New insights on vitamin K biology with relevance to cancer

Abstract

Phylloquinone (K1) and menaquinones (K2 family) are essential for post-translational γ-carboxylation of a small number of proteins, including clotting factors. These modified proteins have now been implicated in diverse physiological and pathological processes including cancer. Vitamin K intake has been inversely associated with cancer incidence and mortality in observational studies. newly discovered functions of vitamin K in cancer cells include activation of the steroid and xenobiotic receptor (SXR) and regulation of oxidative stress, apoptosis and autophagy. Here we provide an update on vitamin K biology, non-canonical mechanisms of vitamin K actions, potential functions of vitamin K dependent proteins in cancer and observational trials on vitamin K intake and cancer.

Expanding physiological roles for vitamin K

Although Vitamin K was discovered almost a century ago as a dietary factor essential for prevention of hemorrhaging, new functions of this vitamin in a wide array of physiological processes are increasingly recognized. There are two naturally occurring substances with vitamin K activity: phylloquinone (See Glossary) (PK, or vitamin K1) and menaquinone (MK or vitamin K2), and both forms act as co-factors for a unique post-translational protein modification called γ-carboxylation. In γ-carboxylation, glutamate (GLU) residues in target proteins are enzymatically converted to carboxyglutamate (GLA) residues. The presence of GLA residues confers calcium-binding ability to these vitamin K dependent proteins (often called “GLA proteins”). Vitamin K dependent γ-carboxylation is essential for the function of several liver derived pro- and anti-coagulant proteins that maintain the balance between efficient clotting activity and pathological thrombosis. While this role of vitamin K is well known, it is now clear that not all GLA proteins function in coagulation. Examples include γ-carboxylated proteins that regulate the balance between normal and pathological calcification and others involved in glucose metabolism and cell signaling. Intriguingly, GLA proteins and the enzymes that produce them have been identified in human tumors and tumor-derived cell lines, but the impact of γ-carboxylation on the functions of these proteins within tumors is vastly understudied. Detection of GGCX, VKORs and GLA proteins in tumors clearly suggests that vitamin K status might have relevance to cancer prevention or therapy. Observational studies have provided some evidence that dietary vitamin K intake can reduce cancer incidence or mortality, but the data are not entirely consistent. A few small randomized trials support the concept that vitamin K supplementation can retard cancer development and/or progression, but additional, larger trials are warranted. It should be noted that it is now clear that the physiological roles of K vitamins are not limited to γ-carboxylation. In particular, K2 exerts non-canonical mechanisms of action in cancer cells including disruption of electron transport and induction of apoptosis. Furthermore, it is increasingly recognized that the two forms of vitamin K differ in their absorption, tissue uptake and metabolism [1]. This review will provide an overview of vitamin K biology with an emphasis on newly discovered metabolic pathways and mechanisms of action with potential relevance to human cancer. With increased attention to the concept of “Personalized Nutrition,” comprehensive understanding of vitamin K actions during wellness and disease is of paramount importance.

Overview of the vitamin K pathway

Forms, diet, absorption and metabolism

“Vitamin K” actually refers to a family of compounds essential for γ-carboxylation of proteins, a post-translational modification critical for optimal blood coagulation. Naturally occurring substances that support γ-carboxylation include phylloquinone (PK, or vitamin K1) and various bacterially-derived menaquinones (MKs, vitamin K2 family) [1–3]. These compounds share a common core (the functional naphthoquinone ring) but differ in their aliphatic side chains (Figure 1A). PK contains a phytyl side chain of 4 prenyl units, while the K2 family contains an unsaturated aliphatic side chain with a variable number (from 4–14) of prenyl units. MK4, which has 4 prenyl units, denotes the most abundant tissue form of K2. PK is abundant in dark green leafy vegetables and plant oils and is the major dietary source of vitamin K [4]. Dihydrophylloquinone (dihydroPK) is a hydrogenated form of PK formed during processing of plant oils which is present in foods such as margarine, mayonnaise, french fries and fast foods. MKs comprise a diverse series of natural compounds produced in bacteria that accumulate in fermented foods [5]. Common dietary sources of MKs in western diets include chicken, egg and dairy products (especially cheeses). Various Asian foods such as Natto (fermented soybeans) are enriched in long chain MKs reflective of the specific bacterial species used for fermentation. While the functions of PK and MK4 have been well studied, less attention has been paid to the potential health effects of the long chain bacterially-derived MKs, in part due to their wide variability in structure [6] and limited accumulation in mammalian tissues other than the liver [7, 8]. Although evidence suggests that long chain MKs are synthesized by gut microbes [9–11], the biological significance of this with respect to vitamin K physiology is unclear since long chain MKs are readily converted into MK4 in tissues [8].

Figure 1. Forms and Function of Vitamin K.

A. Structure of vitamin K1 (Phylloquinone or PK) and vitamin K2 family (menaquinones or MK). B. The gamma-glutamyl carboxylase (GGCX) utilizes vitamin K1 or K2 hydroquinones to drive γ-carboxylation of glutamate (Glu) residues and generate γ-carboxyglutamate (Gla) residues in target proteins. The resulting vitamin K epoxides are sequentially reduced to hydroquinones by vitamin K epoxide reductases (VKOR, encoded by the VKORC1 and VKORC1L1 genes) and an as-yet-unknown vitamin K reductase (VKR). Figure created with BioRender.com.

As lipid soluble substances, K vitamins are absorbed with dietary fats, and their bioavailability is enhanced in the presence of bile salts [2]. They become incorporated into micelles and released as chylomicrons into the systemic circulation via the lymphatic system. Interestingly, regardless of dietary input (PK or various MKs), the major form detected in mammalian tissues is MK4 [8]. The relatively recent discovery of the enzyme UBIAD1 (UbiA Prenyltransferase Domain-Containing Protein 1) that converts PK to MK4 provides an explanation for MK4 accumulation in tissues [12] and studies in mice indicate that dietary PK, MK4, MK7, and MK9 all serve as precursors to tissue MK4 [8]. MK4 is highly enriched in the liver, kidney, adipose, reproductive organs, bone, and pancreas. Although the significance of MK4 accumulation in many extra-hepatic organs remains to be clarified, local pools could support γ-carboxylated protein synthesis or exert alternative biological effects as described below.

Vitamin K status, deficiency and supplements

Assessing an individual’s status is challenging as PK and MKs circulate at low concentrations and require specialized techniques for identification and quantitation [13, 14]. Clinically, serum PK is most commonly used (values < 0.15 μg/L denote deficiency), but this underestimates status as it excludes the contribution of MKs. Functionally, Prothrombin Time (PT, which measures blood clotting efficiency) and other coagulation assays are surrogate measures of vitamin K status but these lack sufficient sensitivity and solely reflect hepatic stores [13]. Direct assessment of the γ-carboxylation state of vitamin K dependent proteins such as osteocalcin more accurately reflects functionality, but these are not routinely performed. Furthermore, changes in γ-carboxylation state of GLA proteins produced in a tissue specific fashion reflect vitamin K abundance in their cells of origin and does not necessarily correlate with whole body status. In addition, some GLA proteins (notably osteocalcin) can be decarboxylated via processes unrelated to vitamin K availability [15].

