Association between Polymorphisms in Vitamin D Pathway-Related Genes, Vitamin D Status, Muscle Mass and Function: A Systematic Review

An association between vitamin D level and muscle-related traits has been frequently reported. Vitamin D level is dependent on various factors such as sunlight exposure and nutrition. But also on genetic factors. We, therefore, hypothesize that single nucleotide polymorphisms (SNPs) within the vitamin D pathway-related genes could contribute to muscle mass and function via an impact on vitamin D level. However, the integration of studies investigating these issues is still missing. Therefore, this review aimed to systematically identify and summarize the available evidence on the association between SNPs within vitamin D pathway-related genes and vitamin D status as well as various muscle traits in healthy adults. The review has been registered on PROSPERO and was conducted following PRISMA guidelines. In total, 77 studies investigating 497 SNPs in 13 different genes were included, with significant associations being reported for 59 different SNPs. Variations in GC, CYP2R1, VDR, and CYP24A1 genes were reported most frequently, whereby especially SNPs in the GC (rs2282679, rs4588, rs1155563, rs7041) and CYP2R1 genes (rs10741657, rs10766197, rs2060793) were confirmed to be associated with vitamin D level in more than 50% of the respective studies. Various muscle traits have been investigated only in relation to four different vitamin D receptor (VDR) polymorphisms (rs7975232, rs2228570, rs1544410, and rs731236). Interestingly, all of them showed only very low confirmation rates (6–17% of the studies). In conclusion, this systematic review presents one of the most comprehensive updates of the association of SNPs in vitamin D pathway-related genes with vitamin D status and muscle traits in healthy adults. It might be used for selecting candidate SNPs for further studies, but also for personalized strategies in identifying individuals at risk for vitamin D deficiency and eventually for determining a potential response to vitamin D supplementation.


Introduction
In recent years, the association between "optimal" serum levels of vitamin D with different healthcare conditions has been given important attention in medical research. Studies have shown that various factors such as season, latitude (ultraviolet B (UVB) availability), air pollution, clothing style, sunshine exposure, skin pigmentation, sunscreen cream, age, diet, and nutritional supplementation directly or indirectly affect vitamin D status [1][2][3].
In the body, vitamin D occurs in different forms with several enzymes being involved in their metabolism. Briefly, vitamin D2 (ergocalciferol) built from the provitamin ergosterol and vitamin D3 (cholecalciferol) originating from 7-dehydrocholesterol Figure 1. Vitamin D pathway, candidate genes (in bold), and associated enzymes. DHCR7 (7dehydrocholesterol) gene encodes 7-DHC (7-dehydrocholesterol) reductase enzyme, which converts 7-DHC to cholesterol; CYP2R1 (cytochrome P450 family 2 subfamily R member 1), CYP3A4 (cytochrome P450 family 3 subfamily A member 4), and CYP27A1 (cytochrome P450 family 27 subfamily A member 1) genes encode 25-hydroxylation cytochrome P450 enzymes responsible for converting provitamin D that is absorbed from the diet or synthesized from the action of sunlight on the skin to the circulating form 25(OH)D (25-hydroxyvitamin D); vitamin D is transported bound to vitamin D binding protein (DBP) (encoded by GC gene); LRP2 and CUBN genes encode plasma membrane receptors megalin and cubilin, respectively (involved in re-absorption of 25(OH)D via receptor mediated endocytosis); CYP27B1 encodes the cytochrome p450 enzyme which coverts 1-alpha-hydroxylates 25(OH)D to the active form 1,25(OH) 2 D (1,25-Dihydroxycholecalciferol, Calcitriol); CASR (calcium sensing receptor) binds calcium in extracellular matrix, impacting calcium homeostasis; Ca homeostasis impacts the synthesis of parathyroid hormone (PTH gene-a protein coding gene) which stimulates the synthesis of 1,25(OH) 2 D from 25(OH)D by upregulating renal 1-α-hydroxylase; CYP24A1 encodes a 24-hydroxylase enzyme which catalyzes the degradation of 25(OH)D and 1,25(OH) 2 D in inactive metabolites; VDR encodes the vitamin D receptor, a nuclear receptor which binds 1,25(OH) 2 D and forms a heterodimer with the gene product of RXR-the retinoid X receptor-to mediate the biological actions of vitamin D.
The final search for PubMed was conducted on 27 November 2018 and followed a protocol developed a priori. The search strategy aimed to identify all articles where the specific genetic variants (i) were aligned with search terms for either vitamin D status (ii), muscle traits (iii), or both (ii AND iii): (i) each of the 14 vitamin D-pathway-related genes ( OR "muscle mass" OR "lean body mass" OR "muscle quality" OR "physical performance" OR SPPB OR "short physical performance battery" OR "handgrip strength" OR "chair stand" OR "arm curl") NOT Review[Publication Type], Filters: Humans; English; German After conducting the initial structured search as outlined above, additional studies were added based on the reference lists of the finally selected studies (hand search).

