The Significance of VDR Genetic Polymorphisms in the Etiology of Preeclampsia in Pregnant Polish Women

For the first time in the Polish population, we aimed to investigate associations between the VDR gene single-nucleotide polymorphisms (SNPs) BsmI (rs15444410), ApaI (rs7975232), FokI (rs19735810), and TaqI (rs731236) and the development of preeclampsia (PE). A case–control study surveyed 122 preeclamptic and 184 normotensive pregnant women. The polymerase chain reaction–restriction fragment length polymorphism (PCR-RFLP) method was performed to examine the maternal VDR FokI, BsmI, TaqI, and ApaI polymorphisms. The VDR BsmI AA homozygous genotype was statistically significantly more frequent in preeclamptic women compared to the control group (p = 0.0263), which was also associated with a 2-fold increased risk of PE (OR = 2.06, p = 0.012). A correlation between the VDR BsmI polymorphism with systolic and diastolic blood hypertension was noted. Furthermore, 3-marker haplotype CTA (TaqI/ApaI/BsmI) was associated with significantly higher systolic (p = 0.0075) and diastolic (p = 0.0072) blood pressure. Association and haplotype analysis indicated that the VDR BsmI A allele could play a significant role in the PE pathomechanism and hence could be a risk factor for PE development in pregnant Polish women. These results indicate the importance of the VDR BsmI polymorphism and reveal that this variant is closely associated with a higher predisposition to hypertension.


Introduction
Preeclampsia (PE) (ICD-10 code 014), a specific condition during pregnancy, still remains a significant clinical problem in perinatal medicine. Hypertensive disorders occur in about 5-10% of pregnancies throughout the world, and preeclampsia complicates the course of 3% of pregnancies in industrialized countries [1]. In Poland, as in other regions of the world, the frequency of this serious complication is estimated at an incidence rate of 2-3% [2,3]. Preeclampsia is connected to a significant risk of the appearance of serious complications in mothers and fetuses, which may require labor induction or a preterm caesarean section. Additionally, preeclampsia could be associated with an increased risk of several disorders in mothers later in life, such as cardiovascular diseases, chronic arterial hypertension, stroke, metabolic syndrome, and chronic kidney disease. Infants born to women with preeclampsia are at increased risk of, first of all, cardiovascular diseases and hypertension, but also insulin resistance, diabetes, and even neurodevelopmental disorders [4][5][6]. Thus, considering the above-mentioned, it undoubtedly becomes clear the activity of antioxidant enzymes such as superoxide dismutase, and neutralizes the activity of NADPH, an enzyme involved in the formation of free oxygen radicals [30].
According to Barrera [31], VDR mediates the effect of calcitriol, which dose-dependently inhibited IL-10 expression in normal and preeclamptic pregnancies. Xu et al. [32] showed that downregulation of VDR expression contributes to increased endothelial inflammatory response in preeclampsia and is associated with upregulation of the expression of VCAM-1 in systemic vessel endothelium in this disease. Xu et al. [32] also revealed that VDR expression has been significantly reduced in cells treated with TNFα. Moreover, it was shown that upregulation of miR-126 expression induced by 1,25(OH) 2 D 3 is mediated through VDR. Thus, this upregulation of miR-126-3p expression may be involved in a mechanism of the anti-inflammatory activity of vitamin D.
The active 1,25(OH)2D form regulates blood pressure by influencing endothelial smooth muscle proliferation [33]. Vitamin D also affects the functioning of the reninangiotensin-aldosterone system (RAAS) by inhibiting the transcription of renin, a proteolytic enzyme that raises blood pressure. Mice experimentally lacking the VDR gene developed arterial hypertension independent of classical regulatory mechanisms such as plasma volume and osmolality [34]. Recent studies have shown that impaired placement causes disturbed expression of the placental renin-angiotensin system, which may lead to systemic disturbances in the functioning of the RAAS system and the development of PE [30]. Vitamin D also modulates the functioning of the immune system, affecting the balance of Th1/Th2 helper lymphocytes, and in this way, it reduces the secretion of pro-inflammatory cytokines and increases the secretion of anti-inflammatory cytokines [35]. This leads to a state of transient immunosuppression necessary for the proper development of pregnancy at an early stage. Given the variety of functions, it is no wonder that disturbances in vitamin D homeostasis lead to many disorders, and vitamin D deficiency is also listed among the risk factors for the development of preeclampsia. It has been shown that pregnant women whose vitamin 25 (OH)D level falls below 20 ng/mL have an increased risk of developing preeclampsia [36].
VDR is a transcription factor that, when bound to the retinoid X receptor, binds to vitamin D responsive elements (VDREs) in the DNA promoter sequence of many genes. Vitamin D acts on cells after binding to a specific steroidal vitamin D receptor (VDR). In normal pregnancy, VDR expression was demonstrated in trophoblasts and temporal villi, in the smooth muscles of the placental vessels, in the nuclei of the fetal villi stromal cells, and in the nuclei of fetal epithelial cells [8]. The VDR gene, cloned for the first time by Baker et al. [37], is located on chromosome 12 (12q11-q13). The most intensively studied VDR polymorphisms are FokI, TaqI, BsmI, and ApaI variants. Three single nucleotide polymorphisms (ApaI, BsmI in intron 8 section, and TaqI in exon 9 (Ile352Ile) are in strong linkage disequilibrium (LD) with each other and additionally with a poly(A) length polymorphism (rs17878969) in the 3 untranslated region of the VDR gene. These variants do not change the amino acid sequence in the VDR protein but can influence the gene expression through regulation of mRNA stability [38]. The most frequently studied variant that results in altered structure of the VDR protein is the FokI variant causing a change of C>T nucleotides (ACG-ATG) in exon 2. In carriers of the C allele (designated F) the VDR protein is three amino acids shorter (424 amino acids) and more active. In contrast, individuals carrying the T allele (designated f ) synthesize a less active, full-length VDR protein (427 amino acids) [39].
Previously, no relationship between PE susceptibility and the maternal and placental VDR BsmI rs1544410, TaqI rs731236, and ApaI rs7975232 polymorphisms has been observed [40]. However, it was revealed that the maternal and placental VDR FokI polymorphism has been associated with lower PE risk in the dominant model (Ff + ff vs. FF) and it was concluded that these genotypes could decrease PE risk [40].
In addition to the above, it is worth noting that the results of the meta-analysis of 11 observational studies showed the relationship between the VDR polymorphism (FokIseven, BsmI-eight, ApaI-five, and TaqI-seven studies) and the risk of preterm birth, and it was shown mainly the protective effect of the BsmI polymorphism against the risk of preterm birth in terms of the allele (A vs. G: OR = 0.74; 95% CI 0.59-0.93) and in the recessive model (AA vs. GG + AG: OR = 0.62; 95% CI 0.43-0.89) [41].
Considering the important role of vitamin D and its receptor in the development of physiological pregnancy and the possible influence of vitamin D deficiency on the occurrence of preeclampsia, it can be assumed that VDR gene polymorphisms influence the development of preeclampsia. In the present study, the four vitamin D receptor gene SNPs FokI (rs2228570), BsmI (rs1544410), ApaI (rs7975232), and TaqI (rs731236) were analyzed to assess their potential value for risk prediction in preeclamptic women.

