Vitamin K-dependent carboxylase DEVELOPMENT OF A PEPTIDE SUBSTRATE*

Rat liver microsomes contain a vitamin K-dependent carboxylase activity that converts specific glutamyl residues of microsomal prothrombin precursor to gamma-carboxyglutamic acid residues. This activity has now been solubilized by treatment with Triton X-100. The pentapeptide, Phe-Leu-Glu-Glu-Val, has been synthesized; and it has been demonstrated that, in the presence of this peptide, the solubilized microsomes catalyze a vitamin K-dependent incorporation of added H14CO3- into a low molecular weight trichloroacetic acid-soluble compound. The carboxylated product has been identified as peptide-bound gamma-carboxyglutamic acid by its chemical stability during acidic and alkaline hydrolysis and by co-chromatography of an alkaline hydrolysate of the product with authentic gamma-carboxyglutamic acid. The conditions for peptide carboxylation appear to be identical with those demonstrated for precursor carboxylation.


SUMMARY
Rat liver microsomes contain a vitamin K-dependent carboxylase activity that converts specific glutamyl residues of a microsomal prothrombin precursor to r-carboxy glutamic acid residues. This activity has now been solubilized by treatment with Triton X-100. The pentapeptide, Phe-Leu-Glu-Glu-Val, has been synthesized; and it has been demonstrated that, in the presence of this peptide, the solubilized microsomes catalyze a vitamin K-dependent incorporation of added HWO,-into a low molecular weight trichloroacetic acid-soluble compound.
The carboxylated product has been identified as peptide-bound ycarboxyglutamic acid by its chemical stability during acidic and alkaline hydrolysis and by co-chromatography of an alkaline hydrolysate of the product with authentic ycarboxyglutamic acid. The conditions for peptide carboxylation appear to be identical with those demonstrated for precursor carboxylation.
Vitamin K functions in the postribosomal modification of liver microsomal protein precursors to form biologically active prothrombin and the other vitamin K-dependent plasma clotting proteins, Factors VII, IX, and X (1,2). This modification involves the formation of y-carboxyglutamic acid residues (3)(4)(5) in these proteins by the carboxylation of specific glutamyl residues in the precursor proteins and we have developed (6,7) a vitamin K-dependent in vitro carboxylase system to study this reaction. This microsomal carboxylase has now been solubilized by detergent treatment (8) and its requirements described. The carboxylase requires the reduced form of vitamin K (or vitamin K and NADH), HC03-, and 0,. It has not been possible to demonstrate a requirement for a nucleoside triphosphate or a biotin-dependent enzyme. Investigations of the mechanism of this reaction and fractionation of the proteins involved in catalyzing the carboxylation have been hampered by the use of the endogenous microsomal precursor protein(s) (9) as a substrate. This report describes the development of a soluble synthetic peptide which serves as a substrate for the carboxylase . ' $ To whom all inquiries should be addressed. 1 The vitamin K-dependent incorporation of added H"CO,-into trichloroacetic acid-precipitable protein is referred to as "carboxylase" activity or "vitamin K-dependent carboxylase." Whether the carboxylating species in the reaction is CO, or HCO,-is not known, nor is it known if the glutsmyl residues are derivatized prior to HCO,-(CO,) attack. Amino acid analyses were obtained on a Technic01 TSM amino acid analyzer.