Inclusion and Exclusion Criteria
Articles were included if studies: (i) were conducted among humans aged >18 years from both sexes; (ii) comprised candidate or genome-wide association studies being cross-sectional, cohort, case control, or intervention studies; (iii) investigated healthy subjects or contained at least a separate control group with normal health status; (iv) measured at least one genotype in a vitamin D pathway-related gene; (v) reported vitamin D status (circulating plasma/serum levels, 25(OH)D, 25(OH)D2, or 25(OH)D3 metabolites) and/or reported results for muscle mass or function, muscle strength parameters or scores for the Short Physical Performance Battery (SPPB); (vi) were published in English or German.
Articles were excluded if they: (i) reported only vitamin D intake or vitamin D metabolites from urine; (ii) investigated children (<18 years old); (iii) investigated participants with severe chronic or acute illnesses with a known impact on either vitamin D status and/or muscle mass and strength (myopathies, hypocalciuric hypercalcemia), or pregnant or lactating women; (iv) were published as case reports, systematic reviews, or meta-analyses.

Study Selection and Data Extraction
Study selection and data extraction was performed by two independent reviewers (EK, BW) in accordance with the above-mentioned inclusion and exclusion criteria. A third reviewer (KHW) was included in case of disagreements.
Data extraction from eligible studies included information as follows: (i) gene, (ii) SNP, (iii) bibliographic information, (iv) study design, (v) participants (age, gender, and ethnicity), (vi) sample size, (vii) main findings and outcomes, (viii) location, sampling season and analysis method of vitamin D level if available.
A narrative synthesis of the findings from the selected studies was performed using two frameworks in accordance with the following specific research questions: (i) association of vitamin D-related genetic polymorphisms and vitamin D status, and (ii) association of vitamin D-related genetic polymorphisms and muscle mass or strength. Results were described qualitatively rather than performing a quantitative meta-analysis as study designs and reported outcomes differed widely in order to quantitatively analyze them.

Risk of Bias (Quality) Assessment
In order to assess the quality of included studies, the STREGA recommendations (STrengthening the REporting of Genetic Association studies guidelines) were applied independently by two reviewers (three if there was any disagreement even after discussion) [25]. The quality of the studies was considered as "high" when the score was 18-22, "moderate-high" when the score was 13-17, and "low" with a score below 12. Results were reported, but no restrictions were made with respect to the inclusion or exclusion of the relevant studies.

Study Selection and Characteristics
In total, 1292 studies were identified from the initial searches in PubMed (n = 1282) and through other sources (n = 10, Figure 2). As the searches were conducted separately for each of the 14 vitamin D pathway-related genes and then combined, 161 articles had to be excluded as duplicates. During title/abstract screening, a further 936 articles were excluded based on the predefined inclusion and exclusion criteria. The remaining 195 articles were assessed for eligibility based on the full texts. Among these, a further 106 studies were not suitable with respect to the research question. Consequently, 89 studies were included in the systematic review, with 77 of them reporting the association of genetic polymorphisms of vitamin D-related genes and vitamin D status, and only 12 studies dealing with genetic variants of vitamin D-related genes and muscle mass and/or function. Study designs included 57 cross-sectional studies, 20 case-control studies, and 12 intervention studies.
Nutrients 2021, 13, 3109 5 of 23 ticles were assessed for eligibility based on the full texts. Among these, a further 106 studies were not suitable with respect to the research question. Consequently, 89 studies were included in the systematic review, with 77 of them reporting the association of genetic polymorphisms of vitamin D-related genes and vitamin D status, and only 12 studies dealing with genetic variants of vitamin D-related genes and muscle mass and/or function. Study designs included 57 cross-sectional studies, 20 case-control studies, and 12 intervention studies.