The major symptom of vitamin K deficiency is impaired blood coagulation leading to excessive bruising and hemorrhaging. Overt deficiency is rare in adults as the vitamin is recycled during γ-carboxylation. However, newborns are at risk of vitamin K deficiency due to minimal tissue stores and low concentrations in breast milk [16]. To prevent the possibility of intracranial bleeding due to vitamin K deficiency, PK is routinely administered by injection to infants born in the US, although recent trends of parents opting-out has become a public health concern [17–19]. Given the emerging data to support differential tissue pools, metabolism and functions, it has been argued that establishment of distinct recommendations for PK and MKs are necessary [20]. However, insufficient data on human vitamin K requirements has precluded the National Academy of Science Food and Nutrition Board from establishing a Recommended Dietary Allowance (RDA). Instead, the board has defined an estimated adequate intake (AI) of 120 and 90 ug/day for men and women, respectively. Sub-optimal vitamin K status in older adults is reversible by PK supplementation [21–23]. In adult populations, low vitamin K status may contribute to osteoporosis, type 2 diabetes and cardiovascular disease although data from human studies is not entirely consistent [3, 21, 23–28]. A potential role of the vitamin K pathway in cancer biology is suggested by laboratory and translational studies as discussed in detail below.

Mechanism of γ-carboxylation and functions of vitamin K dependent proteins

Both PK and MK4 support γ-carboxylation which is catalyzed by gamma-glutamyl carboxylase (encoded by the GGCX gene) [29]. Continuous redox cycling of vitamin K (Figure 1B) by either of two distinct vitamin K oxidoreductase (VKOR) enzymes (VKORC1 and VKORC1L1) serves to minimize loss of the vitamin [30, 31]. The commonly prescribed anti-coagulant warfarin inhibits VKOR-mediated recycling, thus reducing γ-carboxylation of vitamin K dependent coagulation factors. Although the two VKOR enzymes are partially redundant with respect to γ-carboxylation, they differ in enzyme kinetics, sensitivity to warfarin, and tissue distribution [32–35]. Monitoring of vitamin K status is common in individuals on warfarin since genetic variations require careful titration of drug dose to prevent life threatening bleeding [36, 37].

A list of vitamin K dependent GLA proteins is provided in Table 1. Seven of these (Thrombin, Factor VII, Factor IX, Factor X, Protein Z, Protein C and Protein S) are secreted by hepatocytes and act in concert to promote initial clot formation upon injury and to terminate the coagulation cascade when appropriate. Impaired γ-carboxylation of these factors in the absence of sufficient vitamin K underlies the characteristic bleeding observed in deficiency. The presence of clustered GLA domains in coagulation factors confers calcium binding which is required for membrane association and clot formation. Some of these proteins (ie, Protein S, Protein C) regulate cellular signaling programs in addition to their role in coagulation. Other GLA proteins such as osteocalcin (BGLAP), matrix gla protein (MGP), gla rich protein (GRP) and growth arrest specific-6 (GAS6) contribute to diverse physiological processes including bone turnover, prevention of arterial calcification, glucose and body weight regulation, cell cycle/apoptosis, sperm maturation, corneal health and pancreatic cell viability [38–46]. These effects are mediated through diverse mechanisms that are not always dependent on calcium binding to the GLA domain. A family of four transmembrane proline rich GLA proteins (PRRG1,2,3,4) exhibit wide tissue expression but their functions are poorly understood [47–49]. There is conflicting data on the existence of GLA domains in periostin (POSTN) and TGF-β-induced protein (TGFBI or βig-h3) [50–52]. It should be noted that identification of γ-carboxylated proteins and mapping of GLA domains is technically difficult [53], suggesting the likelihood that unidentified GLA proteins exist. This suggestion is supported by the widespread expression of GGCX and VKORs in normal tissues [33]. Specific examples of GLA proteins that may impact cancer cell biology are discussed further below.

Table 1.

Vitamin K dependent γ-carboxylated proteins

Name	Gene ID	Functions	Comments
Factor 2 (Prothrombin)	F2	Pro-coagulant serine protease	Converts fibrinogen to fibrin, abnormal version (DCP) produced in HCC
Factor VII	F7	Initiates the extrinsic pathway of blood coagulation	Complexes with Tissue Factor
Factor IX	F9	Pro-coagulant serine protease	Cleaves Factor X
Factor X	F10	Pro-coagulant serine endopeptidase	Cleaves Thrombin
Protein C	PROC	Anti-coagulant serine protease	Degrades activated Factor V and Factor VIII, binds endothelial protein C receptor (EPCR)
Protein S	PROS1	Anti-coagulant co-factor	Complexes with activated Protein C
Protein Z	PROZ	Anti-coagulant co-factor	Inhibits activated Factor X
Osteocalcin	BGLAP	Extracellular matrix protein involved in bone remodeling & energy metabolism	Secreted by osteoblasts, binds hydroxyapatite
Matrix Gla Protein	MGP	Inhibitor of ectopic cardiovascular calcification	Secreted by chondrocytes and vascular smooth muscle cells
Gla Rich Protein	GRP	Inhibitor of cardiovascular calcification	May have anti-inflammatory actions
Growth Arrest Specific Gene 6	GAS6	Ligand for TYRO/AXL/MER (TAM) receptors	Elevated in many diseases including sepsis and cancer
Proline Rich And Gla Domain 1	PRRG1	Intracellular signaling?	Transmembrane protein
Proline Rich And Gla Domain 2	PRRG2	Intracellular signaling?	Transmembrane protein
Proline Rich And Gla Domain 3	PRRG3	Intracellular signaling?	Transmembrane protein
Proline Rich And Gla Domain 4	PRRG4	Intracellular signaling, axon guidance?	Transmembrane protein
Gamma-Glutamyl Carboxylase	GGCX	Vitamin K dependent γ-carboxylase	γ-carboxylation may affect GGCX stability
Transthyretin	TTR	Serum thyroid hormone binding protein	Transport T4 and retinol
Inter-alpha trypsin inhibitor	ITIH2	Plasma serine protease inhibitor	Implicated in hyaluronan transport and action
Periostin	POSTN	Extracellular matrix protein involved in development, regeneration, wound healing	Conflicting data on GLA modification
TGFβ Induced Protein	TGFBI	Extracellular matrix collagen binding protein that inhibits cell adhesion	Conflicting data on GLA modification

