Nutritional scientist or biochemist?

When invited by the editors to provide a prefatory article for the Annual Review of Nutrition, I attempted to decide what might be unique about my experiences as a nutritional biochemist. Although a large proportion of contemporary nutritional scientists were trained as biochemists, the impact of the historical research efforts related to nutrition within the Biochemistry Department at the University of Wisconsin 50 to 60 years ago was, I think, unique, and I have tried to summarize that historical focus. My scientific training was rather standard, but I have tried to review the two major, but greatly different, areas of research that I have been involved in over my career: inorganic fluorides as an industrial pollutant and the metabolic role of vitamin K. I have also had the opportunity to become involved with the activities of the societies representing the nutritional sciences (American Society for Nutrition), biochemistry (American Society for Biochemistry and Molecular Biology), Federation of American Societies for Experimental Biology, the Food and Nutrition Board, the Board on Agriculture and Natural Resources, and the U.S. Department of Agriculture National Agricultural Research, Extension, Education, and Economics. These interactions can be productive or frustrating but are always time-consuming.

Vitamin K status of premature infants: implications for current recommendations.

OBJECTIVE: Newborn infants are vitamin K deficient. Vitamin K status in full-term infants after intramuscular vitamin K supplementation at birth has been described. Similar information in growing premature infants has not been reported. The objective of this study was to assess vitamin K status in premature infants by measuring plasma vitamin K and plasma protein-induced in vitamin K absence (PIVKA II) from birth until 40 weeks' postconceptional age. METHODS: Premature infants (/=1000 g) via total parenteral nutrition. After hyperalimentation, most received vitamin K-fortified enteral feedings with the remainder receiving unfortified breast milk. Blood was obtained for PIVKA II in cord blood and for PIVKA II and vitamin K at 2 weeks and 6 weeks after birth and at 40 weeks' postconception. RESULTS: Of the 44 infants enrolled, 10 infants in each gestational age group completed the study. The patient characteristics for groups 1, 2, and 3 were as follows: gestational age, 26.3 +/- 1.7, 30.3 +/- 1.3, and 33.9 +/- 1.1 weeks; birth weight, 876 +/- 176, 1365 +/- 186, and 1906 +/- 163 g; and days of hyperalimentation, 28.9 +/- 16, 16.8 +/- 12, and 4.3 +/- 4 days, respectively. At 2 weeks of age, the vitamin K intake and plasma levels were highest in group 1 versus group 3 (intake: 71.2 +/- 39.6 vs 13.4 +/- 16.3 microg/kg/day; plasma levels: 130.7 +/- 125.6 vs 27.2 +/- 24.4 ng/mL). By 40 weeks' postconception, the vitamin K intake and plasma levels were similar in all 3 groups (group 1, 2, and 3: intake, 11.4 +/- 2.5, 15.4 +/- 6.0, and 10.0 +/- 7.0 microg/kg/day; plasma level, 5.4 +/- 3.8, 5.9 +/- 3.9, and 9.3 +/- 8.5 ng/mL). None of the postnatal plasma samples had any detectable PIVKA II. CONCLUSIONS: Premature infants at 2 weeks of age have high plasma vitamin K levels compared with those at 40 weeks' postconceptional age secondary to the parenteral administration of large amounts of vitamin K. By 40 weeks' postconception, these values are similar to those in term formula-fed infants. Confirming "adequate vitamin K status," PIVKA II was undetectable by 2 weeks of life in all of the premature infants. With the potential for unforeseen consequences of high vitamin K levels, consideration should be given to reducing the amount of parenteral vitamin K supplementation in the first few weeks of life in premature infants.vitamin K, PIVKA II, premature, total parenteral nutrition, enteral nutrition.

Vitamin K supplementation reduces serum concentrations of under-gamma-carboxylated osteocalcin in healthy young and elderly adults.

