Association of vitamin D-binding protein polymorphisms and serum 25(OH)D concentration varies among Chinese healthy infants of different VDR-FokI genotypes: A multi-centre cross-sectional study
Abstract
Hypovitaminosis D during infancy is associated with the development of chronic diseases and poor health later in life. While the effect of environmental factors on vitamin D concentration has been extensively explored, this study aimed to explore the effect of genetic factors on vitamin D concentration among Chinese infants. We conducted a multi-centre cross-sectional study in Hong Kong from July 2019 to May 2021. A candidate genetic approach was adopted to study four selected genetic variants of the vitamin D-binding protein (DBP) and vitamin D receptor (VDR) (rs4588, rs7041, rs2282679 and rs2228570) to examine their associations with measured serum 25(OH)D concentration. A total of 378 Chinese infants aged 2–12 months were recruited in this study. Peripheral blood samples were collected from the infants to measure serum 25(OH)D concentration and extract DNA. Results showed that rs7041T and rs2282679C were significantly associated with lower serum 25(OH)D concentration. Further analysis of the DBP variants revealed that the GC1F allele was significantly associated with lower 25(OH)D concentration and identified as the risk DBP isoform in infants. While our results revealed that there is no direct association between VDR-FokI genotype and serum 25(OH)D concentration, a VDR-FokI genotype-specific pattern was observed in the association between DBP isoforms and serum 25(OH)D concentration. Specifically, significant associations were observed in the DBP genotypes GC1F/F, GC1F/2 and GC1S/2 among VDR-FokI TT/TC carriers, but not in VDR-FokI CC carriers. Our findings lay down the basis for the potential of genetic screening to identify high risk of hypovitaminosis D in Chinese infants.
INTRODUCTION
Vitamin D is a steroid hormone that has a significant regulatory role in bone mineralisation, calcium absorption and immune response activation (Lv et al., 2020; Rozmus et al., 2020). Vitamin D, with a circulatory form of 25-hydroxyvitamin D (25[OH]D), is essential for the health and development of infants and children. Previous research showed that vitamin D deficiency in childhood and adolescence was associated with a wide range of physiological conditions and severe diseases such as rickets and seizures (Tan et al., 2020), type 1 diabetes mellitus (T1DM) (Mathieu, 2015), food allergies and asthma (Hennessy et al., 2018), and mental health problems. (Focker et al., 2017) Therefore, maintaining sufficient vitamin D during infancy is particularly important for preventing long-term health problems due to vitamin D deficiency. However, vitamin D concentrations vary widely between individuals due to the differences in environmental exposures and dietary consumption, especially for infants with limited sunlight exposure and dietary choices. While environmental factors have been well explored (Fink et al., 2019), the potential effect of genetic factors on infant vitamin D concentrations warrants further exploration.
Genetic factors play an important role in determining serum vitamin D concentrations (Shea et al., 2009). Twin studies demonstrated that up to 80% of the inter-individual variation in serum vitamin D concentrations can be explained by inherited genetic factors (Arguelles et al., 2009; Lu et al., 2012). A previous clinical study also found that 80% of the 25(OH)D concentrations were genetically inherited in patients with asthma, a known vitamin D deficiency-associated disease (Wjst et al., 2007). Both candidate gene approach and genome-wide association studies (GWAS) have identified the role of genetic variants in vitamin D metabolism, which could influence serum 25(OH)D concentrations and the development of associated diseases (Cui & Liu, 2018; Manousaki et al., 2020).