Polymorphisms in Vitamin D Pathway-Related Genes and Vitamin D Status
As summarized in Table 1, 77 publications that have investigated a potential association between a certain genetic association and vitamin D levels were included in the report. Most of these publications were cross-sectional studies (n = 46); a further 19 studies were conducted as case-control studies and 12 studies as interventions. Altogether, these studies investigated 497 SNPs in 13 different genes (GC, CYP2R1, VDR, CYP24A1, DHCR7, CYP27B1, CYP27A1, CASR, PTH, CYP3A4, RXRA, CUBN, and RXRB). No suitable records were found for the gene lipoprotein receptor-related protein 2 (megalin, LRP2).
The most frequently studied gene comprised the vitamin D binding protein (GC) which was investigated in 56 studies, followed by CYP2R1, coding for a vitamin D 25hydroxylase, which was mentioned in 41 studies, and the vitamin D receptor (VDR) having been subject to 41 studies. In total, 59 SNPs located within 10 different genes showed a significant association with vitamin D levels in at least one study. Most importantly, 23 of these SNPs were confirmed to be related to vitamin D status in at least two other studies (Table 1). For genetic variants in the CYP27A1 gene (vitamin D 25-hydroxylase), CUBN gene (cubilin), and RXRB gene (retinoid-X receptor B), none of the studies reported a significant association with vitamin D level [26,27,30,33,71,101].
Most of the studies were performed in Europe (21 studies from Belgium, Czech Republic, Denmark, Finland, France, Germany, Hungary, Netherlands, Norway, Estonia, UK, France, Italy, Greece, Spain, Scotland, Sweden, United Kingdom), followed by the US (14 studies) and China (11 studies).

Polymorphisms in Vitamin D Pathway-Related Genes and Muscle Mass and Function
As reported in Table 2, 12 publications were included in the qualitative synthesis, whereby 11 studies were conducted as cross-sectional studies and one study as a casecontrol study [103]. All the selected studies were focusing on potential associations between VDR gene polymorphisms and muscle traits, investigating only four different SNPs in this gene [rs7975232 (alias ApaI), rs1544410 (alias BsmI), rs2228570 (alias FokI, including the merged SNP rs10735810), and rs731236 (alias TaqI)].  FF→significantly lower appendicular and total fat-free mass than Ff and ff; FF→significantly lower peak and average isometric quadriceps strength than ff In total, 5342 healthy subjects were included, with the number of participants ranging from 104 (case-control study) [103] to 1970 (cross-sectional study) [110]. The most frequently studied SNP was rs1544410 (BsmI), investigated by 11 studies, whereby five studies were reporting significant associations between its genotypes (BB, Bb, bb) and muscle traits such as knee flexion peak torque [104], knee extensor strength [109], maximal power [110], hamstring strength [112] and quadriceps strength [114]. Four studies included only female subjects [104,110,112,114].

Quality of Included Studies
The STREGA quality score for the studies relating the respective SNPs to vitamin D status was 18.8 ± 2.3 showing low to high quality with a range between 11 and 22. While for studies relating SNPs to muscle traits, the mean STREGA score was 16.8 ± 1.8 with a range between 13 and 19, indicating moderate to high quality.

