liver microsomes and 6.5-fold that of adult ox liver microsomes. The apparent Km for reduced vitamin K and the ionic strength optima of the calf and adult ox enzyme clearly differ from those of the rat enzyme. Pyridoxal phosphate activated the adult ox carboxylase only slightly, whereas the calf enzyme was activated by pyridoxal phosphate as effectively as was the enzyme from the vitamin K-deficient rat. Mn2+ activated the adult ox enzyme 9-fold and calf enzyme 22-fold under optimal conditions (no KCl). Three other divalent metal cations (Ca2+, Ba2+, and Mg2+) activated the adult ox and calf enzymes to about half the extent caused by Mn2+, KCl inhibited this activation. The vitamin K-dependent carboxylase from the dicoumarol-treated calf is apparently more tightly bound to the microsomal membrane than is the adult ox enzyme. In many other respects (pH optimum), temperature optimum, Km values for peptide substrate, substrate specificity, inhibitor effects), the properties of the adult ox and calf enzymes resemble closely those of the rat enzyme.     

Vitamin K-dependent carboxylase: effect of detergent concentrations, vitamin K status, and added protein precursors on activity

Activity of the rat liver microsomal vitamin K-dependent carboxylase has been studied at various concentrations of detergent. The activity which could be solubilized by 0.25% Triton X-100 was low but could be greatly increased if vitamin K-deficient rats were given vitamin K a few minutes before they were killed. At higher concentrations of Triton, more activity was solubilized and this effect was not seen. In vitro carboxylation of endogenous microsomal proteins was decreased by 80-90% if vitamin K was administered 1 min before rats were killed, but the amount of assayable prothrombin precursor was decreased by only 20%. Decarboxylated vitamin K-dependent rat plasma proteins were not substrates for the carboxylase and did not influence peptide carboxylase activity significantly. Purified microsomal prothrombin precursors did, however, stimulate carboxylation of peptide substrate and were used as a substrate for the carboxylase in a preparation from precursor depleted vitamin K-deficient rats.     

Vitamin K-dependent carboxylation of glutamic acid residues to γ-carboxyglutamic acid in lung microsomes

Vitamin K-dependent carboxylation of glutamic acid residues to γ-carboxyglutamic acid was demonstrated in proteins of lung microsomes. The carboxylation was 12% of that in liver microsomes per milligram of mierosomal protein. Carboxylation was very low with microsomes of untreated rats but increased with time up to 42 h after warfarin administration. Carboxylation was highest with microsomes from rats fed a vitamin K-deficient diet. This suggests that a protein(s) accumulates which can be carboxylated in vitro/J. Lung microsomes also catalyzed the vitamin K-dependent carboxylation of the peptide Phe-Leu-Glu-Glu-Leu. The peptide carboxylase activity was 9% of that obtained with liver microsomes. Vitamin K-dependent protein carboxylation required NADH or dithioerythritol, suggesting that vitamin K had to be reduced to the hydroquinone. Accordingly, vitamin K₁ hydroquinone had carboxylating activity without added reducing agents. Menaquinone-3 was considerably more active than phylloquinone. The temperature optimum for carboxylation was around 27 °C.     

Vitamin K-dependent carboxylation of synthetic substrates. Nature of the products

Vitamin K-dependent carboxylation of synthetic Phe-Leu-Glu-Glu-Val by solubilized rat liver microsomes yields a predominant product Phe-Leu-Gla-Glu-Val, as determined by decarboxylation and acid or enzymatic hydrolysis. A small amount of another monocarboxylated product, yet unidentified, is also detected. This product is formed from Phe-Leu-Gla-Glu-Val in an enzymatic vitamin K-independent reaction.     

The vitamin K-dependent carboxylation system in human osteosarcoma U2-OS cells. Antidotal effect of vitamin K1 and a novel mechanism for the action of warfarin.

An osteoblast-like human osteosarcoma cell line (U2-OS) has been shown to possess a vitamin K-dependent carboxylation system which is similar to the system in human HepG2 cells and in liver and lung from the rat. In an 'in vitro' system prepared from these cells, vitamin K1 was shown to overcome warfarin inhibition of gamma-carboxylation carried out by the vitamin K-dependent carboxylase. The data suggest that osteoblasts, the cells involved in synthesis of vitamin K-dependent proteins in bone, can use vitamin K1 as an antidote to warfarin poisoning if enough vitamin K1 can accumulate in the tissue. Five precursors of vitamin K-dependent proteins were identified in osteosarcoma and HepG2 cells respectively. In microsomes (microsomal fractions) from the osteosarcoma cells these precursors revealed apparent molecular masses of 85, 78, 56, 35 and 31 kDa. When osteosarcoma cells were cultured in the presence of warfarin, vitamin K-dependent 14C-labelling of the 78 kDa precursor was enhanced. Selective 14C-labelling of one precursor was also demonstrated in microsomes from HepG2 cells and from rat lung after warfarin treatment. In HepG2 cells this precursor was identified as the precursor of (clotting) Factor X. This unique 14C-labelling pattern of precursors of vitamin K-dependent proteins in microsomes from different cells and tissues reflects a new mechanism underlying the action of warfarin.     

