Marjolein Herfs

Vitamin K-Antagonists Accelerate Atherosclerotic Calcification and Induce a Vulnerable Plaque Phenotype

Background Vitamin K-antagonists (VKA) are treatment of choice and standard care for patients with venous thrombosis and thromboembolic risk. In experimental animal models as well as humans, VKA have been shown to promote medial elastocalcinosis. As vascular calcification is considered an independent risk factor for plaque instability, we here investigated the effect of VKA on coronary calcification in patients and on calcification of atherosclerotic plaques in the ApoE−/− model of atherosclerosis. Methodology/Principal Findings A total of 266 patients (133 VKA users and 133 gender and Framingham Risk Score matched non-VKA users) underwent 64-slice MDCT to assess the degree of coronary artery disease (CAD). VKA-users developed significantly more calcified coronary plaques as compared to non-VKA users. ApoE−/− mice (10 weeks) received a Western type diet (WTD) for 12 weeks, after which mice were fed a WTD supplemented with vitamin K1 (VK1, 1.5 mg/g) or vitamin K1 and warfarin (VK1&W; 1.5 mg/g & 3.0 mg/g) for 1 or 4 weeks, after which mice were sacrificed. Warfarin significantly increased frequency and extent of vascular calcification. Also, plaque calcification comprised microcalcification of the intimal layer. Furthermore, warfarin treatment decreased plaque expression of calcification regulatory protein carboxylated matrix Gla-protein, increased apoptosis and, surprisingly outward plaque remodeling, without affecting overall plaque burden. Conclusions/Significance VKA use is associated with coronary artery plaque calcification in patients with suspected CAD and causes changes in plaque morphology with features of plaque vulnerability in ApoE−/− mice. Our findings underscore the need for alternative anticoagulants that do not interfere with the vitamin K cycle.

Vitamin K intake and status are low in hemodialysis patients

Vitamin K is essential for the activity of γ-carboxyglutamate (Gla)-proteins including matrix Gla28 protein and osteocalcin; an inhibitor of vascular calcification and a bone matrix protein, respectively. Insufficient vitamin K intake leads to the production of non-carboxylated, inactive proteins and this could contribute to the high risk of vascular calcification in hemodialysis patients. To help resolve this, we measured vitamin K(1) and K(2) intake (4-day food record), and the vitamin K status in 40 hemodialysis patients. The intake was low in these patients (median 140 μg/day), especially on days of dialysis and the weekend as compared to intakes reported in a reference population of healthy adults (mean K(1) and K(2) intake 200 μg/day and 31 μg/day, respectively). Non-carboxylated bone and coagulation proteins were found to be elevated in 33 hemodialysis patients, indicating subclinical hepatic vitamin K deficiency. Additionally, very high non-carboxylated matrix Gla28 protein levels, endemic to all patients, suggest vascular vitamin K deficiency. Thus, compared to healthy individuals, hemodialysis patients have a poor overall vitamin K status due to low intake. A randomized controlled trial is needed to test whether vitamin K supplementation reduces the risk of arterial calcification and mortality in hemodialysis patients.

Impaired vitamin K recycling in uremia is rescued by vitamin K supplementation

In chronic kidney disease, vitamin K-dependent proteins, including the calcification inhibitor matrix Gla protein, are largely uncarboxylated indicating that functional vitamin K deficiency may contribute to uremic vascular calcification. Since the effects of uremia on the vitamin K cycle are unknown, we investigated the influence of uremia and vitamin K supplementation on the activity of the vitamin K cycle and extraosseous calcification. Uremia was induced in rats by an adenine-supplemented diet and vitamin K1 or K2 was administered over 4 and 7 weeks. After 4 weeks of adenine diet, the activity of the vitamin K cycle enzyme γ-carboxylase but not the activities of DT-diaphorase or vitamin K epoxide reductase were reduced. Serum levels of undercarboxylated matrix Gla protein increased, indicating functional vitamin K deficiency. There was no light microscopy-detectable calcification at this stage but chemically determined aortic and renal calcium content was increased. Vitamin K treatment reduced aortic and renal calcium content after 4 weeks. Seven weeks of uremia induced overt calcification in the aorta, heart, and kidneys; however, addition of vitamin K restored intrarenal γ-carboxylase activity and overstimulated it in the liver along with reducing heart and kidney calcification. Thus, uremic vitamin K deficiency may partially result from a reduction of the γ-carboxylase activity which possibly contributes to calcification. Pharmacological vitamin K supplementation restored the vitamin K cycle and slowed development of soft tissue calcification in experimental uremia.