Discussion
We have systematically evaluated the available scientific data showing the association of certain genotypes to vitamin D deficiency, and hence, poor muscle status. Variations in GC, CYP2R1, VDR, and CYP24A1 genes were reported most frequently, whereby especially SNPs in the GC (rs2282679, rs4588, rs1155563, rs7041) and CYP2R1 genes (rs10741657, rs10766197, rs2060793) were confirmed to be associated with vitamin D plasma level in more than 50% of the respective studies. Various muscle traits have been investigated only in relation to four different VDR polymorphisms (rs7975232, rs2228570, rs1544410, and rs731236). Interestingly, all of them showed only very low confirmation rates (6-17% of the studies).
Synthesized or consumed with the diet, in the liver, vitamin D is converted in its circulating form 25-hydroxyvitamin D (calcidiol), a process mediated by enzyme 25-hydroxylase, which is encoded by the CYP2R1 gene. Polymorphisms in this gene may impact vitamin D metabolism, while it shows catabolic effects toward Vitamin D2 and D3 by modulating 25-hydroxylase's activity and expression [115]. Accordingly, a significant impact of 9 polymorphisms on this gene (Table 1) with vitamin D status has been found to be replicated in more than two studies. Rs10741657, located in non-coding region 5 -UTR, was the most frequently studied SNP of this gene, with a significant association confirmed in 66% of the conducted studies with the beneficiary allele being the minor one. This data is in line with a recent meta-analysis on the effects of CYP2R1 gene variants on vitamin D levels [116]. Being located in 2-kb CYP2R1 mRNA transcript, it is believed that this SNP is able to change enzyme activities and affect vitamin D metabolism [117]. At the same time, two other SNPs, rs12794714 and rs10766197, located in the coding region of introns with the possibility of altering transcription rate [117] are confirmed in 79 and 60%, respectively, with the minor allele being the risk allele in the two aforementioned SNPs (Supplementary Table S1).
The transport of 25(OH)D toward target tissues for utilization and processes is enabled by vitamin D binding protein (DPB). This protein binds 85 to 90% of vitamin D circulating form, having the role of both carrier and reservoir. The remaining 10-15% of the circulating vitamin D is bound to albumin, or unbound, representing the available fraction based on the free hormone hypothesis [97]. This multifunctional and polymorphic protein is encoded by the GC gene (located on chromosome 4q12-q13), whose DNA sequence may impact the binding ability to vitamin D since its isoforms have different binding affinities, possibly impacting the half-life of circulating 25(OH)D [28]. From at least 120 identified isoforms, the most common ones Gc1f and Gc1s (rs7041 locus), as well as Gc2 (rs4588 locus), refer to the two functional SNPs in exon 11 with rs7041 causing an Asp→Glu amino acid change and rs4588 causing a Thr→Lys exchange in the vitamin D binding protein [118]. Interestingly, those two SNPs were found to be significantly associated with the vitamin D level in 69 and 73% of the respective studies included in this review. In the aforementioned SNPs, the major allele was the beneficiary one in the majority of the studies (70 and 93%, respectively). In addition, this gene's polymorphisms are believed to influence the circulating concentration of DBP, which may alter the bioavailable circulating vitamin D [119]. This association might also be linked to the possible impact of these SNPs in the affinity of DPB to actin while modifying its actin-binding region and affecting 25(OH)D uptake and retention into skeletal muscle cells. Furthermore, it has been shown that C2 myotubes and primary rat muscle fibers express megalin and cubilin receptors, which enable endocytosis of DBP [120]. Studies on megalin and cubilin published after the search closing date of this systematic review did not reveal new findings [121,122]. In addition to the previously mentioned functional polymorphisms in the GC gene, there are also intron-located SNPs, which were found to be associated with vitamin D status, whereas the underlying mechanism remains unclear. One such, rs2282679, an intron variant (in linkage disequilibrium with rs4588) [123], was found to be significantly associated with vitamin D in 77% of the identified studies, all confirming the major allele as the beneficiary one. Finally, rs1155563 (an intron variant) was confirmed in 71% of the studies (of which 71% confirmed the major allele as the beneficiary one).