Specifically, both vitamin D-binding protein (DBP), also known as GC globin (GC), and vitamin D receptor (VDR) are the major components in vitamin D metabolism (Cui & Liu, 2018; Manousaki et al., 2020). The DBP, encoded by the GC gene, is a carrier protein that binds and transports over 85% of vitamin D metabolites (Bikle & Schwartz, 2019; Newton et al., 2019; Gozdzik et al., 2011; Lee et al., 2016; Powe et al., 2013; Rozmus et al., 2020). GC is highly polymorphic such that its binding affinity to vitamin D metabolites is affected by more than 120 variants (Newton et al., 2019; Gozdzik et al., 2011; Lee et al., 2016). Among all variants, rs2282679, rs7041 and rs4588 are the most frequently studied variants that have been shown to correlate with serum concentrations of 25(OH)D (Cooper et al., 2011; Engelman et al., 2008; Kurylowicz et al., 2006). Functional variants in rs7041 and rs4588 were further found to encode three common isoforms with distinctive protein phenotypes, GC1F, GC1S and GC2 (Cleve & Constans, 1988; Newton et al., 2019; Yousefzadeh et al., 2014). These variants are found in complete linkage disequilibrium, resulting in six different haplotypes only (Nazemisalman et al., 2019). Previous studies observed variations in the binding affinity of different DBP isoforms to 25(OH)D, with GC1F having the highest affinity, followed by GC1S and GC2 (Gozdzik et al., 2011). Moreover, the serum concentration of the DBP was found to be significantly lower in those with the GC2 allele (Lauridsen et al., 2001). However, the functional deviations of different DBP isoforms were not fully delineated, which warrants further exploration. VDR, on the other hand, is a nuclear hormone receptor that binds with the vitamin D3 active metabolite, calcitriol, to regulate physiological functions by driving the expression of vitamin D-responsive genes (Nazemisalman et al., 2019). The receptor activity can be altered by some known gene polymorphisms, including FokI (rs2228570), BsmI (rs1544410), ApaI (rs7975232) and TaqI (rs731236) (Zaki et al., 2017). While some studies reported that BsmI, ApaI and TaqI polymorphisms were associated with 25(OH)D concentrations in healthy children (Karpiński et al., 2017; Santos et al., 2012), there is still little evidence for the FokI polymorphism.
Despite a growing body of research on the direct effect of GC and VDR polymorphisms on 25(OH)D concentrations, the majority of existing research has been conducted in non-Chinese populations. Considering the possible differences in the expression of GC and VDR genes across ethnic and age groups (Newton et al., 2019; Yousefzadeh et al., 2014), the investigations on the effect of these polymorphisms on 25(OH)D concentrations, especially in the Chinese Han infant population, would help to strengthen the current research evidence. Furthermore, emerging evidence suggests that GC and VDR polymorphisms can interact to influence 25(OH)D concentrations (Kiani et al., 2019), but more studies are needed to affirm this suggestion. Therefore, this study adopted a replicative genetic association approach among healthy Chinese infants in Hong Kong to (i) estimate the prevalence of the selected GC and VDR gene polymorphisms, (ii) investigate the association of GC and VDR polymorphisms with 25(OH)D concentrations, and (iii) examine the potential moderating effect of VDR polymorphisms on the association between GC polymorphisms and 25(OH)D concentrations.
METHODS
Study design and participants
This is a multi-centre, cross-sectional study in which infants aged between 2 and 12 months were recruited in different districts of Hong Kong using stratified sampling from July 2019 to May 2021. Upon obtaining informed consent, the mothers were asked to complete a questionnaire on their demographics. Peripheral blood samples were collected via venepuncture from the infants by an experienced phlebotomist. An incentive of an HKD200 (approximately USD 25.6) supermarket voucher was given to the participants to cover the cost of transportation upon completion of the study assessment which is supported by the Health and Medical Research Fund (HMRF), Food and Health Bureau, Hong Kong SAR Government. Infants with major congenital malformations or conditions including low birth weight or premature birth were excluded from this study.
Demographics
Maternal demographic data including her age, marital status and parity of the pregnancy, and infant's information including the age, sex, supplementation practice and breastfeeding status were obtained by a questionnaire completed by the mothers.
Ethics approval and consent to participate
The research protocol, including the consent procedures, genetic information assessment and vitamin D measurement method was approved by the Institutional Review Board of the University of Hong Kong/Hospital Authority Hong Kong West Cluster Research Ethics Committee (UW 13-055 & UW 21-268) and the Department of Health, Hong Kong Special Administrative Region Government, China and in accordance with the Declaration of Helsinki. Written informed consent was obtained from each participating infant–mother dyad.