Role of K vitamins in cancer

Unique link between vitamin K and hepatocellular carcinoma

Given the essential role of liver in γ-carboxylation, early studies examined the impact of hepatocellular carcinoma (HCC) on production of active coagulation factors. As liver function declines in HCC, an uncarboxylated, immature form of thrombin (des-γ-carboxy prothrombin or DCP) was found to accumulate in blood [54]. DCP has shown efficacy as a serum biomarker of early-stage HCC [55] and may drive disease progression through both autocrine and paracrine signaling [56]. The impact of vitamin K on DCP and HCC progression has been addressed in several clinical settings (Table 2). Retrospective analyses of HCC patients suggested that MK4 supplementation (which is typically prescribed for prevention of osteoporosis) reduced circulating DCP and extended survival time [57]. In another study designed to assess the long-term effects of MK4 supplementation on bone loss in women with viral liver cirrhosis, development of HCC was evaluated as a secondary outcome. Over 8 years of follow-up, the risk ratio for the development of HCC in the MK4 group compared with the control group was 0.20 [58]. More recently, a phase 2 randomized placebo-controlled trial in HCC patients demonstrated that MK4 supplementation (45 mg/day orally) enhanced the efficacy of the multi-kinase inhibitor sorafenib [59]. MK4 supplementation increased the objective response rate to sorafenib and extended median progression-free survival while decreasing circulating DCP [58]. Collectively, these data support the concept that deregulated γ-carboxylation contributes to development and progression of HCC and that long term vitamin K supplementation is safe and may offer survival benefit in HCC patients.

Table 2.

Clinical trials of vitamin K and cancer

STUDY DESIGN AND OUTCOME MEASURES	SUBJECTS & TREATMENTS	RESULTS	References
Randomized trial of MK4 treatment Primary Outcome: Cumulative proportion of patients with HCC	43 F (mean age 60y) with HCV+ liver cirrhosis Treatment group received 45 mg/d MK4 (no placebo group) Monitored for up to 8y	MK4 supplementation reduced HCC development (log-rank test, p< 0.02) Cox regression analysis indicated reduced risk ratio with MK4 supplementation (univariate 0.20; multivariate 0.13) No adverse effects of MK4 supplementation	[58]
Retrospective analysis of clinical data Outcomes: Progression Free Survival and Overall Survival	HCC patients treated with Sorafenib (M/F, mean age 73y) 36 patients treated with sorafenib alone; 29 patients who had simultaneously received MK4 for bone loss (dose not stated) Monitored for up to 5y	Progression-free survival was prolonged in MK4 group compared with sorafenib alone group (log rank test, p < 0.001) Multivariate analysis with Cox proportional-hazards model indicated MK4 supplementation reduced Hazard Ratio (HR) for both progression free survival (0.16) and overall survival (0.38) MK4 reduced incidence of grade 3/4 adverse events (p<0.05)	[57]
Phase 2 Randomized Trialiii Primary Outcome: Progression Free Survival Secondary Outcomes: Overall Survival, Safety	44 HCC Patients treated with sorafenib (M/F, mean age 72y) 22 assigned to placebo, 22 assigned to MK4 (15mg/3x day/oral) Monitored for 6+y	MK4 increased Objective Response Rate to sorafenib from 4.5% to 27.3% MK4 improved Progression Free Survival (HR = 0.44) and Overall Survival (HR = 0.59) MK4 reduced circulating DCP (p < 0.002) No differences in incidence of adverse events	[59]
ECKO Triali Randomized, double blind placebo-controlled trial Primary Outcomes: Bone density, fractures Secondary Outcome: Development of any cancer	440 F (mean age 59y) Post-menopausal Daily oral supplement of 5mg PK or placebo for 4 years	PK supplement reduced total cancer incidence (Kaplan Meier curves; log rank test HR = 0.25) PK supplement increased serum PK 10X and decreased undercarboxylated osteocalcin Higher mean serum PK correlated with lower cancer incidence (p < 0.05)	[26]

Vitamin K Supplementation for non-hepatic cancers

No intervention studies have been specifically designed to assess the impact of vitamin K supplementation on cancers other than HCC. However, the ECKO trial ⁱ (Evaluation of the Clinical use of vitamin K supplementation in postmenopausal women with Osteopenia) assessed cancer incidence as a secondary endpoint in women randomized to placebo or PK supplementation (5mg/day) for 4 years (Table 2) [26]. PK supplementation increased serum PK and decreased circulating undercarboxylated osteocalcin (indicating enhanced γ-carboxylation activity in bone). Higher mean serum PK levels over the duration of the study correlated with lower total cancer incidence (the majority were breast cancers). This data suggested that sub-optimal vitamin K status is common in post-menopausal women and that supplementation enhanced serum PK pools, improved osteocalcin γ-carboxylation and reduced cancer risk in this population.

Population data on vitamin K intake and non-hepatic cancers

The relationship between dietary PK and various MKs as assessed by food frequency questionnaires (FFQ) and incidence or mortality of non-hepatic cancers has been explored in several large observational studies (Table 3). In the PREDIMED ⁱⁱ (Prevención con Dieta Mediterránea) study of adults at high risk for cardiovascular disease, baseline PK (but not MK) intake was inversely associated with risk of cancer-related deaths [60]. In the Danish Diet, Cancer, and Health study, moderate to high (87–192 μg/d) PK intake at enrolment was associated with lower risk of cancer-related mortality, but only in current/former smokers. Increased intake of either PK or MKs during follow-up was also associated with reduced risk of cancer mortality [61]. In the Heidelberg cohort of the European Prospective Investigation of Cancer (EPIC) trial, intake of MKs (but not PK) was associated with reduced cancer mortality. In contrast, no association between intake of either PK or MKs and cancer mortality was found in the Netherlands EPIC cohort, possibly because >90% of participants had intakes above the recommended AI set by the European Food Safety Authority [62].

Table 3.