Furthermore, another important part of the vitamin D pathway chain is the vitamin D receptor, a high-affinity nuclear receptor encoded by the VDR gene. Vitamin D exerts its biological roles when its active metabolite 1alpha,25-dihydroxyvitamin D3 [1alpha,25(OH)2D3] binds to VDR, causing a transactivation function of VDR [124]. The resulting complex 1,25D-VDR-RXR then binds to vitamin D response elements in the DNA [125]. Consequently, VDR is involved in the regulation of many cellular functions such as phospholipid metabolism, apoptosis, cell differentiation, and oxidative stress. It also affects the expression of the vitamin D metabolism-related genes CYP27B1 and CYP24A1 [126,127]. Several studies aimed to prove the association between VDR gene polymorphism and vitamin D status. Among one of the most studied SNPs, Fok1 is a polymorphism located at the start codon of the coding part, whose polymorphic form produces a protein shorter by amino acids [128], altering the length of the VDR [106]. However, an association with vitamin D level has been confirmed in only 16% of the studies included in this systematic review. The exact role of other frequently studied polymorphisms in VDR is not fully elucidated. Three SNPs located in the 3 end of the VDR gene rs1544410 (Bsm1), rs731236 (Taq1), and rs7975232 (Apa1), considered to be in high linkage disequilibrium with 3 UTR polymorphisms [129]. An effective role of these genetic variants on vitamin D level is very unlikely as they have been associated with the vitamin D level in only 14, 8, and 17% of the studies, although investigated in 28, 24, and 18 studies, respectively. Although the regions around these SNPs are not translated to the VDR protein, they might have a role in mRNA stability because of their neighborhood to the poly-A tail [130].
Taken together, there seems to be limited evidence that genetic variations in the VDR gene will exert a meaningful association with vitamin D level, given the small number of studies showing significant associations in relation to the high number of studies investigating a potential association. This is further supported by the notation that even in those studies showing a significant association, there was no clear direction with respect to the identification of a beneficiary allele (Supporting File S1).
In addition, two SNPs on the DHCR7 genes have been highly investigated. This gene encodes the enzyme 7-dehydrocholesterol reductase, a key metabolite enzyme that catalyzes the conversion of 7-dehydrocholesterol to cholesterol [28]. Rs12785878 and rs1790349 were found to be significantly associated with the vitamin D level in 32 and 50% of the included studies, while the major allele could be identified to be the beneficiary one in 83 and 80% of the studies. Therefore, no clear conclusion can be drawn on the involvement of these SNPs in affecting vitamin D level, which is confirmed in recent studies on adolescents with rs12785878 genotype showing no association to hypovitaminosis D [131], but an interaction between 25(OH)D levels and rs12785878 genotype in DHCR7 on overall survival of patients with metastatic colorectal cancer [132].
While vitamin D deficiency is an important public health topic, at least some vitamin D-related gene polymorphisms seem to play an important role in vitamin D status. However, an in-depth analysis of the study characteristics revealed (see Supplementary  Table S1) that the included studies were characterized by heterogeneous methodology, including varying sample sizes, age groups, and, most importantly, different vitamin D measurement techniques. After concerns were raised about the accuracy of different vitamin D assays and the possibilities of misleading assessment of vitamin D levels [133], EFSA recommended liquid chromatography-tandem mass spectroscopy (LC-MS/MS) as the reference method in regard to 25(OH)D concentrations [4]. Notwithstanding, LC-MS/MS was found to be used in only 17 out of 77 studies (22.1%) included in this qualitative analysis, while the radioimmunoassay method (RIA) was the most used one (in 22 of 77 studies, 28.6%). It also must be noted that the search for genotypes differed as well-as most of the studies used a candidate genotype approach, although some SNPs were identified via SNP arrays [26][27][28]30,37,38,52,60,62,65,74,80,81,97].
As vitamin D level has been shown to be low in a significant proportion of adults worldwide [134], vitamin D supplementation remains an important method to achieve optimal levels. Its ability to enhance muscle strength [135], physical performance [17,18], including the lowering of the risk of falling in older adults with low serum 25(OH)D [17,136], has already been demonstrated. Notwithstanding, genetics might also affect the metabolic response toward vitamin D supplementation [80]. Accordingly, existing data implicate that polymorphisms in CYP2R1 (rs10766197, rs10741657), GC (rs4588, rs7041, rs2282679,) and VDR gene (rs2228570,) are associated with vitamin D dose-response, in view of the fact that these were individually replicated in at least two different intervention studies [27,35,40,41,54,56,60,74,78,80]. Although this was not the primary aim of this systematic review, it might comprise important information towards the necessity of personalized vitamin D treatment due to a possible intra-individual variability. It should be noted that we have not encountered studies investigating the direct link between vitamin D pathway-related gene polymorphisms and the effect of vitamin D supplementation on muscle traits. Although, data supporting the impact of these genes' polymorphisms in vitamin D status implicates their potential effect of vitamin D status in health outcomes such as muscle performance, particularly in vitamin D supplementation improvement in older adults with 25(OH)D levels <37-45 nmol/L [17,18].