Serum 25(OH)D concentration of infants
Serum was separated from the collected peripheral blood samples for assessment of vitamin D concentration. Serum 25(OH)D concentration, determined by liquid chromatography–tandem mass spectrometry method (LC–MS/MS), was used to assess the vitamin D concentration in infants. The LC–MS/MS has been found to have high specificity and precision in previous studies (Vitamin D Roundtable on the NHANES Monitoring of Serum 25D: Assay Challenges Options for Resolving Them, 2010). The AB Sciex Triple Quad QTRAP 5500+ LC–MS/MS system (AB Sciex Pte. Ltd., Framingham, MA, USA) was used to simultaneously detect the concentration of 25-hydroxyvitamin D3 (25[OH]D3), 25-hydroxyvitamin D2 (25[OH]D2) and 3-epi-25 hydroxyvitamin D (3-Epi-25[OH]D3). The total serum 25(OH)D concentration, defined as the sum of 25(OH)D3 and 25(OH)D2 and excluding 3-Epi-25(OH)D3, was quantified by referring to the standard curves generated by serial dilutions of accredited standards (MilliporeSigma, St. Louis, MO, USA). Results obtained through the LC–MS/MS method in this study have been certified by achieving satisfactory performance (within ±25% of the target value) in the proficiency test conducted by Vitamin D External Quality Assessment Scheme (DEQAS, Endocrine Laboratory, Charing Cross Hospital, London, UK) (Carter et al., 2017). The inter-assay coefficient of variability (n = 24) was found to be 9.09% in this study.
DNA extraction and genotyping of selected genetic variants
Leucocytic DNA was extracted from the whole blood samples using QIAamp DNA Mini Kit (QIAGEN, Valencia, CA, USA), according to the manufacturer's protocol. The extracted DNA quality was checked with 1% agarose gel electrophoresis, and the quantity was measured by NanoDrop™ spectrophotometer (ThermoFisher Scientific, Waltham, MA, USA).
Four selected genetic variants, VDR-Fok I rs2228570, GC rs2282679, GC rs4588 (T436K) and GC rs7041 (D432E), were genotyped using the extracted leucocytic DNA samples. The genetic polymorphisms were common variants and have previously shown significant associations with serum 25(OH)D concentrations (Cooper et al., 2011; Engelman et al., 2008; Kurylowicz et al., 2006; Orton et al., 2008). Experiments were carried out using TaqMan single nucleotide polymorphism (SNP) genotyping assay, according to the manufacturer's protocol (Applied Biosystems Inc., Foster City, CA, USA). Genotyping by allelic discrimination was validated and verified by Sanger sequencing. Restriction Fragment Length Polymorphism (RFLP) was performed to determine the rs7041-rs4588 haplotypes and expressed by different DBP isoforms. Polymerase chain reaction (PCR) was performed using primers pairs (forward primer: 5′-CTGGACTTCCAATTCAGCAG-3′; reverse primer: 5′-AATGGCATCTCAATAACAGG-3′) to amplify the target genomic region including rs7041 and rs4588 at the conditions 95°C initiations for 10 min, 50 cycles with 94°C denaturation for 30 s, primers annealing at 60°C 1 min and extension was allowed at 72°C 1 min. The amplification product was subsequently double digested by 1 U StyI and HaeIII restriction enzyme at 37°C for 1 h (New England Biolabs Inc., Ipswich, MA, USA). The DBP protein isoforms were analysed by the band size of the digested product after 2% agarose gel electrophoresis and a representative result is illustrated in Figure S1. Different sizes of DNA fragments were produced if the specific alleles in rs7041 and rs4588 were presented in the PCR products, which generated the restriction sites recognised by HaeIII and StyI; HaeIII recognised the variant rs7041G alleles in GC1/S and generated a fragment with size 273 bp + 194 bp, whereas StyI recognised the variant rs4588A alleles in GC2 and generated a fragment with size 281 bp + 187 bp. Since there are no variant rs7041 and rs4588 alleles in GC1/F, the product was undigested with a size of 457 bp.