Observational studies of dietary vitamin K and cancer

Study	Details	Approach	Findings	References
EPIC – Heidelberg Prospective cohort study Prostate Cancer Incidence	11,319 M 40–65y Mean follow-up 8.6y	145 item FFQ at baseline Quartiles of energy adjusted PK, MK4, MK5–9, MK9–14 calculated from food tables	MK (but not PK) intake inversely associated with total and advanced prostate cancer Strongest association with MK5–9 from fermented dairy products	[63]
EPIC – Heidelberg Nested case-control study Prostate Cancer Incidence	250 cases 494 controls	Serum total and undercarboxylated osteocalcin	Undercarboxylation of serum osteocalcin in high grade & advanced stage prostate cancer	[64]
EPIC – Heidelberg Prospective cohort study Total Invasive Cancer Incidence and Mortality	24,340 M/F 35–64y Mean follow-up > 10y	148 item FFQ at baseline Quartiles of energy adjusted PK, MK4, MK5–9, MK10–14 calculated from food tables	MK4 and MK5–9 (but not PK) intake inversely associated with cancer mortality in M only (especially prostate and lung cancers)	[144]
PREDIMED Trialii Prospective cohort study Cancer Mortality	7,216 M/F 55–80y Population with high cardiovascular risk consuming Mediterranean diets Median follow-up 4.8y	137 item FFQ at baseline + annually Quartiles of energy adjusted PK and MK calculated from food tables	PK intake inversely associated with cancer mortality Increased PK or MK intake during follow-up associated with reduced cancer mortality	[60]
EPIC – Netherlands Prospective cohort study Cancer Mortality	33,289 M/F 20–70y Mean follow-up 16.8y	178 item FFQ at baseline Energy adjusted PK and MK (total-, short chain and long chain) intakes as quartiles and as continuous variables	No association between intakes of PK or MKs and cancer mortality	[62]
Danish Diet, Cancer and Health Study Prospective cohort study Cancer Mortality	56,048 M/F 52–60y Mean follow-up 21y	192 item FFQ at baseline Energy adjusted quintiles of PK intake	Non-linear inverse association between PK intake and cancer mortality Restricted to former/current smokers	[61]
PLCO Trialiv Prospective cohort study Pancreatic Cancer Incidence and Mortality	10,165 M/F Mean follow-up 8.8y	137 item FFQ at baseline Energy adjusted quintiles of PK, dihydroPK and total MK (4–14) intake	Non-linear inverse association between PK or dihydroPK intake and pancreatic cancer incidence No association of MK intake with pancreatic cancer incidence	[65]
PLCO Trialiv Prospective cohort study Breast Cancer Incidence and Mortality	51,662 US F Mean follow-up 13.6y	124 item FFQ within 3 years of recruitment Quintiles of intake: Total K = sum of PK + MKs; Total MK = sum of MK4-MK14	Total MK (but not PK) intake associated with increased breast cancer incidence and mortality (non-linear inverted Ushaped dose response)	[66]

Three studies have reported on vitamin K intake in relation to site-specific cancer incidence. In the Heidelberg EPIC male cohort, an inverse association between incidence of advanced prostate cancer and MK intake (particularly MK5–9 from dairy products) but not PK intake was observed [63]. In a nested case-control follow-up study, undercarboxylation of osteocalcin (a marker of inadequate vitamin K status) was significantly higher in cases with advanced or high-grade prostate cancer vs controls [64]. In the PLCO (Prostate, Lung, Colorectal, and Ovarian Cancer Screening) trial, intake of PK but not MK was inversely associated with the risk of pancreatic cancer [65]. Contrary to these studies, Wang et al [66] reported a positive association between intake of MKs (but not PK) and risk of breast cancer incidence and death, especially for luminal-like, triple negative, and early-stage disease.

In summary, while most observational trials support the concept that high vitamin K intake reduces risk of cancer development or mortality, the data are inconsistent with respect to the specific forms that might be beneficial for various types of cancer. Interpretation of these data are complicated by genetic variability in vitamin K pathway genes, altered vitamin K metabolism or function in different tumor types, the use of different strategies to evaluate vitamin K intake and content of foods, the presence of confounding dietary factors, and limited assessment of vitamin K status. Given the promising results from the small ECKO intervention trial, additional effort should be devoted to assessing the impact of vitamin K supplementation on cancer end points.

Expression/function of GGCX and VKORs in human cancers/cells

Laboratory studies to evaluate the vitamin K pathway in carcinogenesis have used cellular models and human tumor databases. GGCX, VKORC1 and VKORC1L1 were up-regulated in cellular models of carcinogenesis, in cell lines derived from aggressive human breast cancers and in tissue arrays from human breast tumors [67]. Analysis of The Cancer Genome Atlas (TCGA) datasets revealed that amplification and up-regulation of vitamin K pathway genes (GGCX, VKORC1 and/or VKORC1L1) or low expression of the MK4 biosynthesis enzyme UBIAD1 in human breast tumors correlated with reduced overall survival. Bioinformatic approaches have linked GGCX to bladder cancer susceptibility and lung cancer metastasis [68, 69]. Collectively, these observations suggest that aberrant activity of the vitamin K cycle and/or deregulation of UBIAD1-mediated MK4 synthesis might impact cancer progression.

Use of vitamin K antagonists and cancer

Given that human tumors express GGCX and VKORs, one mechanism by which vitamin K could impact cancer is through altered γ-carboxylation. However, data on the impact of the vitamin K antagonist warfarin (which blocks γ-carboxylation) on cancer risk, progression and survival is inconsistent. Some studies suggested reduced risk and/or better disease outcomes for prostate, lung, and breast cancer in warfarin users [70–75]. Confounding factors in these studies include individual variations in the therapeutic efficacy of warfarin and the limited ability of warfarin to inhibit VKORC1L1 which is expressed in many normal tissues and in cancers [32, 33, 67]. These studies are also complicated by the high frequency of thromboembolism and other clotting disorders in cancer patients which affect disease progression. Some analyses [71, 76] have reported anti-cancer effects of functionally distinct anti-coagulants such as heparin (which would argue against an effect mediated via vitamin K inhibition) whereas others demonstrated the opposite relationship [73]. Overall, the available data does not clarify the relationship between use of vitamin K antagonists and cancer incidence, progression, or survival.

Expression and relevance of γ-carboxylated proteins in cancer

As discussed below, several γ-carboxylated proteins are expressed in tumors, but the impact of carboxylation state on the functions of these proteins in the context of cancer biology has yet to be fully characterized. Cell culture media typically lack vitamin K and therefore γ-carboxylation is not optimized in cell culture studies unless PK or MKs are intentionally added. Indeed, exposure of breast cancer cells to physiological concentrations of PK induced γ-carboxylation of multiple proteins which was blocked by warfarin [67]. Continuous culture of breast cancer cells in PK promoted a more aggressive phenotype characterized by emergence of stem cell features [67]. This data suggests that vitamin K might drive γ-carboxylation of proteins that affect cancer cell behavior. Select GGCX targets in cancer (Table 4) are discussed below with the caveat that little is known about the impact of γ-carboxylation on their functions in the context of carcinogenesis. Further studies to identify specific GLA proteins produced by cancer cells in the presence of vitamin K are needed to fully understand their role in tumor biology.

Table 4.

Expression and actions of select vitamin K dependent proteins in cancer.