While some outcomes of vitamin D deficiency, such as osteomalacia and osteoporosis are well known, the implications of vitamin D in muscle strength and function are still being investigated. Studies on vitamin D pathway-related genotypes and muscle traits were exclusively focused on the VDR gene. Genomic and non-genomic pathways might explain the effects of vitamin D on muscles. While the genomic effect is mediated through the already mentioned 1,25D-VDR-RXR complex, the non-genomic effect involves intracellular calcium and phosphate homeostasis resulting from transcriptional regulation of specific proteins in organs such as intestines, bone, and parathyroid gland [125]. In this respect, it has been shown that muscle fibers of VDR-null mice were smaller, more variable in size, and accompanied by abnormal expression of myoregulatory transcription factors (myf5, myogenin, and E2A). Hence, it is believed that VDR may be involved in transcriptional down-regulation of these factors during muscle differentiation [137]. Whereas these implications and underlying biology are still being studied [138], this review shows rs2228570 and rs1544410 within the VDR gene to be the most frequently investigated polymorphisms also with respect to their impact on skeletal muscle traits. Interestingly, the identified studies showed non-conclusive results as, i.e., upper body strength (major allelethe beneficiary one) [106] was controversially affected by the FokI (rs2228570) genotype than the lower body strength (minor allele-the beneficiary one) [103,108,111,113]. Nevertheless, this demands further mechanistic investigations. For BsmI (rs1544410) genotype, the positive impact of major alleles in lower body muscles was confirmed in four studies [104,109,110,112]. A recent study, not being included in the systematic review focusing on further SNPs in the VDR gene (rs9729, rs17882106, rs7136534, rs11568820, rs10735810, rs4516035, and 11574024) did not reveal new findings, as neither muscle strength nor physical performance were associated to these genotypes [139]. To date, studies investigating the direct impact of vitamin D pathway-related genes (other than VDR gene) and muscle traits are still lacking.
Finally, this systematic review highlighted that there are promising candidate SNPs in vitamin D pathway-related genes that might impact vitamin D level and eventually muscle traits. However, it should be noted that heterogeneity among the selected studies represents a potential limitation, which also caused the decision to refrain from conducting a metaanalysis. Despite this limitation, the strength lies in the extensive information on individual SNPs in most of the relevant vitamin D pathway-related genes. To extract this detailed information from all included studies caused the rather long duration from the underlying systematic search to reporting the results. However, no conflicting results were found when comparing the outcomes of our study to recently published data [121,122,[140][141][142][143][144][145][146][147][148][149]. The focus of recently published data remains in the same gene's polymorphisms: GC, CYP2R1, VDR, CYP24A1, and CYP27B1. Except for these genes, Fediriko et al., 2019 and Jorde et al., 2019 identified potentially novel SNPs in vitamin D-related candidate genes (LRP2 and CUBN), but none of those were statistically significant [121,122].

Conclusions
To the best of our knowledge, this systematic review presents a very comprehensive update of the association of polymorphisms in vitamin D pathway-related genes with vitamin D status in healthy adults. While especially SNPs in the GC (rs2282679, rs4588, rs1155563, rs7041) and CYP2R1 genes (rs10741657, rs10766197, rs2060793) were confirmed to be associated with vitamin D levels in more than 50% of the respective studies, various muscle traits have been investigated only in relation to four different VDR polymorphisms (rs7975232, rs2228570, rs1544410, and rs731236) and outcomes remain inconclusive. Taken together, these data could be used in various ways: (1) to use the identified SNPs as candidate genes to be validated in further studies, (2) to identify individuals at potential risk, and (3) to optimize potential interventions with all these suggestions being important for precision nutrition.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The data presented in this study are available in Supplementary  Table S1.