Subjects
A total of 306 Caucasian women of Polish origin (122 with preeclampsia and 184 heathy normotensive pregnant women) were enrolled in the study. All patients were recruited from the Division of Perinatology and Women's Diseases, Poznan University of Medical Sciences. Preeclampsia was diagnosed according to American College of Obstetricians and Gynecologists (ACOG) criteria [42]. The exclusion criteria were as follows: age younger than 19 or older than 35, diabetes mellitus, hemolytic disease, habitual smokers, multifetal pregnancy, chronic hypertension, endocrinological diseases, kidney diseases, or obesity. Depending on the time of disorder appearance, preeclampsia is classified either as early onset (EOPE), which requires delivery before 34 weeks' gestation, or late-onset (LOPE), with delivery at or after 34 weeks.
In each woman with preeclampsia, venous blood was taken for laboratory tests leading to preeclampsia diagnosis routinely after their admission to hospital. The laboratory tests such as urea, uremic acid, total protein, creatinine, liver enzymes ALT, AST, and maternal proteinuria were performed in III trimester of gestation. In the laboratory analysis of the above tests, the protocol methods used routinely in our hospital were applied.
The participants in the control group were in the third trimester of a healthy pregnancy and had to be normotensive, without any fetal disorder, pathological states, or multiple pregnancies. All patients were Caucasians of Polish origin. The enrolled women were informed in detail about the study and provided their written consent to collect venous blood samples for diagnostic and scientific purposes including genotyping. The study was approved by the Local Bioethical Committee at Poznan University of Medical Sciences (No: 1129/18).