GAS6 and Protein S (PROS1)
Pancreatic cancer	γ-carboxylated GAS6 promoted aggressive pancreatic cancer invasion, survival, drug resistance and metastasis via AXL. Warfarin inhibited GAS6-mediated AXL activation and blocked the progression and spread of pancreatic cancer in vivo. PROS1 exerted pro-apoptotic actions via TAM receptors in pancreatic ductal carcinoma cells; PROS1 competed with GAS6 for MERTK signaling.	[82, 86]
Breast Cancer	γ-carboxylated GAS6 activates pro-invasive activity of AXL via PEAK1 driven disassembly of focal adhesions	[83]
Prostate cancer	PROS1 increased androgen independence in vitro and in vivo. Proteomic screening identified PROS1 enrichment in castrate resistant prostate cancers – evidence for marker of aggressive disease.	[84, 90]
Squamous carcinoma	PROS1 stimulated proliferation and migration in vitro via regulation of AXL; PROS1 knockdown reduced tumor growth in vivo.	[89]
Glioblastoma	↑ PROS1 level associated with cell proliferation, migration, invasion. Silencing PROS1 inhibited apoptosis, proliferation, migration. Glioma cell secretion of PROS1 attracted neural stem/progenitor cells via TYRO signaling.	[85, 145]
Lung cancer	PROS1 activated TYRO and MERTK to promote proliferation, migration, anchorage independence; PROS1 stimulated endothelial cell tube formation.	[87]
Matrix Gla Protein (MGP)
Colon cancer	High MGP expression correlated with increased proliferation, invasion, migration and tumor growth and impaired apoptosis. MGP increased proliferation via NFκB signaling.	[97, 98, 103, 108]
Breast cancer	High tumor MGP associated with poor prognosis. MGP promoted epithelial mesenchymal transition, proliferation, invasion, and migration in aggressive breast cancer cells, but had opposite effects in early-stage breast cancer cells.	[96, 107, 146]
Ovarian cancer	↑ MGP expression in micro-dissected ovarian cancers. MGP markedly up regulated in drug resistant cells.	[100, 101]
Gastric cancer	High tumor MGP associated with poor prognosis; high intracellular MGP promoted cell proliferation via IL2-STAT5 signaling.	[104]
Glioma	↑ MGP in aggressive glioma cells; manipulation of MGP altered migration and invasion in vitro and in vivo.	[93]
Osteosarcoma	Ectopic expression of MGP increased lung metastastic ability independent of γ-carboxylation; high serum MGP detected in osteosarcoma patients with lung metastases.	[105]
Protein C (PROC) and Activated Protein C (aPC)
Breast cancer	aPC mediated cytoprotective actions via EPCR and PAR signaling. EPCR was robustly expressed in triple negative breast cancers. EPCR identified as stem cell marker in triple negative breast cancer; silencing of EPCR reduced cancer stem cell numbers. EPCR associated with metastases and adverse clinical outcomes in invasive ductal carcinoma.	[116, 147, 148]
Melanoma	Treatment with activated PROC decreased adhesion and migration in vitro and reduced liver and lung metastases in mouse models.	[112]
Lung cancer	High EPCR levels correlated with poor prognosis in early-stage lung adenocarcinoma; silencing EPCR or blocking aPC/EPCR interaction impaired pro-metastatic activity.	[113]
Colon cancer	High EPCR expression secondary to gene amplification and DNA hypomethylation; aPC triggered ERK signaling via EPCR.	[120]

Growth arrest-specific 6 and Protein S

Growth arrest-specific 6 (GAS6) and Protein S (PROS1) are structurally similar GLA proteins that bind and activate a family of tyrosine kinases collectively known as TAM receptors (TYRO3, AXL, and MERTK) [77]. GAS6 was overexpressed in several types of cancer and activated oncogenic signaling pathways downstream of TAM receptors including mitogen-activated protein kinase (MAPK), NFκB and PI3K/AKT [78–81]. Mechanistic studies [82] demonstrated that γ-carboxylation of GAS6 was necessary for AXL to promote aggressive pancreatic cancer invasion, survival, drug resistance and metastasis and that warfarin treatment inhibited GAS6-mediated AXL activation and blocked the progression and spread of pancreatic cancer in vivo. γ-carboxylation of GAS6 has also been linked to the pro-invasive actions of AXL in aggressive breast cancer [83]. Of particular interest, inhibition of GAS6-mediated AXL activation in vivo was achieved at doses of warfarin that did not affect coagulation factors [82], suggesting the possibility that γ-carboxylation of GAS6 in tumors might be uniquely sensitive to warfarin.

High expression of PROS1 was found in aggressive prostate cancers, glioblastomas and pancreatic cancer cell lines, and PROS1 stimulated proliferation and migration in vitro while its knockdown reduced tumor growth in vivo [84–86]. MERTK and TYRO3 mediated effects of PROS1 on proliferation, migration, anchorage-independent growth and angiogenesis in lung cancer cells [87–90]. Although PROS1 γ-carboxylation is necessary for activation of TAM receptors in general [91], no studies have specifically compared the effects of carboxylated and uncarboxylated PROS1 on cancer cells. The relative concentrations of TAM ligands may be critical in mediating pro-tumorigenic effects via MERTK, as either overexpression of PROS1 or knock down of GAS6 inhibited proliferation and promoted apoptosis in cancer cells comparably to that achieved with a MERTK-specific inhibitor [86].

Matrix Gla Protein

Although best characterized as a calcification inhibitor, MGP, which contains five GLA residues [92], has recently been implicated in cancer. MGP was overexpressed in glioblastoma relative to normal brain [93, 94] and was also enriched in colon, breast and renal cancers [95–97]. Several oncogenic transcription factors (ID4, HOXC8) have been identified as drivers of high MGP expression in cancer [94, 96]. MGP drives proliferation, invasion, angiogenesis and metastasis of various cancers [93, 96, 98, 99] and may contribute to paclitaxel and topotecan resistance in ovarian cancer [100, 101]. In mechanistic studies, expression of MGP in glioblastoma cells increased migration in vitro and tumor spread in vivo, whereas MGP knockdown had the opposite effect [102]. Mechanisms by which MGP promoted the cancer phenotype included activation of JAK2/STAT5, TGFβ, NFκB and intracellular calcium signaling [103–105]. In contrast to these studies suggesting MGP driving cancer progression, MGP expression was found to be reduced in colorectal adenocarcinomas relative to adjacent normal tissue [106]. Furthermore, MGP repression promoted proliferation, invasion, migration and increased sensitivity to chemotherapeutics in cancer cells [98, 107, 108]. These disparate results may reflect differences in the γ-carboxylation state of MGP, although the pro-tumorigenic effects of MGP in osteosarcoma cells were not blocked by warfarin and were retained in cells expressing MGP mutants that lack γ-carboxylation sites [105].