Anticoagulant Effect of Dietary Fish Oil in Hyperlipidemia A Study of Hepatic Gene Expression in APOE2 Knock-in Mice

Objective—In hyperlipidemia, dietary fish oil containing n-3 polyunsaturated fatty acids (PUFA) provokes plasma triacylglycerol lowering and hypocoagulant activity. Using APOE2 knock-in mice, the relation of these fish-oil effects with altered gene expression was investigated. Methods and Results—Male APOE2 knock-in mice, fed regular low-fat diet, had elevated plasma levels of triacylglycerol and coagulation factors. Plasma lipids and (anti)coagulant factors reduced on feeding the mice with fish oil (n-3 PUFA) or, to a lesser degree, with sunflowerseed oil (n-6 PUFA). The fish-oil diet provoked a 40% reduction in thrombin generation. Microarray (Affymetrix) and single-gene expression analysis of mouse livers showed that fish oil induced: (1) upregulation of genes contributing to lipid degradation and oxidation; (2) downregulation of genes of &ggr;-glutamyl carboxylase and of transcription factors implicated in lipid synthesis; (3) unchanged expression of coagulation factor genes. After fish-oil diet, vitamin K–dependent coagulation factors accumulated in periportal areas of the liver; prothrombin was partly retained in uncarboxylated form. Only part of the changes in gene expression were different from the effects of sunflowerseed oil diet. Conclusions—The hypocoagulant effect of n-3 PUFA is not caused by reduced hepatic synthesis of coagulation factors, but rather results from retention of uncarboxylated coagulation factors. In contrast, the lipid-lowering effect of n-3 PUFA links to altered expression of genes that regulate transcription and fatty acid metabolism.

Gla-Rich Protein Is a Potential New Vitamin K Target in Cancer: Evidences for a Direct GRP-Mineral Interaction

Gla-rich protein (GRP) was described in sturgeon as a new vitamin-K-dependent protein (VKDP) with a high density of Gla residues and associated with ectopic calcifications in humans. Although VKDPs function has been related with γ-carboxylation, the Gla status of GRP in humans is still unknown. Here, we investigated the expression of recently identified GRP spliced transcripts, the γ-carboxylation status, and its association with ectopic calcifications, in skin basal cell and breast carcinomas. GRP-F1 was identified as the predominant splice variant expressed in healthy and cancer tissues. Patterns of γ-carboxylated GRP (cGRP)/undercarboxylated GRP (ucGRP) accumulation in healthy and cancer tissues were determined by immunohistochemistry, using newly developed conformation-specific antibodies. Both GRP protein forms were found colocalized in healthy tissues, while ucGRP was the predominant form associated with tumor cells. Both cGRP and ucGRP found at sites of microcalcifications were shown to have in vitro calcium mineral-binding capacity. The decreased levels of cGRP and predominance of ucGRP in tumor cells suggest that GRP may represent a new target for the anticancer potential of vitamin K. Also, the direct interaction of cGRP and ucGRP with BCP crystals provides a possible mechanism explaining GRP association with pathological mineralization.

Insights into the association of Gla-rich protein and osteoarthritis, novel splice variants and γ-carboxylation status.