BACKGROUND: Subclinical vitamin K insufficiency, manifested by under-gamma-carboxylation of the bone matrix protein osteocalcin, may be common. OBJECTIVE: Our objective was to delineate the prevalence of submaximal gamma-carboxylation as assessed by response to phylloquinone supplementation and to evaluate the effect of this intervention on skeletal turnover in healthy North American adults. DESIGN: Healthy subjects (n = 219), approximately equally distributed by sex and age (18-30 y and >/=65 y), received daily phylloquinone (1000 microg) or placebo for 2 wk. Serum undercarboxylated osteocalcin (ucOC) and total osteocalcin, N:-telopeptides of type I collagen (NTx), bone-specific alkaline phosphatase (BSAP), and phylloquinone concentrations were measured at baseline and after weeks 1 and 2. RESULTS: At baseline, the mean serum phylloquinone concentration was lower in the young than in the old group; there was no effect of sex. Concomitantly, baseline %ucOC was highest in the young and lowest in the old men (P: < 0.0001) but did not differ significantly by age in women. After supplementation, serum phylloquinone concentration increased approximately 10-fold (P: < 0.0001) at week 1 (from 0.93 +/- 0.08 to 8.86 +/- 0.70 nmol/L, x+/- SEM); this was sustained through week 2. Among all supplemented groups, mean %ucOC decreased from 7.6% to 3. 4% without significant differences by age or sex; 102 of 112 subjects had a >1% decrease. Phylloquinone supplementation reduced serum osteocalcin but did not alter NTx or BSAP concentration. CONCLUSIONS: Usual dietary practices in this population did not provide adequate vitamin K for maximal osteocalcin carboxylation. Phylloquinone supplementation reduced serum osteocalcin concentration but did not alter other markers of serum bone turnover.

N-glycosylation contributes to the intracellular stability of prothrombin precursors in the endoplasmic reticulum.

Rat prothrombin (rFII) and human prothrombin (hFII) are processed differently during their biosynthesis in a manner dependent upon gamma-glutamyl carboxylation, a vitamin K-dependent posttranslational modification of prothrombin precursors. The role of N-glycosylation in the cellular processing of prothrombin was examined in hepatoma (H-35 and HepG2) and transformed kidney (HEK293) cell lines. Aglyco-rFII obtained by tunicamycin treatment was degraded in warfarin-treated H-35 cells, but not in vitamin K-treated cells. Fully glycosylated rFII is also selectively retained and degraded in warfarin-treated H-35 cells. When rFII and hFII were transiently expressed in tunicamycin-treated HEK293 cells, rFII but not hFII was degraded to generate a 48-KD species. This degradation was independent of gamma-carboxylation, indicating that the sensitivity of aglyco-rFII and aglyco-hFII toward this specific proteolysis differs in HEK293 cells. By expressing chimeric rFII/hFII constructs in tunicamycin-treated HEK293 cells, it was shown that the kringle 2 structure of prothrombin was responsible for this difference. The 48-KD species was not observed in tunicamycin-treated H-35 cells, suggesting that this specific proteolytic processing of aglyco-rFII is also cell-type specific. Expression of rFII in other tunicamycin-treated nonhepatic cells suggests that this cell-specific difference in processing might be determined by a difference between hepatic and nonhepatic cells. The site-directed mutagenesis utilized in these studies also establishes that the N-glycosylation sites of rFII are at residues Asn77, 101, 378, and 518.

Comparison of phylloquinone bioavailability from food sources or a supplement in human subjects.