Protein C

Protein C (PROC) is a GLA protein that acts as a natural anticoagulant through a processed form known as activated Protein C (aPC). aPC binds to the endothelial protein C receptor (EPCR), a key regulator of blood coagulation that is also expressed on many cancer cells where it regulates migration, invasion, angiogenesis and apoptosis [109–114]. In human and murine xenograft breast cancer models, EPCR silencing reduced the growth of primary and metastatic tumors [115]. Furthermore, high EPCR expression correlated with reduced patient survival and poor outcome, including distant metastases, in breast cancer cohorts [115]. Consistent with this data, EPCR has been identified as a marker of the population of breast cancer stem cells that are enriched in aggressive subtypes of the disease and exhibit tumor initiating properties [116, 117]. These data are consistent with findings that EPCR overexpression was associated with increased tumor aggressiveness in an animal model of human triple negative breast cancer cells [110]. Poor prognosis in early-stage lung cancer patients was also associated with high level of EPCR [113]. In contrast, PROC mRNA was not consistently expressed whereas EPCR and PROS were reduced in biopsies of malignant vs benign ovarian cancer [111].

Mechanistically, activation of EPCR by aPC increased invasion and chemotaxis of breast cancer cells [118, 119]. In lung cancer models, the aPC/EPCR complex promoted cell survival by triggering the AKT/ERK signaling pathways, and metastasis in vivo was reduced when the aPC/EPCR interaction was blocked. Similar results were reported in models of colorectal and gastric cancers [114, 120]. Conflicting date however was reported in a transgenic mouse model of melanoma, where overexpression of EPCR/aPC reduced liver and lung metastases [112]. Thus, the role of PROC in cancer biology is complex and appears to vary depending on its processing, carboxylation state, the specific tumor context and the interactions between aPC and EPCR.

Cellular and molecular effects of K vitamins in cancer model systems

In addition to modulation of γ-carboxylation, K vitamins might directly affect the cancer cell phenotype via alternative, non-canonical mechanisms. Both in vitro and in vivo approaches have addressed the impact of PK and various MK compounds on cancer cells and tumor growth. The in vitro studies have utilized different dosing regimens (5μM – 200μM) and different treatment times (<24h – 96h). Depending on the experimental design, the effects of K vitamins on cancer cells have included cell cycle arrest, apoptosis and/or autophagy. In studies where both forms of the vitamin have been compared, MKs generally exerted more potent anticancer effects than PK. In many studies where MK4 in the range of 10–20μM reduced cancer cell viability in vitro, PK had no effect except at non-physiologic concentrations (over 150μM). Neither PK nor MK4 have been shown to affect the growth of normal cells in culture.

Induction of apoptosis and autophagy by vitamin K compounds.

Anticancer effects of vitamin K in vitro have been linked to apoptosis secondary to oxidative stress, MAPK signaling and mitochondrial dysfunction (Figure 2). In pancreatic cancer cells, both PK and MK4 activated ERK, up-regulated pro-apoptotic BAX and induced caspase-dependent apoptosis which was blocked by MAPK inhibitor [121]. Similarly, PK induced mitochondrial mediated apoptosis in colon cancer cells by sustained activation of ERK secondary to up-regulation of BAX and downregulation of anti-apoptotic BCL2 [122]. MK4 mediated apoptosis may also involve binding of MK4 to pro-apoptotic BAK, direct effects on mitochondrial membrane depolarization and reactive oxygen species (ROS) [123, 124], as either ROS neutralization by antioxidants (N-acetyl cysteine (NAC) and alpha-tocopherol) or BAK knockdown prevented MK4 mediated mitochondrial disruption and apoptosis [125].

Figure 2. Mechanisms of induction of apoptosis and autophagy by MK4.

MK4 induces apoptosis through production of ROS which activates ERK, JNK/p38 MAPK, and initiates binding of MK4 to BAK. Activation of ERK results in up-regulation of BAX, decrease in mitochondrial membrane potential (MMP), and cytochrome c release. JNK/p38 MAPK activation results in down-regulation of BCL2. MK4 covalently bound to BAK induces decrease in MMP and cytochrome c release. The outcome of mitochondrial dysfunction and cytochrome c release is caspase activation which results in apoptosis. ROS production can be blocked by N-acetyl-cysteine (NAC) and alpha-tocopherol which can ultimately block MK4 mediated apoptosis. MK4 also induces metabolic stress by reducing ATP production and increasing lactate production which activates AMPK and in turn blocks MTORC1 (responsible for protein synthesis and proliferation) and initiates autophagy and formation of autophagosomes with up-regulation of LC3B-II. 3-Methyladenine (3MA) can block autophagosome formation. Arrows denote activation; blunt ended arrows denote inhibition. Figure created with BioRender.com.

Also shown in Figure 2 is the induction of autophagy by MK4. Autophagy is an evolutionary conserved mechanism for degradation of unnecessary/dysfunctional molecules via autophagosomes which in some contexts triggers cell death [126, 127]. In colon and breast cancer cells, MK4 induced autophagic cell death as evidenced by up-regulation of the autophagosome marker LC3B-II [128, 129]. In Jurkat leukemic cells (but not in normal lymphoblasts), MK4 increased phosphoethanolamine, a precursor to phosphatidylethanolamine which is necessary for LC3B-II generation [130]. Both 3-methyladenine (autophagy inhibitor) and the antioxidant NAC blocked MK4 mediated autophagic cell death indicating ROS involvement. In mitochondria, MK4 also reduced basal respiration and ATP production while increasing lactate production [67, 131]. Prolonged metabolic stress induced by MK4 in bladder cancer cells triggered autophagic cell death through AMPK activation, mTORC1 inhibition and ULK1 (unk-51-like autophagy-activating kinase 1) activation leading to autophagosome formation [131]. Depending on cellular context, apoptosis and autophagy may be simultaneously triggered by MK4.

Effects of K vitamins on cancer phenotypes

At concentrations that do not induce cell death, MK4 inhibited cancer cell migration, spheroid formation and expression of stem cell markers. In contrast, PK (at equivalent near-physiologic concentrations) increased proliferation and stem cell features whereas MK4 reduced cell viability, migration and spheroid formation in breast cancer models [67]. MK4 similarly inhibited anchorage independent growth and spheroid formation in castration resistant prostate cancer cells [132]. MK4 blocked invasion of HCC cells via inhibition of RhoA, a small GTPase that regulates actin cytoskeleton and motility [133]. These studies suggest that MK4 may be more effective than PK in reducing tumor progression, at least in some cancer types.