Data analysis
Descriptive statistics were computed to summarise the demographic characteristics of the mother–child dyads and the minor allele frequency (MAF) of each selected genetic variant in the study sample. Data were presented as mean (standard deviation [SD]) for continuous variables, median (interquartile range) for skewed variables and frequency (percentage) for categorical variables. The Hardy–Weinberg Equilibrium (HWE) was examined by the Chi-Square test. A series of multiple linear regression models were employed to examine the relationship between various genetic variants (VDR-FokI, rs2282679, rs4588 and rs7041) and serum 25(OH)D concentration among infants per added copy of each minor allele. Similar approaches were used to analyse the effect of carrying any minor allele on serum 25(OH)D concentration. In addition, multiple linear regression models were also conducted to study the association between serum 25(OH)D concentration and (i) distinctive protein phenotypes (GC1F [rs7041T/rs4588C], GC1S [rs7041G/rs4588C] and GC2 [rs7041T/rs4588A]) and (ii) the rs7041-rs4588 combined haplotypes (GC1F/F, GC1F/S, GC1F/2, GC1S/S, GC1S/2 and GC2/2). Subgroup analyses were then conducted to examine the association between the GC rs7041-rs4588 combined haplotype and serum vitamin D concentration in infants and by the genotypes of VDR-FokI (TT/TC vs. CC). All the regression models were adjusted for the child's sex, age, breastfeeding history and supplementation practice and mean UV index at recruitment. All statistical tests were two-sided, with a p-value of less than 0.05 denoting statistical significance, using SPSS Statistics for Windows software (version 26.0, SPSS Inc., Chicago, IL, USA).
RESULTS
Participants' demographics
This study included 378 mother–child dyads. The average maternal age was 33.1 years (SD = 4.2), whereas the average infant's age was 6.4 months (SD = 2.8). About half of the infants were boys (56.1%) (Table 1). The average height z-score and weight z-score of the infants were 0.04 and 0.06 respectively. Among 378 infants, 275 (72.8%) were vitamin D sufficient (total 25[OH]D concentration ≥50 nmol/L), 67 (17.7%) vitamin D insufficient (total 25[OH]D concentration <50 nmol/L) and 36 (9.5%) vitamin D deficient (total 25[OH]D concentration <25 nmol/L). Participants were recruited across different months throughout the year, with the most participants recruited from September to November (33.6%). Our survey found that only around 10% of the mothers provided vitamin D supplementation to their infant child. About half of the infants were breastfed in the previous 7 days (53.4%).
| Mean (SD) | n (%) | |
|---|---|---|
| Mother characteristics | ||
| Age, years | 33.1 (4.2) | |
| Parity | ||
| 1 | 216 (59.2%) | |
| 2 | 123 (33.7%) | |
| 3 or more | 21 (5.8%) | |
| Child characteristics | ||
| Sex | ||
| Boys | 212 (56.1%) | |
| Girls | 166 (43.9%) | |
| Age, months | 6.4 (2.8) | |
| Serum vitamin D concentration, nmol/L | 65.7 (28.3) | |
| Vitamin D status | ||
| Sufficient (≥50 nmol/L) | 275 (72.8%) | |
| Insufficient (<50 nmol/L) | 67 (17.7%) | |
| Deficient: (<25 nmol/L) | 36 (9.5%) | |
| Height, z-score | 0.04 (1.1) | |
| Weight, z-score | 0.06 (1.0) | |
| Birth weight, kg | 3.1 (0.3) | |
| Gestational week at birth | ||
| 37 weeks | 48 (12.7%) | |
| 38 weeks | 127 (33.6%) | |
| 39 weeks | 104 (27.5%) | |
| 40 weeks | 71 (18.8%) | |
| 41 weeks or more | 28 (7.4%) | |
| Recruitment time | ||
| December to February | 90 (23.8%) | |
| March to May | 70 (18.5%) | |
| June to August | 91 (24.1%) | |
| September to November | 127 (33.6%) | |
| Supplementation practice | ||
| Yes | 45 (11.9%) | |
| No | 333 (88.1%) | |
| Breastfed in past 7 days | ||
| Yes | 202 (53.4%) | |
| No | 176 (46.6%) | |
Genetic prevalence of the selected variants in and its association with serum 25(OH)D concentration
Table 2 displays the proportion of infants with specific genotypes of selected GC and VDR variants. The selected polymorphisms, except for rs2228570 (p = 0.019), obeyed the Hardy–Weinberg Equilibrium (HWE) (p > 0.05). Results showed that all three selected GC variants were significantly associated with serum 25(OH)D concentration (p < 0.05) (Table 3 and Figure 1). Compared to rs7041GG (8.2%), infants with rs7041TT were found to have lower 25(OH)D concentrations (β = −13.3, p = 0.007). For rs2282679, it was found that C carrier had a lower 25(OH)D concentration than those with rs2282679AA (p < 0.001). Specifically, infants with rs2282679AC (β = −9.2, p < 0.001) and rs2282679CC (β = −13.5, p = 0.011) had significantly lower 25(OH)D concentrations than those with rs2282679AA. For rs4588, rs4588A carriers were found to have lower 25(OH)D concentrations than those with rs4588CC (p = 0.018). No significant differences were detected in 25(OH)D concentrations among infants with different genotypes of rs2228570.