RESULTS
Structural studies of bovine prothrombin have established that the ten y-carboxyglutamic acid residues are located at residues 7, 8, 15, 17, 20, 21, 26, 27, 30, 33 (5). Inspection of the published sequence does not reveal any unique sequences that might serve as recognition signals for the carboxylase, and a decision was made to use the pentapeptide, Phe-Leu-Glu-Glu-Val," corresponding to residues 5 to 9 of the bovine prothrombin precursor as a substrate for the carboxylase. When Tritonsolubilized microsomes from vitamin K-deficient rats were incubated with Hr4C0,-under conditions (8) previously shown to result in the carboxylation of endogenous prothrombin precursor, the incorporation of radioactivity into any non-trichloroacetic acid-precipitable nonvolatile material was low, and it was not dependent on the addition of vitamin K. However, when the peptide was added to the incubation mixture (Table   I), there was a 5-to lo-fold stimulation of radioactive incorporation into non-protein material in the presence of the vitamin, but no additional incorporation in the absence of the vitamin.
The data in Fig. 1  The 'material (70% of the radioactivity) in the first radioactive peak eluted from the column lost 50% of its radioactivity upon 6 N HCl hydrolysis, while the remainder of the material eluted from the column did not. The first radioactive peak was concentrated by lyophilization and after addition of 5 mg of unlabeled peptide, it was rechromatographed on the same Bio-Gel P-2 system. The radioactivity (Fig. 2) emerged ' The pentapeptide, Phe-Leu-Glu-Glu-Val, is referred to as peptide. as a single peak which eluted slightly ahead of the added peptide. It has previously been shown (16) that y-carboxyglutamic acid is stable to alkaline hydrolysis. The material in the peak tubes (Fig. 2) was subjected to acid or base hydrolysis; and, after the addition of carrier glutamic acid or y-carboxyglutamic acid, the digests were subjected to thin layer chromatography.
The data (Fig. 3) indicate that most of the radioactivity in the alkaline hydrolysate co-chromatographed with y-carboxyglutamic acid and that following acid (1 ml) of the same sample as inA and 12 N HCl were mixed and heated at 100" for 8 h. The sample was evaporated to dryness and redissolved i n H20. Aliquots of the hydrolysate and authentic glutamic acid were co-chromatographed and detected as described i n A.
hydrolysis the remaining radioactivity was associated with glutamic acid.
Previous studies (7,8) have demonstrated that carboxylation of the endogenous microsomal prothrombin precursor requires vitamin K and NADH (or vitamin K hydroquinone) and that the carboxylation is not inhibited by the addition of the ATP analog, AMP-P(NH)P.
The data in Table II indicate that carboxylation of the peptide is subject to the same responses. Carboxylation does not proceed in the absence of NADH and the data in Table II suggest that ATP is not required in the reaction.
The vitamin K-dependent carboxylation of endogenous proteins in this system has been shown (6) to be minimal when microsomes are prepared from normal rats, presumably because of the low level of precursor (9) in these animals. The data in Table III confirm this observation and indicate that the peptide carboxylase activity is also higher if the enzyme is prepared from vitamin K-deficient rat liver microsomes. The increase in peptide carboxylation in deficient rats was not, however, as marked as the increase in protein carboxylation. The data in Table III also indicate that the total vitamin Kdependent carboxylation of the peptide was about three times higher than that of the endogenous protein acceptors, and that the addition of the peptide did not substantially inhibit the carboxylation of the endogenous precursor.

DISCUSSION
These data have established that a low molecular weight peptide which contains glutamic acid residues will serve as a substrate for this unique carboxylase system. It is likely that other peptides will be found to be better substrates for the enzyme, but Phe-Leu-Glu-Glu-Val is a sufficiently good substrate to be useful in studies of the enzyme. Little progress can be made in purifying the protein(s) involved in the carboxylase activity or in delineating the molecular mechanism of the carboxylation event without a suitable substrate with which to follow the reaction. This demonstration of the availability of such a compound should greatly aid these studies. The lack of an ATP dependence in the solubilized precursor carboxylation system (8) strongly suggested that Coenzyme A esters of the glutamyl residues involved were not being formed prior to carboxylation.
There is, however, evidence (17) that the precursor protein which is being carboxylated may be much more basic than the protein which has been isolated and characterized (91, and this raises the possibility that the endogenous precursors were in some manner "activated" before the vitamin was added. The data presented here (Table II) appear to rule out the possibility of thioester formation and give support to the hypothesis that the carboxylation reaction may be driven by the vitamin hydroquinone.
Alternatively, the vitamin may function as a CO,(HCO,J carrier is this reaction. The availability of a substrate such as the one described here should greatly aid studies into both the role of the vitamins in this reaction, and the molecular details of the carboxylase reaction.