Phylloquinone (K) absorption was assessed in 22- to 30-y-old human subjects consuming a standard test meal [402 kcal (1682 kJ), 27% energy from fat]. The absorption of phylloquinone, measured over a 9-h period as the area under the curve (AUC), was higher (P < 0.01) after the consumption of a 500- microgram phylloquinone tablet [27.55 +/- 10.08 nmol/(L. h), n = 8] than after the ingestion of 495 microgram phylloquinone as 150 g of raw spinach [4.79 +/- 1.11 nmol/(L. h), n = 3]. Less phylloquinone (P < 0.05) was absorbed from 50 g of spinach (AUC = 2.49 +/- 1.11 nmol/(L. h) than from 150 g of spinach. Absorption of phylloquinone from fresh spinach (165 microgram K), fresh broccoli (184 microgram K) and fresh romaine lettuce (179 microgram K) did not differ. There was no difference in phylloquinone absorption from fresh or cooked broccoli or from fresh romaine lettuce consumed with a meal containing 30 or 45% energy as fat.

Dietary intake and adequacy of vitamin K.

The current daily recommended dietary allowance for vitamin K is 1 microg/kg. Reliable measurements of vitamin K content in foods are now available, and data from 11 studies of vitamin K intake indicate that the mean intake of young adults is approximately 80 microg phylloquinone/d and that older adults consume approximately 150 microg/d. The vitamin K concentration in most foods is very low (<10 microg/100 g), and the majority of the vitamin is obtained from a few leafy green vegetables and four vegetable oils (soybean, cottonseed, canola and olive) that contain high amounts. Limited data indicate that absorption of phylloquinone from a food matrix is poor. Hydrogenated oils also contain appreciable amounts of 2', 3'-dihydrophylloquinone of unknown physiological importance. Menaquinones absorbed from the diet or the gut appear to provide only a minor portion of the human daily requirement. Measures of the extent to which plasma prothrombin or serum osteocalcin lack essential gamma-carboxyglutamic acid residues formed by vitamin K action, or the urinary excretion of this amino acid, provide more sensitive measures of vitamin K status than measures of plasma phylloquinone or insensitive clotting assays.

Conversion of dietary phylloquinone to tissue menaquinone-4 in rats is not dependent on gut bacteria.

The ability of male rats to accumulate menaquinone-4 (MK-4) in tissues when fed a vitamin K-deficient diet supplemented with intraperitoneal phylloquinone (K) as the sole source of vitamin K for 14 d was assessed. In both conventionally housed controls and gnotobiotic rats, supplementation with the equivalent of 1500 microg vitamin K/kg diet increased (P < 0.001) tissue MK-4 concentrations above those of controls fed a vitamin K-deficient diet. MK-4 concentrations were approximately 5 ng/g (11 pmol/g) in liver, 14 ng/g in heart, 17 ng/g in kidney, 50 ng/g in brain and 250 ng/g in mandibular salivary glands of gnotobiotic rats. MK-4 concentrations in conventionally housed rats were higher than in gnotobiotic rats in heart (P < 0.01), brain (P < 0.01) and kidney (P < 0.05) but lower in salivary gland (P < 0.05). Cultures of a kidney-derived cell line (293) converted K to the expoxide of MK-4 in a manner that was dependent on both time of incubation and concentration of vitamin K in the media. A liver-derived cell line (H-35) was less active in carrying out this conversion. These data offer conclusive proof that the tissue-specific formation of MK-4 from K is a metabolic transformation that does not require bacterial transformation to menadione as an intermediate in the process.

Improving the vitamin K status of breastfeeding infants with maternal vitamin K supplements.