There have been a limited number of in vivo studies addressing the impact of K vitamins on tumor growth, and none have utilized dietary approaches. In mouse xenograft studies, treatment with MK4 administered in water at a calculated dose of 20 mg/kg/d significantly reduced growth of established HCCs as well as both androgen-dependent and androgen-independent prostate tumors [133, 134]. These data suggest that effective concentrations of MK4 can be achieved with oral delivery. In another study, MK4 injected directly into bladder cancer xenografts slowed tumor growth and increased long-term survival rates [131]. Proteins involved in autophagy and apoptosis were increased in MK4 treated tumors, supporting the in vitro studies implicating these processes as mediators of the anti-cancer effects of MK4.

Activation of SXR by MK4 and interaction with UBIAD1

The nuclear steroid and xenobiotic receptor (SXR, also known as pregnane X receptor or PXR) has been identified as an alternative mediator of MK4 actions in some cells [135]. Examples include regulation of SXR target genes involved in bile acid synthesis in HCC [136] and those involved in cholesterol efflux and steroid catabolism in prostate cancer [137]. A complex network linking MK4, SXR and UBIAD1 (the MK4 biosynthetic enzyme), has emerged (Figure 3). UBIAD1, which has been identified as a tumor suppressor, interacts with enzymes involved in cholesterol synthesis and storage. The loss of UBIAD1 in prostate cancer cells reduced MK4 synthesis which in turn decreased SXR transcriptional regulation of cholesterol efflux and transport resulting in prostate cancers with an aggressive phenotype and elevated intracellular cholesterol. Moreover, ectopic expression of UBIAD1 in renal cancer cells with low or absent UBIAD1 not only altered lipid metabolism through activation of SXR, but also significantly decreased proliferation and colony formation while increasing caspase activation [138]. The tumor suppressor activities of UBIAD1 are thus consistent with the anti-cancer effects of MK4 described above.

Figure 3. MK4 biosynthesis by UBIAD1 and activation of SXR.

UBIAD1 is the biosynthetic enzyme responsible for MK4 production from K vitamins (PK and MKs) and formation of cellular pools of MK4. The side chain of K vitamins is cleaved by an unknown enzyme to form menadione (Men) which is prenylated by UBIAD1 utilizing the isoprenoid geranylgeranyl pyrophosphate (GGpp) to produce MK4. UBIAD1 translocates from the ER to the Golgi when sterol synthesis is high, thus regulating sterol synthesis and maintaining isoprenoid synthesis. MK4 can bind and activate the transcription factor SXR and up-regulate expression of many genes including those involved in cholesterol catabolism and efflux, thus reducing cancer phenotype. Figure created with BioRender.com.

Therapeutic targeting of vitamin K pathway

Drug resistance, which is commonly associated with treatment failures in cancer, may be lessened by combining therapies with different mechanisms of action. As noted in Table 2, supplementation of HCC patients with MK4 enhanced the activity of the kinase inhibitor sorafenib. The mechanism of synergy between vitamin K and sorafenib was studied in HCC cells in vitro and in vivo. At doses that were ineffective alone, co-treatment with sorafenib and PK induced apoptosis through caspase activation, DNA fragmentation and inhibition of the Raf/MEK/ERK pathway. Marked regression of transplantable HCC was achieved in vivo after intraperitoneal injections of PK (2mg/kg daily) and sorafenib [139]. Other promising vitamin K related synthetic analogs have been developed as potential therapeutics [140]. For example, PPM18 and ester derivatives of reduced MK4 (MKH4) display enhanced cellular uptake and more potent anticancer activity via mitochondrial-mediated apoptosis than natural MK4 against HCC, bladder cancer and leukemic cells in vitro [141–143]. Further studies to clarify the mechanism of action for natural and synthetic vitamin K compounds and their potential relevance to cancer prevention and therapy are warranted.

Concluding remarks

Recent research on vitamin K has expanded our view of its physiological importance into new directions including potential role in age-related diseases including osteoporosis, type 2 diabetes and cancer (see Clinician’s Corner). Although overt deficiency is rare, sub-optimal vitamin K status appears to be common (especially in aging populations) and may be associated with an increased risk of chronic disease. A limited number of small trials suggest that supplementation with either PK and MK4 may reduce development or progression of cancer. In population studies, high dietary vitamin K has been linked to reductions in cancer mortality and incidence but there is considerable uncertainty regarding the specific forms that may be beneficial. Limitations of published trials on vitamin K and cancer include difficulty in assessing long-term intakes, incompleteness of food composition databases for the various PK and MK forms and the likelihood of individual variation in vitamin K metabolism. While in vitro studies support anti-cancer effects of PK or MK4, many of these have utilized supra-physiological concentrations that are not likely achievable in vivo. Pre-clinical studies to assess the impact of dietary vitamin K on animal models of cancer should be prioritized to clarify the specific forms and metabolites that accumulate in tumors and alter disease progression in vivo. Randomized controlled trials to assess the impact of dietary and/or supplemental vitamin K on cancer end points are also warranted.

CLINICIAN’S CORNER.

• Vitamin K is a required cofactor for hepatic γ-carboxylation of coagulation proteins; overt deficiency is associated with hemorrhaging. Breast fed neonates are susceptible to vitamin K deficiency due to lack of stores at birth and limited transfer of vitamin K to milk. Vitamin K deficiency may be precipitated by anti-coagulant therapies such as warfarin and attention to dietary vitamin K is required for such patients.

• Phylloquinone (K1) is the major dietary form, but it is converted into menaquinone (K2) in tissues. A variety of menaquinones with vitamin K activity are produced by bacteria. These are present in fermented foods and are also synthesized by some intestinal-resident bacteria.

• Additional physiological functions of vitamin K dependent proteins include optimization of skeletal remodeling, inhibition of soft tissue calcification, regulation of energy metabolism and others. Sub-optimal vitamin K status is common in adults and may contribute to chronic diseases such as osteoporosis, type 2 diabetes and cardiovascular disease. Vitamin K supplementation is safe and well tolerated.

• Enzymes of the vitamin K cycle and vitamin K dependent proteins have been identified in human tumors. HCC is characterized by production of an abnormal coagulation factor and HCC patients may benefit from vitamin K supplementation. Some supplementation trials and dietary intake studies support a protective effect of vitamin K against cancer development and mortality, but additional randomized intervention studies are warranted to confirm these associations.

• Some cancers up-regulate vitamin K pathway activity and aberrantly produce vitamin K dependent proteins, but the relevance of these changes in relation to disease progression require further study. Basic research studies have demonstrated anti-cancer actions of vitamin K that are mediated via newly discovered mechanisms unrelated to γ-carboxylation.

Additional basic research is needed to identify changes in vitamin K actions that occur during tumor progression. Genomic databases indicate that some advanced cancers (ie, breast and pancreatic) exhibit up-regulation of vitamin K cycle enzymes, and laboratory studies have demonstrated that several known γ-carboxylated proteins exert oncogenic actions. It is possible that in these cancers, vitamin K supplementation may be contra-indicated. Basic research to define the impact of γ-carboxylation on the function of vitamin K dependent proteins in the tumor microenvironment is essential.