Vitamin K-dependent gamma-glutamyl carboxylase activity in the chick embryonic chorioallantoic membrane.

During embryonic development of the chick, the onset of calcium transport by the chorioallantoic membrane (CAM) is concomitant with the appearance of a calcium-binding protein (CaBP). The development-specific expression of the CaBP in the CAM is inhibited by vitamin K antagonism in ovo with the anticoagulant, warfarin. However, the CaBP remains immunologically detectable in the CAM of warfarin-treated embryos, suggesting the presence of a precursor form of the CaBP. Previously, we have demonstrated that CaBP expression in CAM organ cultures is inducible by vitamin K. Furthermore, the CaBP contains several residues of the modified amino acid, gamma-carboxyglutamic acid (gamma-CGlu), which has been shown to be formed by vitamin K-dependent carboxylation of glutamic acid in several plasma clotting proteins. This study reports the presence of a post-translational, vitamin K-dependent gamma-glutamyl carboxylase activity in the CAM. Our results show that explants of CAM incorporate H14CO3 in an age-specific and vitamin K-dependent manner. Incorporation of H14CO3 by the CAM is further potentiated by warfarin treatment of the embryos, presumably owing to an elevation of the amount of endogenous uncarboxylated protein precursor(s). Among the subcellular (nuclear, mitochondrial, microsomal, and soluble) fractions of the CAM, only microsomes exhibit specific incorporation of of H14CO3 into gamma-CGlu. The CAM microsomal carboxylation activity is post-translational, vitamin K-dependent, specific for prenylated homologs of vitamin K, sensitive to warfarin, and appears to be unrelated to the activities of biotin-dependent carboxylases or phosphoenolpyruvate carboxykinase. Optimal carboxylation activity occurs after incubation of the microsomes with H14CO3 for 60 min at 37 degrees C in the presence of over 100 microgram of vitamin K1/ml.     

Vitamin K-dependent carboxylase: Evidence for cofractionation of carboxylase and epoxidase activities,and for carboxylation of a high-molecular-weight microsomal protein

The vitamin K-dependent carboxylase from rat liver microsomes has been fractionated by submitting a crude preparation of this activity to chromatography on different column supports. A constant ratio of vitamin K epoxidation and vitamin K-dependent carboxylation was observed in all column fractions with good carboxylase activity, supporting the hypothesis that these two activities are carried out by the same enzyme complex. The preparation obtained (Complex B) is stable for several days when left on ice and has the same general properties as those observed in Triton X-100-solubilized microsomes. When antiserum raised against Complex B was incubated with Complex B, a twofold increase in carboxylase activity was observed. Benzidine staining showed that an appreciable pool of the antibody population was directed against hemeprotein(s). These data and spectral analyses indicated that a major contaminant of the preparation in cytochrome P-450. Although endogenous prothrombin precursors were absent in the crude starting preparation, a constant ratio of endogenous substrate carboxylation and carboxylation of a soluble substrate was observed during fractionation. A protein with a molecular weight of approximately 120,000 which copurified with Complex B was identified as substrate for the carboxylase.     

Mechanism of action of t-butyl hydroperoxide in the inhibition of vitamin K-dependent carboxylation

t-Butyl hydroperoxide has been studied as a possible competitive inhibitor of the vitamin K-dependent carboxylation of the pentapeptide PheLeuGluGluIle. Under standard carboxylating conditions the concentrations of reduced phylloquinone and phylloquinone were followed by high-pressure liquid chromatography during 30-min incubations of Triton-solubilized microsomes from rat liver. Under these conditions supporting linear rates of carbon dioxide fixation for 20–30 min, the vitamin KH₂ concentration decreased exponentially to less than 5% of its initial value in 30 min principally due to autooxidation. In the presence of 10 mm t-butyl-OOH, however, the oxidation of vitamin KH₂ was greatly accelerated with none being detected after 7 min. In general, the rate of carboxylation of peptide paralleled the KH₂ concentration. After cessation of carboxylation in the presence of t-butyl-OOH the readdition of KH₂ stimulated additional ¹⁴CO₂ fixation. A known competitive inhibitor of vitamin K, 2-chlorophylloquinone, did not accelerate the oxidation of KH₂ but nonetheless inhibited the vitamin K-dependent carboxylation in a competitive manner. These data have led us to conclude that t-butyl-OOH is not a competitive inhibitor of the vitamin K-dependent carboxylase at the active site of the enzyme but merely acts to promote the oxidation of KH₂.     