OBJECTIVE: To increase the phylloquinone (vitamin K1) concentration of human milk with maternal oral phylloquinone supplements such that both the phylloquinone intake of breastfed infants and their serum concentrations of phylloquinone would approach those of formula-fed infants who are known to be at much less risk for hemorrhagic disease of the newborn. DESIGN: Two stages: stage I, longitudinal, randomized study of 6 weeks' duration; and stage II, longitudinal, randomized, double-blind, placebo-controlled study of 12 weeks' duration. SETTING: Patients from a private pediatric practice in Madison, WI. PATIENTS: Stage I: sequential sampling of 20 lactating mothers to determine the level of maternal supplementation needed in stage II. Ten mothers received 2.5 mg/d oral phylloquinone, and 10 mothers received 5.0 mg/d oral phylloquinone. Stage II: sequential sampling of 22 human milk-fed infants and lactating mothers. All infants received 1 mg of phylloquinone at birth. Eleven mothers received a placebo; 11 mothers received 5 mg/d phylloquinone. MEASUREMENTS AND RESULTS: In stage I, both 2.5 and 5.0 mg/d phylloquinone significantly increased the phylloquinone content of human milk at both 2 and 6 weeks. As expected, 5.0 mg had a greater effect (mean +/- SD, 58.96 +/- 25.39 vs 27.12 +/- 12.18 ng/mL at 2 weeks). In stage II, the vitamin K-supplemented group had significantly higher maternal serum phylloquinone concentrations, higher phylloquinone milk concentrations, and higher infant plasma phylloquinone concentrations at 2, 6, and 12 weeks compared with the placebo group. At 12 weeks infant phylloquinone intakes were significantly higher for the vitamin K group than the placebo group (9.37 +/- 4.55 vs 0.15 +/- 0.07 microgram/kg per day). This corresponded to a plasma phylloquinone concentration in the vitamin K group of 2.84 +/- 3.09 vs 0.34 +/- 0.57 ng/mL in the placebo group. At 12 weeks, the prothrombin times did not differ between the groups, but the des-gamma-carboxy-prothrombin (partially carboxylated prothrombin thought to be a measure of vitamin K deficiency) was significantly elevated in the placebo group compared with the vitamin K group (1.48 +/- 1.19 vs 0.42 +/- 0.55 ng/mL). CONCLUSION: In exclusively breastfed infants who receive intramuscular phylloquinone at birth, the vitamin K status as measured by plasma phylloquinone and des-gamma-carboxy-prothrombin concentrations is improved by maternal oral supplements of 5 mg/d phylloquinone through the first 12 weeks of life.

Structural features of the kringle domain determine the intracellular degradation of under-gamma-carboxylated prothrombin: studies of chimeric rat/human prothrombin.

Vitamin K antagonists such as warfarin inhibit the vitamin K-dependent gamma-glutamyl carboxylation during protein processing and block the secretion of under-gamma-carboxylated prothrombin (FII) in the rat but not in the human or bovine. Under-gamma-carboxylated prothrombin is also secreted from warfarin-treated human (HepG2) cell cultures but is degraded in the endoplasmic reticulum in warfarin-treated rat (H-35) cell cultures. This differential response to warfarin has been shown to be determined by the structural difference in the proteins rather than by the origin of the cell line. When recombinant rat prothrombin (rFII) and human prothrombin (hFII) were expressed in a transformed human kidney cell line (HEK293), secretion of rFII but not hFII was drastically decreased in response to warfarin. To determine the structural signal required for this differential response, chimeric cDNAs with the propeptide/Gla domains, kringle domain, and serine protease domain exchanged between rFII and hFII were generated (FIIRHH and FIIHRR, FIIRRH and FIIHHR, FIIRHR and FIIHRH) and expressed in both warfarin-treated HEK293 cells and HepG2 cells. The presence of the hFII kringle domain changed the stability of rFII to that of hFII, and the rFII kringle domain changed the stability of hFII to that of rFII. The kringle domain therefore is critical in determining the metabolic fate of under-gamma-carboxylated prothrombin precursors during processing. Prothrombin contains two kringle structures, and expression of additional rFII/hFII chimeras (FIIHrhH and FIIHhrH, FIIRrhR, and FIIRhrR) was used to determine that the first of the two kringles plays a more important role in the recognition process.

Assessment of vitamin K status in human subjects administered "minidose" warfarin.