| Genetic polymorphism | n | % | Hardy–Weinberg equilibrium | ||
|---|---|---|---|---|---|
| χ 2 | p | ||||
| rs7041 | |||||
| Alleles | T | 70.6 | 0.414 | 0.812 | |
| G | 29.4 | ||||
| Genotypes | TT | 186 | 49.2 | ||
| TG | 162 | 42.9 | |||
| GG | 30 | 8.2 | |||
| rs4588 | |||||
| Alleles | A | 24.4 | 0.0313 | 0.984 | |
| C | 75.6 | ||||
| Genotypes | AA | 22 | 5.8 | ||
| AC | 141 | 37.3 | |||
| CC | 215 | 56.9 | |||
| rs2282679 | |||||
| Alleles | C | 27.2 | 0.287 | 0.866 | |
| A | 72.8 | ||||
| Genotypes | CC | 26 | 6.9 | ||
| AC | 154 | 40.7 | |||
| AA | 198 | 52.4 | |||
| rs2228570 (FokI) | |||||
| Alleles | T | 32.1 | 7.936 | 0.019 | |
| C | 67.9 | ||||
| Genotypes | TT | 51 | 13.5 | ||
| TC | 141 | 37.3 | |||
| CC | 186 | 49.2 | |||
| Crude | Adjusteda | |||
|---|---|---|---|---|
| β (95% CI) | p | β (95% CI) | p | |
| rs7041 | ||||
| TT | −14.3 (−25.2, −3.4) | 0.010 | −13.3 (−23.1, −3.6) | 0.007 |
| TG | −8.9 (−19.9, 2.1) | 0.115 | −7.7 (−17.5, 2.2) | 0.127 |
| TT/TG | −11.8 (−22.4, −1.1) | 0.030 | −10.7 (−20.3, −1.1) | 0.029 |
| GG | 1.0 | – | 1.0 | – |
| rs4588 | ||||
| AA | −7.6 (−20.0, 4.9) | 0.235 | −10.8 (−22.1, 0.5) | 0.062 |
| AC | −5.4 (−11.5, 0.6) | 0.078 | −5.6 (−11.0, −0.2) | 0.040 |
| AA/AC | −5.7 (−11.6, 0.1) | 0.055 | −6.3 (−11.6, −1.1) | 0.018 |
| CC | 1.0 | – | 1.0 | – |
| rs2282679 | ||||
| CC | −12.3 (−23.9, −1.0) | 0.034 | −13.5 (−23.9, −3.1) | 0.011 |
| AC | −10.2 (−16.1, −4.3) | <0.001 | −9.2 (−14.5, −3.9) | <0.001 |
| CC/AC | −10.5 (−16.2, −4.8) | <0.001 | −9.8 (−15.0, −4.7) | <0.001 |
| AA | 1.0 | – | 1.0 | – |
| rs2228570 (FokI) | ||||
| TT | 0.9 (−7.9, 9.7) | 0.844 | 2.8 (−5.4, 10.8) | 0.505 |
| TC | −2.0 (−8.2, 4.2) | 0.528 | 0.4 (−5.2, 6.0) | 0.887 |
| TT/TC | 1.2 (−4.5, 7.0) | 0.674 | 1.0–0.7 (−4.3, 6.3) | 0.710 |
| CC | 1.0 | – | 1.0 | – |
- a Adjusted for sex, age, breastfeeding history and supplementation practice of the infants and mean UV index at recruitment.