Genotyping
The preliminary data on the selected VDR gene SNPs are shown in Table 1. Genomic DNA was extracted from blood samples stored in EDTA-coated tubes using the QIAamp DNA Mini Kit according to the manufacturer's instructions (Qiagen, Hilden Germany). The concentration of genomic DNA was quantified using a NanoDrop Spectrophotometer (Thermo Scientific, Waltham, MA, USA). Genotyping of the VDR gene polymorphisms was performed in the Molecular Biology Laboratory of Poznan Medical Science University by polymerase chain reaction and restriction fragment length polymorphism (PCR-RFLP). The rs7975232 and rs731236 polymorphic sites of the VDR gene were analyzed by a method described in the study of Pani et al. [43]. A 740-bp PCR product was digested with ApaI (Eurx, Gdańsk, Poland) at 25 • C overnight. The same PCR product was also digested overnight with the restriction enzyme TaqI (Eurx, Gdańsk, Poland) at 65 • C. The 825-bp fragment encompassing the rs1544410 polymorphic site in intron 8 was amplified using primers and PCR conditions described by Morrison et al. (1994) [44]. Mva12691 (BsmI) (Thermo Scientific, Waltham, MA, USA) enzymes were added directly to the PCR products and digested at 37 • C. The rs2228570 was identified using the method described by Harris et al. [45]. The products of PCR amplification were cleaved with the restriction enzyme FokI (Eurx, Gdańsk, Poland) at 37 • C overnight. DNA fragments obtained after respective restrictive enzyme digestion and the DNA size marker were electrophoresed on a 2.5% agarose gel with Midori Green Advanced DNA Stain (Nippon Genetics, Europe GmbH, Düren, Germany). For documentation of results, gel pictures were taken under ultraviolet light.

Statistical Analysis
Statistical analyses were conducted using R version 4.0.3 version (R Foundation for Statistical Computing, Vienna, Austria, accessed on 20 October 2020) [46]. For continuous variables, normality was checked by the Shapiro-Wilk test. Continuous variables normally distributed were expressed as mean ± SD and in the absence of normal distribution as median and interquartile range (IQR). Categorical variables were expressed as numbers or percentages. Clinical characteristics between groups were compared using Student's t-test or Mann-Whitney U test. Pearson's χ 2 and the Fischer test for nominal variables was used. For each SNP, the Hardy-Weinberg equilibrium (HWE) was assessed using Pearson's goodness-of-fit χ 2 statistic. Differences in allele and genotype frequencies between the case and control subjects, odds ratios (ORs) and associated 95% confidence intervals (95% CIs) were evaluated using the SNPassoc package for R [47]. Two-tailed values of p < 0.05 were accepted to be statistically significant. The Bonferroni multiple-comparison correction method was employed to calculate the corrected p value. The general linear model (GLM) was applied to compare blood pressure levels between genotypes and haplotypes.
The haplotype-based association analysis was performed using Haploview version 4.2 software ((Cambridge, MA, USA, https://www.broadinstitute.org/haploview/haploview (accessed on 1 October 2020)). Linkage disequilibrium (LD) parameters r 2 and D and haplotype frequencies were calculated. Distribution of haplotypes was compared in case and control groups with chi-squared tests. Significant p values were corrected using the 10,000-fold permutation test.

Baseline Demographics and Clinical Characteristics of Study Population
The preeclamptic women and controls had similar mean age (30.12 ± 5.52 vs. 30.58 ± 4.41 years, respectively, p = 0.4507). In the PE group lower gestational age, neonatal birth weight and Apgar score in the first and fifth minute, as well as higher systolic/diastolic blood pressure and pre-/post-pregnancy BMI index compared to the control group (all, p < 0.001), were noted. Accordingly, pre-pregnancy BMI and primiparity were selected as a covariate to be controlled for in subsequent analysis. In preeclamptic women delivery by caesarean section was more frequent (90.16 vs. 33.15% in controls, p < 0.001), and more women in this group were primiparas (61.48 vs. 39.13% in controls, p = 0.0001). Clinical characteristics of the study population are summarized in Table 2.
In the PE group early onset preeclampsia was diagnosed in 57 women (46.72%) and 65 (53.28%) cases were diagnosed as late-onset. Comparing early and late-onset groups, the most interesting observation was lower gestational age in the early onset subgroup (30.60 vs. 36.54 weeks, p < 0.001) and systolic blood pressure (175.65 vs. 166.97 mmHg, p = 0.0171). The two subgroups also differed in neonatal birthweight and Apgar score in the first and fifth minute (all, p < 0.001). For the PE group also blood chemistry tests were performed. In the early onset group a statistically significantly lower level of total protein (5.47 vs. 5.81 g/dL in LOPE, p = 0.0089) and significantly higher levels of ALT (53.83 vs. 23.08 in LOPE, p = 0.0411) and AST (52.17 vs. 25.50 in LOPE, p = 0.0492) were observed. The distribution of selected clinical and laboratory parameters in PE cases and controls is shown in Table 3.