Vitamin K is required to convert specific glutamyl residues in a limited number of proteins to gamma-carboxyglutamyl residues. The response of various measures of vitamin K insufficiency to the administration of 1 mg/d of the vitamin K antagonist warfarin was studied in two groups of nine older (55-75 y) or younger (20-28 y) subjects. The most consistent and extensive alteration was an increase in the concentration of serum under-gamma-carboxylated osteocalcin followed by increased immunochemical detection of plasma under-gamma-carboxylated prothrombin (PIVKA-II), and by a decreased urinary excretion of gamma-carboxyglutamic acid. Plasma concentrations of prothrombin were altered by this treatment but prothrombin times, factor VII activity, prothrombin F-1 x 2 concentrations, and a less sensitive assay for under-gamma-carboxylated prothrombine were not. The concentration of serum under-gamma-carboxylated osteocalcin was lower when subjects consumed 1 mg vitamin K/d than when they consumed their normal diet.

Vitamin K status influences brain sulfatide metabolism in young mice and rats.

The established role of vitamin K in nutrition is as a cofactor in the post-translational conversion of specific glutamyl to gamma-carboxyglutamyl (Gla) residues in a limited number of proteins. Administration of the vitamin K antagonist warfarin has previously been shown to decrease brain sulfatide concentrations and decrease brain galactocerebroside sulfotransferase (GST) activity in young mice. A dietary deficiency of vitamin K has now been shown to decrease (P < 0.01) brain sulfatide concentrations of 30-d-old mice significantly (by 21%). Male 21-d-old rats fed an excess of vitamin K for 7 or 14 d had 26 and 31% (P < 0.05) greater GST activity and 15 and 18% (P < 0.05) greater brain sulfatide concentrations, respectively, than controls fed a vitamin K-deficient diet. Male 21-d-old rats fed a diet containing 500 mg of phylloquinone/kg diet had an intermediate response and were vitamin K sufficient by normal criteria. The vitamin K response was observed when either phylloquinone or menaquinone-4 was fed as a source of the vitamin. These data suggest that in addition to its recognized role in Gla synthesis, vitamin K status is important in the maintenance of normal complex lipid sulfatide metabolism in young rats and mice.

Differential effects of warfarin on the intracellular processing of vitamin K-dependent proteins.

The vitamin K-dependent carboxylation of specific glutamyl residues to gamma-carboxyglutamyl residues occurs during the endoplasmic reticulum processing of a limited number of proteins. The fate of the under-gamma-carboxylated proteins during protein processing was studied. When human hepatoma (HepG2) cells were grown in the presence of warfarin, under-gamma-carboxylated prothrombin was secreted into the medium. In contrast, prothrombin secretion from a rat hepatoma (H-35) cell line was blocked by warfarin, and intracellular forms which were retained were degraded. When rat prothrombin (rFII) was stably transfected into warfarin treated HepG2 cells, endogenous human prothrombin (hFII) was secreted in an under-gamma-carboxylated form, while rFII accumulated intracellularly. These data indicate that retention and degradation of under-gamma-carboxylated prothrombin by human hepatocytes is related to a structural difference in rFII and hFII. When rFII and hFII were transfected into a warfarin treated transformed human embryonic kidney cell line (293), both proteins were secreted in an under-gamma-carboxylated form and intracellular retention was not observed. However, the secretion of rFII was greatly diminished. Cellular retention of under-gamma-carboxylated forms is therefore tissue specific, but degradation is not.

Comparative distribution, metabolism, and utilization of phylloquinone and menaquinone-9 in rat liver.

The liver of most species contains a spectrum of bacterially produced menaquinone homologs as well as the major dietary form of vitamin K, phylloquinone. The relative utilization of phylloquinone and menaquinone-9 (MK-9) as substrates for the microsomal vitamin K-dependent gamma-glutamyl carboxylase was determined in a rat model. Vitamin K 2,3-epoxide, the co-product of the carboxylation reaction, is recycled to the quinone form of the vitamin by a microsomal vitamin K epoxide reductase. This enzyme activity was blocked by warfarin administration, and the appearance of the hepatic epoxides of phylloquinone and MK-9 was determined as a measure of their utilization by the carboxylase. When the liver contained equimolar amounts of phylloquinone and MK-9, four times as much phylloquinone epoxide as MK-9 epoxide was present in the liver 1 hr after warfarin administration. These data suggest that hepatic MK-9 is not as efficiently utilized as phylloquinone. The data obtained have also demonstrated a previously unrecognized difference in phylloquinone and menaquinone metabolism. MK-9 epoxide, and to a lesser extent MK-9, was preferentially localized in the mitochondria, while higher concentrations of phylloquinone were found in the microsomes.