Association between the combined genetic effects of vitamin D-binding protein and vitamin D receptor variants and infants' serum 25(OH)D concentration
As shown in Table 4, our RFLP results found that about 70% of infants were GC1F (rs7041T/rs4588C) carriers; about half of them were GC1S (rs7041G/rs4588C) carriers, and 43.1% were GC2 (rs7041T/rs4588A) carriers. When we considered the rs7041-rs4588 combined haplotype, the prevalence of GC1F/F, GC1F/S, GC1F/2, GC1S/S, GC1S/2 and GC2/2 were 22.8%, 26.2%, 20.9%, 7.9%, 16.4% and 5.8%, respectively.
| n (%) | Crude | Adjusteda | |||
|---|---|---|---|---|---|
| β (95% CI) | p | β (95% CI) | p | ||
| Alleles | |||||
| GC1F (rs7041T/rs4588C) | 264 (69.8%) | −3.1 (−9.5, 3.1) | 0.323 | −2.0 (−7.8, 3.7) | 0.488 |
| GC1S (rs7041G/rs4588C) | 191 (50.5%) | 6.8 (1.0, 12.5) | 0.022 | 6.9 (1.7, 12.0) | 0.010 |
| GC2 (rs7041T/rs4588A) | 163 (43.1%) | −5.1 (−10.9, 0.7) | 0.086 | −6.3 (−11.6. -1.1) | 0.018 |
| Genotypes | |||||
| GC1F/F | 86 (22.8%) | −12.3 (−24.1, −0.6) | 0.039 | −10.1 (−20.6, 0.4) | 0.059 |
| GC1F/S | 99 (26.2%) | −7.5 (−19.0, 4.0) | 0.202 | −6.6 (−16.9, 3.7) | 0.209 |
| GC1F/2 | 79 (20.9%) | −15.8 (−27.7, −4.0) | 0.009 | −15.5 (−26.1, −4.9) | 0.004 |
| GC1S/S | 30 (7.9%) | 1.0 | – | 1.0 | – |
| GC1S/2 | 62 (16.4%) | −11.2 (−23.5, 1.1) | 0.074 | −9.2 (−20.1, 1.8) | 0.101 |
| GC2/2 | 22 (5.8%) | −15.9 (−31.5, −0.4) | 0.044 | −17.8 (−31.9, −3.9) | 0.012 |
- a Adjusted for sex, age, breastfeeding history and supplementation practice of the infants and mean UV index at recruitment.
Our regression analyses showed that GC1S-carrying infants had significantly higher serum 25(OH)D concentrations than infants without the GC1S variant (β = 6.9, p = 0.010). Moreover, when analysing the association between DBP genotypes and serum 25(OH)D concentration, the adjusted regression model found that GC1F/2 (β = −15.5, p = 0.004) and GC2/2 (β = −17.8, p = 0.012) carriers had lower serum 25(OH)D concentrations than GC1S/S carriers (Table 4). In addition, the associations between DBP variants and serum 25(OH)D concentration were found to vary by the VDR-FokI genotype (Table 5). Specifically, significant associations were observed in the DBP genotypes GC1F/F, GC1F/2 and GC1S/2 among VDR-FokI TT/TC carriers, but not in VDR-FokI CC carriers. The lowest mean serum 25(OH)D concentration was observed in the GC1F/2-VDR-FokI T carriers (β = −21.7, p = 0.007).
| DBP genotypes | VDR-FokI | |||||
|---|---|---|---|---|---|---|
| TT/TC (n = 192) | CC (n = 186) | |||||
| n (%) | β (95% CI) | p | n (%) | β (95% CI) | p | |
| GC1F/F | 47 (24.5%) | −19.7 (−35.1, −4.4) | 0.012 | 39 (21.0%) | 0.6 (−13.5, 14.8) | 0.928 |
| GC1F/S | 52 (27.1%) | −13.1 (−28.3, 2.0) | 0.090 | 47 (25.3%) | −0.2 (−13.8, 13.4) | 0.980 |
| GC1F/2 | 33 (17.2%) | −21.7 (−37.7, −5.9) | 0.007 | 46 (24.7%) | −10.2 (−23.9, 3.6) | 0.147 |
| GC1S/S | 14 (7.3%) | 1.0 | – | 16 (8.6%) | 1.0 | – |
| GC1S/2 | 36 (18.8%) | −21.6 (−37.3, −5.9) | 0.007 | 26 (14.0%) | 3.6 (−11.4, 18.6) | 0.636 |
| GC2/2 | 10 (5.2%) | −19.1 (−397, 1.5) | 0.069 | 12 (6.5%) | −10.9 (−30.1, 8.4) | 0.269 |
- Note: Adjusted for sex, age, breastfeeding history and supplementation practice of the infants and mean UV index at recruitment.