With respect to mechanistic insight, support for non-canonical mechanisms of action of K vitamins in tumor suppression has accumulated. The K2 family is synthesized by bacteria, including those present in the gut microbiome, and the long chain MKs may be especially potent in promoting anti-cancer processes such as apoptosis and autophagy. As noted in the Outstanding Questions, more complete understanding of the metabolism and cellular mechanisms of action for specific vitamin K compounds is ultimately needed to enable translation of these findings into strategies for cancer prevention or treatment.

Outstanding Questions.

• What is the importance of vitamin K in extra-hepatic tissues? Do undiscovered vitamin K dependent proteins exist that contribute to normal physiology? Are there unrecognized developmental windows in the life course (infancy, pregnancy, lactation, aging) where vitamin K status and functions are critically important?

• What is the optimal vitamin K status for cancer prevention, and can it be achieved by diet or supplements? Is there a threshold to the relationship between dietary vitamin K and cancer incidence or mortality? What are the recommended doses and form of vitamin K in the prevention of cancers? Are there concerns regarding vitamin K toxicity or adverse reactions?

• What changes in the vitamin K cycle occur during cancer progression? What vitamin K dependent proteins are synthesized by tumor cells and how does γ-carboxylation affect their function? Does γ-carboxylation of cancer-specific GLA proteins contribute to human cancer development or progression? What is the potential role of vitamin K in regulation of cancer stem cells?

• What forms or metabolite(s) of vitamin K accumulate in human tumors, and which are most potent in mediating anti-cancer effects? Are non-canonical mechanisms involved?

• Is UBIAD1 a tumor suppressor in human tumors via biosynthesis of MK4? If so, does deregulation of its activity in tumors contribute to disease progression?

• What are the comparative effects of dietary PK and MK in animal models of cancer?

• Do cancers respond differently to vitamin K based on sub-type or stage of disease?

• What is the bioavailability of different forms of vitamin K? Which form of vitamin K is most bioavailable? Are the long chain menaquinones synthesized by the gut microbiome absorbed sufficiently to impact tissue pools? Do long chain menaquinones possess unique properties that could be translated into therapeutic strategies for cancer or other diseases? If so, are novel mechanisms involved?

Highlights.

• Phylloquinone (PK) and menaquinones (MKs) are natural compounds essential for the vitamin K cycle of γ-carboxylation, a rare protein post-translational modification.

• Seventeen γ-carboxylated proteins have been identified that regulate blood coagulation, soft tissue calcification, bone health and energy metabolism, but the functional impact of γ-carboxylation and the widespread role of these proteins in health and disease have yet to be fully characterized.

• Overt vitamin K deficiency sufficient to impair blood clotting is rare, but sub-clinical status is common especially in elderly populations.

• Observational studies suggest that low vitamin K intake increases cancer risk, but there is conflicting data on the specific forms of vitamin K that might be beneficial and the types of tumors that might be impacted by vitamin K status.

• Aggressive tumors exhibit aberrant expression of genes required for vitamin K metabolism and γ-carboxylation. γ-carboxylated proteins have been identified in tumors and cancer cell lines.

GLOSSARY

AI

adequate intake

AMP-activated protein kinase (AMPK)

key enzymatic regulator of cell energy homeostasis

Apoptosis

process of programmed cell death that is triggered by multiple stimuli

Autophagy

process by which damaged and redundant components are degraded in autophagolysosomes

Cancer stem cells (CSCs)

a subpopulation of cancer cells with abilities of self-renewal and differentiation, associated with tumor initiation, progression and drug resistance

DCP

des-γ-carboxy prothrombin, an immature (uncarboxylated) form of thrombin produced by hepatocellular carcinoma cells

Gamma-glutamylcarboxylase (GGCX)

endoplasmic reticulum localized enzyme responsible for vitamin K-dependent γ-carboxylation of proteins that function in various biological processes

GAS6 (Growth Arrest Specific 6)

γ-carboxylated protein that binds TAM receptors and stimulates cell proliferation

GLA protein

protein with γ-carboxylated residues

Matrix metalloproteinase (MMPs)

a family of zinc-dependent proteases capable of degrading extracellular matrix proteins

Matrix Gla Protein (MGP)

γ-carboxylated protein that regulates calcification

Menaquinones

vitamin K2 family of compounds with 4–14 prenyl units in side chain; can be bacterially derived or generated in mammalian tissues from PK

Mitogen-activated protein kinases (MAPKs)

protein kinases including ERK, p38 and JNK that regulate cell growth, differentiation, survival, and immune function in response to extracellular stimuli

Osteoblasts

differentiated bone cells; secrete GLA proteins involved in bone formation and remodeling

Pathologic calcification

abnormal deposition of calcium in soft tissues including blood vessels

Phosphatidylinositol-3-kinase (PI3K)/protein kinase B (Akt) pathway

enzymatic regulator of signal transduction and biological processes involved in cell growth, cell cycle and apoptosis

Phylloquinone

vitamin K1, most abundant dietary form of vitamin K, present in plant-based foods

Protease-activated receptors (PAR)

G protein coupled receptors that are activated by proteolytic cleavage of their N-terminal extracellular domain

Protein C (PROC)

γ-carboxylated protein with anti-coagulant activity; activated Protein C (aPC) is the processed form of PROC that binds the endothelial protein C receptor (EPCR)

Protein S (PROS1)

γ-carboxylated protein with anti-coagulant activity that binds TAM receptors

RDA

recommended dietary allowance

Reactive oxygen species (ROS)

oxygen-containing radicals that when in excess can damage organelles and membranes

Signal transducer and activator of transcription (STAT)

transcription factor that contributes to signal transduction by cytokines, hormones and growth factors

Spheroids

three-dimensional cell clusters used to mimic in vivo tissue organization

Steroid and xenobiotic receptor (SXR)/Pregnane X receptor (PXR)

a nuclear hormone receptor activated by drugs, hormones, and xenobiotic compounds

TAM receptors

the TYRO, AXL and MERTK family of membrane receptors

Thrombosis

formation or presence of a blood clot in a blood vessel or heart

UBIAD1 (UbiA Prenyltransferase Domain-Containing Protein 1)

enzyme that synthesizes MK4 from longer chain MKs or PK

Unc-51 like autophagy activating kinase (ULK1)

a serine/threonine kinase that is involved in the initiation of autophagy

Vitamin K antagonists

drugs such as warfarin that reduce blood clotting by inhibiting the vitamin K dependent γ-carboxylation of coagulation factors

Vitamin K oxidoreductase (VKOR)

endoplasmic reticulum localized enzymes that catalyze the reduction of vitamin K epoxides in the vitamin K cycle