Multi-site-specificity of the vitamin K-dependent carboxylase: in vitro carboxylation of des-gamma-carboxylated bone Gla protein and Des-gamma-carboxylated pro bone Gla protein.

The vitamin K-dependent carboxylase processes multiple glutamic acid residues to gamma-carboxyglutamic acid (Gla) residues in a limited number of proteins. The targeted proteins are synthesized with an amino-terminal propeptide which has been shown to play an important role in gamma-carboxylation. The specificity of the enzyme for each potential Gla site, the direction of carboxylation, and the influence of a bound propeptide on these events are not understood. Des-gamma-carboxy forms of bone Gla protein (BGP), which contain potential Gla residues at positions 17, 21, and 24, were employed as model substrates to determine the multi-site-specificity of the enzyme. Recombinant bovine des-gamma-carboxylated proBGP (rdproBGP) and heat-decarboxylated BGP (dBGP), lacking a propeptide, were used as substrates for a bovine liver carboxylase, and the in vitro reaction products were analyzed for the formation of 14CO2 Gla. The di-Gla species was found to be the predominant product of in vitro carboxylation of both rdproBGP and dBGP at less than saturating concentrations of each substrate. Carboxylation of both substrates occurred preferentially at the more C-terminal potential Gla sites, residues 21 and 24. A similar pattern of carboxylation was observed with a rat bone cell carboxylase, suggesting no species or tissue variation in the enzyme specificity. Some tricarboxylated product accumulated during carboxylation of rdproBGP but not dBGP, suggesting that the covalently bound propeptide directs more complete carboxylation of the Gla domain. In addition, monocarboxylated rdproBGP was found to accumulate in the absence but not in the presence of a free noncovalently attached propeptide, indicating that free propeptide affects more efficient carboxylation of rdproBGP.

Vitamin K nutrition and osteoporosis.

Although the abundance of vitamin K-dependent proteins in bone suggests an important function, the precise role of vitamin K in skeletal health remains to be determined. Serum concentrations of vitamin K are reportedly reduced in older individuals and persons with osteoporotic fracture. Whether this is causally related to vitamin K insufficiency or simply reflects inadequate nutritional status is unclear. Circulating levels of undercarboxylated osteocalcin may be a sensitive marker of vitamin K inadequacy and have been reported to be increased in both postmenopausal women and individuals who sustained hip fracture. It is also possible that vitamin K indirectly affects the skeleton via control of renal calcium excretion. The effect of vitamin K antagonists (oral anticoagulants) on both renal calcium excretion and bone density is controversial. Thus, many of the reports implicating a role for vitamin K insufficiency in the development of osteoporosis are conflicting. This review summarizes current knowledge regarding a possible role of vitamin K insufficiency in the pathogenesis of osteoporosis.

The importance of menaquinones in human nutrition.

Bacterially produced menaquinones, 2-methyl-1,4-naphthoquinones with an unsaturated polyisoprenoid chain at the 3-position, are biologically active forms of vitamin K that are present in high concentrations in the human lower bowel. Menaquinones are found in human liver and circulate in human plasma at much higher concentrations than previously thought. Numerous case reports of antibiotic-induced, vitamin K-responsive hypothrombinemias have been taken as evidence that menaquinones contribute importantly to satisfying the human vitamin K requirement. However, more recent production of symptoms of vitamin K insufficiency in normal human subjects by dietary restriction of vitamin K argues against their nutritional significance. Current data support the view that menaquinones may partially satisfy the human requirement but that their contribution is much less than previously thought.