SCOPE Gla-rich protein (GRP) is a vitamin K dependent protein, characterized by a high density of γ-carboxylated Glu residues, shown to accumulate in mouse and sturgeon cartilage and at sites of skin and vascular calcification in humans. Therefore, we investigated the involvement of GRP in pathological calcification in osteoarthritis (OA). METHODS AND RESULTS Comparative analysis of GRP patterning at transcriptional and translational levels was performed between controls and OA patients. Using a RT-PCR strategy we unveiled two novel splice variants in human-GRP-F5 and F6-potentially characterized by the loss of full γ-carboxylation and secretion functional motifs. GRP-F1 is shown to be the predominant splice variant expressed in mouse and human adult tissues, particularly in OA cartilage, while an overexpressing human cell model points it as the major γ-carboxylated isoform. Using validated conformational antibodies detecting carboxylated or undercarboxylated GRP (c/uc GRP), we have demonstrated cGRP accumulation in controls, whereas ucGRP was the predominant form in OA-affected tissues, colocalizing at sites of ectopic calcification. CONCLUSION Overall, our results indicate the predominance of GRP-F1, and a clear association of ucGRP with OA cartilage and synovial membrane. Levels of vitamin K should be further assessed in these patients to determine its potential therapeutic use as a supplement in OA treatment.

Vitamin-K-Dependent Protection of the Renal Microvasculature: Histopathological Studies in Normal and Diseased Kidneys

Vitamin-K-dependent carboxylation of matrix Gla protein (MGP) protects the macrocirculation against calcification. We recently reported in a multiethnic population study that the estimated glomerular filtration rate, a microvascular trait, decreased and the risk of chronic kidney disease increased with higher circulating levels of inactive dephospho-uncarboxylated MGP, a marker of vitamin K deficiency. These findings highlighted the possibility that vitamin K might have a beneficial effect on the renal microcirculation. To substantiate these epidemiological findings, we undertook a pilot study, in which we stained renal tissue samples obtained by biopsy from 2 healthy kidney donors and 4 patients with nephropathy for carboxylated and uncarboxylated MGP and calcium deposits. Three patients had renal calcifications, which were consistently associated with carboxylated and uncarboxylated MGP. Normal renal tissue was devoid of microcalcifications and staining for carboxylated and uncarboxylated MGP. Pending confirmation in a larger study covering a wider range of renal pathologies, these histopathological findings suggest that MGP might inhibit calcification not only in large arteries, as was known before, but in renal tissue as well, thereby highlighting potentially new avenues for promoting renal health, for instance by vitamin K supplementation.

Gas6 protein: its role in cardiovascular calcification.

BackgroundCardiovascular calcifications can be prevented by vitamin K and are accelerated by vitamin K antagonists. These effects are believed to be mainly mediated by the vitamin K-dependent matrix Gla protein. Another vitamin K-dependent protein, Gas6, is also expressed in vascular smooth muscle cells (VSMC). In vitro Gas6 expression was shown to be regulated in VSMC calcification and apoptotic processes.MethodsWe investigated the role of Gas6 in vitro using VSMC cultures and in vivo in young and old Gas6-deficient (Gas6-/-) and wildtype (WT) mice. In addition, Gas6-/- and WT mice were challenged by (a) warfarin administration, (b) uninephrectomy (UniNX) plus high phosphate diet, or (c) UniNX plus high phosphate plus electrocautery of the residual kidney.ResultsIn vitro VSMC from WT and Gas6-/- mice exposed to warfarin showed increased apoptosis and calcified similarly. In vivo, aortic, cardiac and renal calcium content in all groups was similar, except for a lower cardiac calcium content in Gas6-/- mice (group a). Von Kossa staining revealed small vascular calcifications in both WT and Gas6-/- mice (groups a-c). In aging, non-manipulated mice, no significant differences in vascular calcification were identified between Gas6-/- and WT mice. Gas6-/- mice exhibited no upregulation of matrix Gla protein in any group. Cardiac output was similar in all treatment groups.ConclusionsTaken together, in our study Gas6 fails to aggravate calcification against the previous assumption.