Griseofulvin-induced aggregation of microtubule protein.

The mechanism of the vitamin K-dependent post-translational carboxylation of the gamma-carbon atom of glutamic acid residues in proteins remains obscure. Experiments were performed in vivo and in vitro in an attempt to establish a role for biotin in the transfer of the carboxyl group. Weanling male rats were fed on a biotin-deficient diet until severe biotin deficiency was induced. Their degree of biotin deficiency was documented by assaying for liver acetyl-CoA carboxylase activity, which was about 15% of normal. However, one-stage and two-stage prothrombin times measured on the plasmas were normal. In addition, the liver microsomal fraction did not contain any more prothrombin precursor than did that of normal rat liver. Experiments were done in vitro in which vitamin K-dependent fixing of 14CO2 was measured in the liver microsomal fraction from vitamin K-deficient male rats in the presence or absence of avidin. No evidence for an avidin-sensitive critical biotin-containing site was obtained. Thus neither series of experiments suggests a role for biotin; the data are compatible with carboxyl transfer occurring either through a carboxylated vitamin K intermediate; or via a yet to be identified intermediate, or perhaps via CO2 itself.     

Solubilization and characterization of vitamin K epoxide reductase from normal and warfarin-resistant rat liver microsomes

Two procedures have been developed for the solubilization of vitamin K epoxide reductase from rat liver microsomal membranes using the detergent Deriphat 160 at pH 10.8. The methods are applicable to both normal and Warfarin-resistant-strain rat liver microsomes and yield material suitable for further purification. The preparations retain dithiothreitol-dependent vitamin K quinone reductase activity as well as vitamin K epoxide reductase and are free of vitamin K-dependent carboxylase and epoxidase activities. Optimal epoxide reductase activity is obtained at 0.1 M KCl and pH 9 in the presence of sodium cholate. Artifactual formation of vitamin K metabolites was eliminated through the use of mercuric chloride to remove excess dithiothreitol prior to extraction and metabolite assay. Using the solubilized enzyme, valid initial velocities were measured, and reproducible kinetic data was obtained. The substrate initial velocity patterns were determined and are consistent with a ping-pong kinetic mechanism. The kinetic parameters obtained are a function of the cholate concentration, but do not vary drastically from those obtained using intact microsomal membranes. At 0.8% cholate, the enzymes solubilized from normal Warfarin-sensitive- and Warfarin-resistant-strain rat livers exhibit respective values of Vmax = 3 and 0.75 nmol/min/g liver; Km for vitamin K epoxide = 9 and 4 microM; and Km for dithiothreitol of 0.6 and 0.16 mM.     

Vitamin K-dependent carboxylase: requirements for carboxylation of soluble peptide and substrate specificity.

Rat liver microsomes contain a triton X-100 solubilizable vitamin K-dependent carboxylase activity that converts specific glutamyl residues of precursor proteins to γ-carboxyglutamyl residues. This activity has been studied utilizing synthetic peptides as substrates for the enzyme. When compared to the carboxylation of the endogenous microsomal precursors, the peptide carboxylase activity is more sensitive to the action of various inhibitors, and requires a higher concentration of vitamin K for maximal activity. The apparent K_m for the peptide Phe-Leu-Glu-Glu-Leu was found to be 4 mM. Substrate specificity depends on residues adjacent to the carboxylated Glu residues and macromolecular recognition sites.     

The stereochemistry of hydrogen abstraction in vitamin K-dependent carboxylation

The stereochemistry of the hydrogen abstraction in the vitamin K-dependent carboxylation of synthetic peptides has been investigated; the carboxylation rates of various peptidic substrates containing a stereospecifically 4-monodeuterated glutamic acid residue have been compared to that of nondeuterated peptides. A significant isotope effect was found only with the substrates containing (4S)-4-deuterated glutamic acid. These data reveal that the rat liver microsomal vitamin K-dependent carboxylase acts stereospecifically in abstracting the 4-pro-S hydrogen of the glutamyl residue. The low values of the measured isotope effects indicate that the hydrogen abstraction does not constitute a limiting step in the carboxylation mechanism.     