Mechanism of the abnormal vitamin K-dependent gamma-carboxylation process in human hepatocellular carcinomas.

BACKGROUND: An important marker for hepatocellular carcinoma is the presence of des-gamma-carboxy (abnormal) prothrombin. However, the molecular basis for the reduced carboxylation of prothrombin is unknown. METHODS: Two groups of patients were defined according to the absence (Group I, n = 7) or presence (Group II, n = 8) of des-gamma-carboxy prothrombin. The enzymatic activity of gamma-carboxylase and the total microsomal prothrombin concentration were determined in all tumors. The kinetic parameters for the synthetic peptide Phe-Leu-Glu-Glu-Leu (FLEEL) were measured in eight tumors. The gamma-carboxylase mRNA expression was evaluated by Northern blot analysis in 12 of 15 tumors. In addition, the total vitamin K content (K1, K1 epoxide, and menaquinones 4-10) in 10 tumors was investigated by high performance liquid chromatography. RESULTS: Concentrations of menaquinones 4-10 were normal in the nontumorous part of the liver but significantly decreased (P = 0.02) in all the tumors (Groups I and II). This decrease was more severe in Group II (P = 0.02). The tumors in Group I had normal or increased gamma-carboxylase activity and increased mRNA expression (P < 0.02) as compared with their nontumorous counterparts. The tumors in Group II were heterogeneous. Five tumors displayed low gamma-carboxylase activity, associated with low mRNA expression in two, whereas two others had high gamma-carboxylase activity and mRNA expression. The concentration of FLEEL at half-maximal velocity was normal in all the tumors examined (Groups I and II), and a relation was found between the level of expression of gamma-carboxylase and the maximal velocity for FLEEL carboxylation in the tumors in Group II (r = 0.98; P < 0.01). The microsomal content of normal prothrombin was within normal limits in all tumors (Groups I and II). CONCLUSIONS: Tumor vitamin K content has a critical role in the synthesis of des-gamma-carboxy prothrombin. Furthermore, the gamma-carboxylase defect, which is observed in some secreting tumors, is the result of the defective gene expression of a normal enzyme and not the consequence of the presence of a competitive inhibitor. It is possible that a 75% reduction in gamma-carboxylase gene expression could take a part in the secretion of des-gamma-carboxy prothrombin, but this mechanism is not predominant.

Prothrombin synthesis and degradation in rat hepatoma (H-35) cells: effects of warfarin.

Vitamin K is a substrate for the enzyme catalyzing the carboxylation of specific glutamyl residues to gamma-carboxyglutamyl residues in hepatic precursors of a limited number of plasma proteins, including prothrombin. The gamma-carboxylation of these proteins can be blocked by the anticoagulant warfarin; and in the bovine and human, warfarin treatment results in the secretion of under-gamma-carboxylated forms of prothrombin into plasma. In the rat, this response is not seen, but plasma prothrombin concentrations are drastically decreased. This response has now been studied in rat hepatoma (H-35) cells in which prothrombin secretion is decreased 90% by incubation in the presence of warfarin. Neither prothrombin mRNA levels nor the apparent rate of prothrombin message translation were decreased when cells were cultured in the presence of warfarin rather than of vitamin K. The pool of intracellular prothrombin precursors is increased threefold by warfarin treatment, and this pool is rapidly secreted when vitamin K is administered. In contrast, continued incubation in the presence of warfarin resulted in the degradation of 60% of this pool in 24 hours. When transport of secretory proteins to the golgi apparatus was blocked with Brefeldin A, this precursor pool was gamma-carboxylated in the presence of vitamin K and no degradation occurred. Lysosomal enzyme inhibitors did not block the degradation, and the data suggest that, in rat hepatocytes, under-gamma-carboxylated prothrombin is specifically targeted to a pathway of protein degradation located in the endoplasmic reticulum.