DISCUSSION
This study provides evidence for the influence of genetic factors on serum 25(OH)D concentrations in Chinese infants aged 2–12 months. Among the four studied variants, we found that all three selected DBP variants (GC rs2282679, GC rs4588 and GC rs7041) were significantly associated with serum 25(OH)D concentrations. Furthermore, our moderation analyses revealed that the associations between DBP variants and serum 25(OH)D concentrations of the infants could differ by their genotype of VDR-FokI.
Consistent with previous studies (McGrath et al., 2010; Zhou et al., 2019), we found a negative association between the common GC variants and infant serum 25(OH)D concentration. Carriers of the rs7041 homozygous TT variant were identified as the risk population associated with a reduction in serum 25(OH)D concentration among Chinese infants. Similarly, a significant association was also detected between rs2282679 and reduced 25(OH)D concentration. The rs2282679 C carriers were identified to have significantly lower serum 25(OH)D concentrations than the homozygous rs2282679A carriers. When compared to other studies, we found that this association could be stronger among southern Chinese than European populations (Cheung et al., 2013). However, unlike other previous studies (Li et al., 2014; Powe et al., 2013; Sadat-Ali et al., 2016; Zhou et al., 2019), our results showed that the association between rs4588 and serum 25(OH)D concentration was very weak. This could be because previous studies were conducted on populations including various races or ethnicities including Chinese Han (Li et al., 2014; Sadat-Ali et al., 2016; Zhou et al., 2019), but ethnic differences among our study participants were minimal. In addition, previous studies focused mainly on adults, whereas our study focused on infants alone. We postulated that the effect of the rs4588 genotype on serum 25(OH)D concentration could be subjected to environmental influences such as sunlight exposure, and dietary and supplementation practices in which large variations are more likely to be observed among adults than infants. Apart from the DBP variants, our study also found that there is no direct association between VDR-FokI and 25(OH)D concentration, which has also been reported in previous studies on Jordanian (Atoum et al., 2015) and Indian populations (Bhanushali et al., 2009). These findings suggest that the influences of VDR-FokI on infant serum 25(OH)D concentration may be indirect through changes in other unmeasured factors.
Our analyses on different DBP isoforms identified GC1F, or 432D allele, as the risk protein variant which was also found to have a relatively high prevalence in Chinese infants. The haplotype distribution in this study was similar to those reported in African American populations, but the major haplotype in Caucasian populations was reported to be GC1S (Newton et al., 2019). Furthermore, our study observed that infants with GC1S were more likely to have higher serum 25(OH)D concentration. A similar observation was also found in African American children (Newton et al., 2019). A previous study found that the GC2 variant was associated with lower vitamin D3 concentrations among Canadian adults of East Asian and European ancestry, but such an association was not observed in South Asians (Engelman et al., 2010; Gozdzik et al., 2011; Powe et al., 2013). However, our results identified a significant negative association between 25(OH)D concentrations and the GC2 variant. Further analyses showed that the isoform GC1F/2 was significantly associated with a lower 25(OH)D concentration. Previous in vivo studies have suggested circulating DBP acts as an important reservoir for vitamin D metabolites, which can help minimise the risk of vitamin D deficiency (Nykjaer et al., 1999; Safadi et al., 1999; Zella et al., 2008). However, the functional consequences of DBP variants, particularly in terms of infant serum 25(OH)D concentrations, remain largely unclear. Scientific evidence regarding the effect of GC and VDR genotypes in affecting serum 25(OH)D among infants is limited. Some knowledge gaps, including the infant-specific synthetic rate of the DBP, protein concentration and half-life in the circulation, remain unanswered. Our findings can provide initial insights into the infant-specific associations between DBP variants and serum 25(OH)D concentration in the Han Chinese population.
