**3. Epidemiological and clinical studies on vitamin K and aging-related skeletal diseases**

A traditional Japanese food, "natto" (fermented soybeans) contains high concentrations of MK-7, a form of vitamin K2 (menaquinone), synthesized by microorganisms. Epidemiological study conducted in Japan revealed negative correlation of Natto intake and incidence of hip fracture [13], which drew attention toward possible link between vitamin K and osteoporosis. Later, among several nutrients including vitamin D and calcium, vitamin K was shown to be the only nutrient that is significantly correlated with hip fracture incidence in Japanese population [14]. Furthermore, the fracture-preventing effect of vitamin K was observed in several clinical studies in Japan, which was confirmed by meta-analysis [15]. Based on these results, vitamin K2 is used for treatment of osteoporosis in several Asian countries. We previously reported a functional single nucleotide polymorphism (SNP) in *GGCX* that causes higher enzymatic activity correlated with higher bone mineral density in elderly Japanese women [16], suggesting bone-protective function of vitamin K is related to GGCX activity. Osteocalcin, one of the substrates of GGCX, is specifically expressed in osteoblastic lineage. The concentration of undercarboxylated form of osteocalcin (ucOC) in serum was reported to be positively correlated with fracture risk [17]. Measurement of ucOC has been clinically used to decide the indication of vitamin K for treatment of osteoporosis in Japan. These support the contribution of GGCX activity to bone-protective effect.

Recently, another mode of GGCX-dependent vitamin K function was reported in the study of proapoptotic effect of vitamin K. Handa et al. found proapoptotic protein Bak was covalently modified by vitamin K epoxide and regulated by its modification [5]. This function is dependent on GGCX-mediated vitamin K function since GGCX activity is required to generate vitamin

On the other hand, we discovered GGCX-independent mode of vitamin K function mediated by transcriptional regulation [3] as compared to posttranscriptional modifications explained above. Vitamin K was found to be one of the ligands of the nuclear receptor, SXR, and its murine ortholog, PXR. This receptor is also called NR1I2 according to standardized nomenclature designated by the nuclear receptor committee. In 1998, SXR/PXR was cloned as a novel nuclear receptor that is mainly expressed in the liver and intestine [6]. At first, its functions were characterized as a ligand-dependent transcription factor which is activated by various pharmaceutical agents and xenobiotic compounds [7]. It was originally classified as an orphan receptor since the endogenous ligand was not known when it was cloned. It was later shown that some kinds of secondary bile acids (such as lithocholic acid) could be endogenous ligands for this receptor [8, 9]. It forms a heterodimer with 9-cis-retinoid acid receptor (RXR) on ligand stimulation. This complex then binds to SXR-responsive elements (SXRE) in the promoter or enhancer regions of target genes (**Figure 1**). Some of its target genes are the drug-metabolizing enzyme, such as *CYP3A4*, and the ABC (ATP-binding cassette) family transporter, *MDR1*. Because of that, a function of SXR/PXR is considered as a xenobiotic sensor-inducing genes involved in detoxification and drug excretion [10] and named as such. The discovery of novel vitamin K function as a ligand for SXR/PXR indicated that physiological and pathological

There is another mode of vitamin K function which modulates activation of signal transduction pathway. This is inferred by existence of some genes induced by vitamin K, not by SXR agonist, rifampicin [11]. This induction was not affected by knocking down of GGCX suggesting that this is γ-carboxylation-independent pathway. Expression of those genes was suppressed by protein kinase A (PKA) inhibitor, showing the novel vitamin K function as a modulator of PKA

Inhibition of another protein kinase, protein kinase C (PKC) α and ε, by vitamin K was also reported [12]. Inhibition of IKK (inhibitor of nuclear factor kappa B kinase) and subsequent inhibition of NFkB (nuclear factor kappa B) were observed. Whether this function of vitamin

**3. Epidemiological and clinical studies on vitamin K and aging-related**

A traditional Japanese food, "natto" (fermented soybeans) contains high concentrations of MK-7, a form of vitamin K2 (menaquinone), synthesized by microorganisms. Epidemiological study conducted in Japan revealed negative correlation of Natto intake and incidence of hip fracture [13], which drew attention toward possible link between vitamin K and osteoporosis.

K is independent of mechanisms described above remains to be elucidated.

processes mediated by PXR/SXR would be affected by vitamin K.

K epoxide (**Figure 1**).

24 Vitamin K2 - Vital for Health and Wellbeing

activity (**Figure 1**).

**skeletal diseases**

Vitamin K also has some epidemiological evidences in relationship with another skeletal disease, osteoarthritis. Low vitamin K intake was correlated to the prevalence of osteoarthritis both in North America and in Japan [18–20]. Unfortunately, therapeutic effect of vitamin K for established osteoarthritis was not proven by a trial [21], suggesting that the study period was too short or vitamin K has only preventive effect.

#### **4. Paradoxical GGCX-mediated vitamin K functions on bone metabolism**

It is difficult to evaluate vitamin K function on bone tissue mediated by GGCX *in vivo* due to its dominant effect on coagulation activity. For example, it is impossible to measure bone mineral density of adult mice systemically lacking GGCX because *Ggcx*-knockout mice die before birth or on the day of birth with massive bleeding [22]. To overcome this obstacle, we utilized Cre/loxP system which enables tissue-/organ-specific knockout of GGCX dependent on promoter activity [23] and generated osteoblast-specific GGCX-deficient mice by crossing with *Col1a1*-Cre mice [24]. Since osteoblasts express several substrates of GGCX including osteocalcin, we assumed bone-protective effect of vitamin K is mediated by GGCX activity in osteoblasts. Surprisingly, the bone mineral density was increased in osteoblast-specific GGCXdeficient mice and aberrant mineralization was observed in these mice by ultrastructural analysis. This result indicates that GGCX in osteoblast may not contribute to bone-protective effect of vitamin K. Moreover, it is contradicting to the clinical studies on *GGCX* SNPs or ucOC described above. We speculate that GGCX activity in other tissue is responsible for boneprotective effect of vitamin K and/or vitamin K function mediated by SXR/PXR that would be more important in the bone tissue. Further studies are necessary to clarify this enigma. It is noteworthy that osteocalcin-deficient mice have been shown to have mechanically stronger bone than wild-type mice [25], suggesting that the decrease of carboxylated osteocalcin, rather than increase of ucOC, has "bone strengthening effect."
