Vitamin K active compounds – study of content and bioavailability

Vitamin K describes a group of fat-soluble vitamers namely phylloquinone (PK) and menaquinones (MKs). Vitamin K was first discovered as a blood coagulation factor in chicken. Evidence is now emerging correlating vitamin K with a range of other health effects, such as improvement of bone health, prevention of cardiovascular diseases, and prevention of cancer. At present, limited knowledge of the health effects of vitamin K and limited knowledge of the content of vitamin K in food, makes it impossible to set recommended intake values for both PK and MKs. However, an adequate intake value of 70 μg PK/day has been set by EFSA.

PK is found in green vegetables, while MK-4 is primarily found in products of animal origin, whereas long chained MKs are produced by bacteria and found in fermented products. Various methods have been utilised to quantify vitamin K through time, however these methods tend to focus on a specific food group or a limited selection of vitamin K vitamers. The aim of this project was therefore to develop and validate a method for quantification of a broad range of vitamin K vitamers in a broad range of food types. A method including extraction, lipase treatments, SPE clean-up and LC-ESI-MS/MS quantification of PK, MK-4, MK-7 and MK-9 using deuterium (d) labelled PK (d7-PK), d7-MK-4, d7-MK-7 and d7-MK-9 as internal standards (ISs) was developed, and subsequently extended and validated to quantify MK-5, MK-6, MK-8 and MK-10. The validation of the method included a method comparison with a laboratory, which used a validated LC- fluorescence detection (FLD) method. The two methods were in agreement for the quantified vitamin K content in the food matrices. The resulting precision of the LC-ESI-MS/MS method was < 15% for PK, MK-4, MK-7 and MK-9, < 20% for MK-5, MK-8 and MK-10, and < 25% for MK-6, and a trueness based on recoveries at 80-120%. The LC-ESI-MS/MS was validated on samples representing a broad range of food matrices e.g. dairy, meat, vegetables and microalgae, however it was shown that the method was not applicable for extraction of vitamin K vitamers in supplements. Three different methods were compared for extraction of vitamin K vitamers from supplements. A method described for PK in the U. S. Pharmacopeia was evaluated also to be optimal for MK-4, MK-7, and MK-9. The successful validation of the method included a comparison study with another laboratory, which used LC-FLD for quantification. As only standard reference material exist for PK, method comparisons in collaboration with other institutions was utilised in this project to assess the trueness of the quantified content of MKs in the different food matrices and supplements.

Another aim of the PhD project was to determine the bioaccessibility and absorption of vitamin K vitamers in food and supplements. The starting point was to perform the internationally, standardised static in vitro digestion model INFOGEST 2.0 to investigate the bioaccessibility of vitamin K vitamers in food. It was found, that an addition of a standard meal to every digest was necessary in order to avoid concentration dependency of the bioaccessibility determined using the INFOGEST 2.0 method. The optimised INFOGEST 2.0 method for determination of the bioaccessibility of vitamin K is referred to as the INFOGEST 2.0 – vit K method. The bioaccessibility of vitamin K vitamers was then determined in different food matrices. It was observed that the bioaccessibility of vitamin K in cheese differed significantly between vitamers and between the different cheeses. When the bioaccessibility of vitamin K vitamers was determined in oil, egg, broccoli, natto, microalgae (N. Oceanica) and supplements it was observed that the highest bioaccessibility was observed in egg and oil followed by supplements and the lowest bioaccessibility was observed in N. Oceanica and broccoli. Neither the fat content nor the content of vitamin K in the food matrices could explain the significant differences in bioaccessibility. More studies are needed to fully understand the factors affecting the bioaccessibility of vitamin K vitamers in food and supplements.

The absorption of vitamin K was studied in collaboration with the Institute of Pharmacy at the University of Copenhagen. The mixed micelle phase from the INFOGEST 2.0 – vit K model was then applied in a transwell Caco-2 cell model in an attempt to assess the absorption of vitamin K vitamers. However, the Caco-2 cell monolayer was ruptured when exposed to the mixed micelle phase. Standards of PK, MK-4, MK-7 and MK-9 incorporated into micelles was instead utilised and the permeability and content in Caco-2 cells and filters were determined in collaboration with the department of Pharmacy at the University of Copenhagen. It was observed that the apparent permeability of all four vitamin K vitamers were low (< 1*10^-6 cm/s) indicating the vitamin K vitamers have a low absorption. It was further observed that the apparent permeability was significantly higher for MK-4 compared to PK, MK-7 and MK-9 (P-value < 0.05). It was further tested whether the transporter protein P-glycoprotein (P-gp), which has been identified for efflux transport of PK, mediates transport of MK-4. No indication of P-gp mediated transport of MK-4 was observed, whereas indication of passive diffusion of MK-4 was observed.