In addition, this study is among the first to use a genetic approach to examine the potential interaction between two important elements of the vitamin D metabolic pathway, GC and VDR, on 25(OH)D concentrations in healthy infants. While our results revealed that there is no direct association between VDR-FokI genotype and serum 25(OH)D concentration, a VDR-FokI genotype-specific pattern was observed in the association between DBP isoforms and serum 25(OH)D concentration. Specifically, we found that the effects of DBP isoforms (GC1F/F, GC1F/2 and GC1S/2) on serum 25(OH)D concentrations were significantly stronger among VDR-FokI T carriers than non-carriers. This indicates the potential digenic GC-VDR interactions in the determination of the serum 25(OH)D concentration among infants in Hong Kong. Both GC and VDR encode for essential elements in the vitamin D metabolism with distinctive functions (McGrath et al., 2010). While DBP is primarily responsible for the transportation of vitamin D in the circulation, the vitamin D receptor plays a key role in regulating cytochrome P450 enzymes located on mitochondria. Our findings suggest that vitamin D metabolism (such as synthesis and activation) may be less effective among the GC1F/F-VDR-FokI T infant carriers. For example, the expression and activity of vitamin D enzymes such as CYP27B1 and CYP24A1 could be affected in GC1F-FokI T carriers. However, GC1F DBP was frequently reported to have a higher affinity for serum 25(OH)D and therefore is potentially more efficient in delivering 25(OH)D to various target organs. (Abbas et al., 2008; Arnaud & Constans, 1993; Lauridsen et al., 2005). Furthermore, the VDR-FokI genotype-specific association between GC isoforms and serum 25(OH)D could be due to the difference in metabolising the substrates in the pathway, with the involvement of the altered VDR translation initiation site encoded by the VDR-FokI polymorphisms. VDR-FokI may encode two potential translation start sites: the M1 form encoded by T alleles at the first start codon is three amino acids longer than the M4 form encoded by the C allele (Whitfield et al., 2001) The M4 protein form was found to have higher transcriptional activity than the M1 form, possibly because of the higher affinity to the Transcription Factor II B (TFIIB) binding to the VDR-RXR complex (Whitfield et al., 2001). Moreover, the C allele was also associated with a higher copy number of VDR mRNA. (Ogunkolade et al., 2002) Thus, the T allele carriers would express fewer cellular VDR receptors in accommodating the more quickly produced active ligands. Future functional and repetitive studies are needed to confirm our hypothesis and to elucidate the exact mechanisms underlying the responses of serum 25(OH)D concentrations towards different SNPs and associated genes identified in this study, especially in infants.
This study has several limitations. First, our study sample was relatively small and may not have adequate power to detect the effect of interactions between the selected variants. Second, because of potential racial and age effects, the results may not be generalisable to other populations and age groups. Third, we did not examine the vitamin D status and genotype of the mother of the infants, which could be a potential factor affecting the vitamin D insufficiency risk of the infants. Further research would benefit from assessing the maternal influences on the vitamin D insufficiency risk in infants. Last, rs2228570 did not obey the HWE in our study, possibly due to our relatively small sample size. Further studies with a larger sample size could provide more solid evidence of the association between DBP and VDR genetic variants and serum 25(OH)D concentrations among infants. Despite these limitations, this study provides initial evidence of the genetic influences on the serum 25(OH)D concentrations of Hong Kong infants. GC1/F subtype, which is encoded by rs7041T, was found to be associated with lower serum 25(OH)D concentrations. In addition, we observed a significant interactive effect of DBP-VDR FokI on serum 25(OH)D concentrations in Chinese infants, particularly for those GC1F/F-VDR-FokI T carriers. Further studies are needed to support the development of genetic screening in identifying infants susceptible to hypovitaminosis D in this community.
ACKNOWLEDGEMENTS
The authors thank the support from the Family Health Service of the Department of Health, Hong Kong SAR, as well as all participants in this study.
FUNDING INFORMATION
This work was supported by Health and Medical Research Fund (HMRF) (Reference No.: Vit D-HKU & 18192311). The funding sources were not involved in the study design, data collection, analysis and interpretation; writing of the manuscripts; and the decision to submit the manuscript for publication.
CONFLICT OF INTEREST STATEMENT
The authors declare that they have no competing interests.
Open Research
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.