A conserved motif within the vitamin K-dependent carboxylase gene is widely distributed across animal phyla

The vitamin K-dependent gamma-glutamyl carboxylase catalyzes the posttranslational conversion of glutamic acid to gamma-carboxyglutamic acid, an amino acid critical to the function of the vitamin K-dependent blood coagulation proteins. Given the functional similarity of mammalian vitamin K-dependent carboxylases and the vitamin K-dependent carboxylase from Conus textile, a marine invertebrate, we hypothesized that structurally conserved regions would identify sequences critical to this common functionality. Furthermore, we examined the diversity of animal species that maintain vitamin K-dependent carboxylation to generate gamma-carboxyglutamic acid. We have cloned carboxylase homologs in full-length or partial form from the beluga whale (Delphinapterus leucas), toadfish (Opsanus tau), chicken (Gallus gallus), hagfish (Myxine glutinosa), horseshoe crab (Limulus polyphemus), and cone snail (Conus textile) to compare these structures to the known bovine, human, rat, and mouse cDNA sequences. Comparison of the predicted amino acid sequences identified a nearly perfectly conserved 38-amino acid residue region in all of these putative carboxylases. In addition, this amino acid motif is also present in the Drosophila genome and identified a Drosophila homolog of the gamma-carboxylase. Assay of hagfish liver demonstrated vitamin K-dependent carboxylase activity in this hemichordate. These results demonstrate the broad distribution of the vitamin K-dependent carboxylase gene, including a highly conserved motif that is likely critical for enzyme function. The vitamin K-dependent biosynthesis of gamma-carboxyglutamic acid appears to be a highly conserved function in the animal kingdom.     

Vitamin K epoxidase activity of rough and smooth microsomes from rat liver

The distribution of vitamin K epoxidase activity in rough and smooth microsomes has been studied and compared to the prothrombin precursor and vitamin K-dependent carboxylase activity. All three activities were high in rough microsomes as compared to the low levels found in smooth microsomes. The results are in agreement with the suggestion that there might be a linkage between the vitamin K-dependent carboxylation and epoxidation reaction in vivo.     

Vitamin K-dependent carboxylase: minimized escape of CO2 from solution may prolong linearity of the reaction rate

Escape of 14CO2 from the reaction mixture into the gas phase may seriously affect the accuracy of in vitro measurement of vitamin K-dependent carboxylase activity (and probably that of other carboxylases as well). In this paper we describe the effect of (a) the volume of the test tubes in which the reaction is performed, (b) the addition of an excess of NaH12CO3 in parallel with standard amounts of NaH14CO3, and (c) the incubation temperature. In this way optimal conditions are defined and used for the carboxylation of various peptide and protein substrates. It is shown that both a prosequence and an internal recognition site contribute to the effective recognition of a substrate by carboxylase. The maximal efficiency of carboxylation was 1-2% with substrates lacking both signals and 20-50% if only one was present. This indicates the need for developing peptide substrates containing both recognition signals for vitamin K-dependent carboxylase.     

Synthesis of fluoro- and hydroxy-derivatives of vitamin K as substrates or inhibitors of the liver microsomal vitamin K-dependent carboxylase.

The rat liver microsomal vitamin K-dependent carboxylase catalyzes the carboxylation of glutamyl to gamma-carboxyglutamyl residues in the presence of reduced vitamin K, O2 and CO2. The specificity of the enzyme for the vitamin substrate has been probed by the synthesis of a series of fluoro- hydroxy- and methoxy-analogs. 2-Fluoro-methyl-3-phytyl-1,4-naphthoquinone and 2-methyl-3-(1'-fluorodecyl)-1,4-naphthoquinone were synthesized but found to be unstable under enzyme assay conditions. The reduced (naphthohydroquinone) forms of 2-hydroxy-methyl-3-phytyl-1,4-naphthoquinone, 2-methoxymethyl-3-phytyl-1,4-naphthoquinone and 2-methyl-3-(1'-hydroxy-decyl)-1,4-naphthoquinone were inactive as substrates, but inhibitors of the enzyme. The two hydroxy analogs were shown to be low Ki (less than 10 microM) inhibitors of the reduced 2-methyl-3-phytyl-1,4-naphthoquinone-dependent activity of the enzyme. The oxidized forms of these compounds did not inhibit the enzyme and they had no activity as in vivo anticoagulants.     

Synthesis of menaquinone-2 derivatives as substrates for the liver microsomal vitamin K-dependent carboxylase

The rat liver microsomal vitamin K-dependent carboxylase catalyzes the carboxylation of glutamyl to gamma-carboxyglutamyl residues in the presence of reduced vitamin K, O2 and CO2. The specificity of the enzyme for the vitamin substrate has been probed by the synthesis of a number of menaquinone-2 (2-methyl-3-geranyl-1,4-naphthoquinone) derivatives. The 2-des-methyl and 2-ethyl-MK-2 derivatives had very low activity as substrates. The 6- or 7-methyl-MK-2 derivatives and (6,7)-chloro-MK-2 were relatively high Vmax substrates with Km values increased over that seen for K-2. The 5- or 8-methyl-MK-2 derivatives were low Vmax substrates but also demonstrated low Km values. Although these observations suggested that 5-methyl-MK-2 might be a competitive inhibitor of the carboxylation reaction, it was not an effective inhibitor of either phylloquinone or 6-methyl-MK-2-dependent carboxylation.