Associations between SNPs and Preeclampsia Risk
The genotyping completion rates were 100% and all SNPs conformed to HWE in the control group (p > 0.05). The genotype distribution of BsmI rs1544410 differs significantly from that expected by HWE only in the preeclamptic group (p = 0.021). Higher minor allele frequency (MAF) for BsmI rs1544410 was found in PE cases (0.467 vs. 0.353 in controls, OR = 1.61, 95% CI = 1.15-2.23, p = 0.005) and this association remained significant after applying correction for multiple testing (p = 0.020). Additionally, for TaqI rs731236, the C allele was associated with increased risk of PE (OR = 1.42, 95% CI: 1.02-1.99, p = 0.038), but the association was not significant after Bonferroni correction (Table 4). The risk of PE associated with a specific VDR genotype was confirmed by testing the effect of pre-pregnancy BMI and primiparity as confounding factors. Significantly higher frequencies of homozygous BsmI AA genotype carriers were found between PE cases and control women, respectively 27.0 vs. 15.2%, p = 0.0263 (recessive model: OR = 0.48, 95% CI = 0.27-0.85, p = 0.0119). This association remained significant after adjusting for maternal BMI (p adj. = 0.0113, recessive model p adj. = 0.0048).
The dominant models for rs1544410 (BsmI) and rs7975232 (ApaI) were significantly associated with PE after adjusting (p adj. = 0.0334 and p adj. = 0.0425, respectively). The prevalence of the maternal VDR rs731236 (TaqI) TC and CC genotype were higher in preeclamptic group than the healthy pregnant women. When the mode of inheritance was dominant, the OR (95% CI) for TT vs. TC + CC was 0.63 (0.39-1.00, p = 0.0498, p adj. = 0.0375). The distribution of VDR genotypes in PE women and controls is summarized in Table 5. The prevalence of the maternal VDR rs7975232 GG genotype was higher in the healthy pregnant women than the preeclamptic group (25.5 vs. 18.9%). When the mode of inheritance was dominant, the OR (95% CI) for TT vs. TG + GG was 1.63 (0.97-2.74, p = 0.0660, p adj. = 0.0503). In the next stage a stratified analysis was performed. Women with preeclampsia were divided into two subgroups: early onset preeclampsia (57 cases), and late-onset preeclampsia (65 cases). Relative to the controls, significantly increased risk of EOPE was found to be associated with the TT genotype of ApaI (rs7975232) polymorphism in a codominant model, compared with GG genotype (p= 0.0724, OR = 2.85, 95%CI = 1.08-7.57, p adj. = 0.036). Compared with GG genotype, markedly increased risk of preeclampsia was associated with the TT + TG genotypes in a recessive model (p = 0.0275, OR = 2.45, 95% CI = 1.04-5.78, p adj. = 0.0185). No significant association was observed between ApaI polymorphism and risk of late-onset PE (p > 0.05). Significantly increased risk of early onset preeclampsia was also identified to be associated with the A allele of the BsmI (rs1544410) polymorphism (p = 0.0451 in the dominant model and p adj. = 0.0305 in the recessive model). For this variant a difference was also found between late-onset preeclampsia and controls in the recessive model (p = 0.0165, OR = 0.43, 95% CI = 0.22-0.85). For the FokI (rs2228570) polymorphism, the only statistically significant difference was observed between the EOPE and control groups in the recessive model after adjusting for BMI (p adj. = 0.0402). Our data indicated no significant difference in the allele frequencies of TaqI (rs731236) and FokI (rs2228570) polymorphisms between early and late-onset preeclampsia compared to the control group (Table 6).

Association between VDR Gene Variants and Blood Pressure
Linear regression analysis was used to examine genotype associations with blood pressure for all 306 women (Table 7). One SNP, rs1544410 (BsmI), was found to be associated with both systolic blood pressure (SBP) and diastolic blood pressure (DBP) in all genetic models evaluated. This association remained significant after adjusting for maternal BMI and primiparity (SBP p-value 0.0033, 0.0172 and 0.0015, DBP p-value 0.0026, 0.0073 and 0.0020 in the codominant, dominant, and recessive model, respectively). Carriers of the A allele in rs1544410 showed higher blood pressure in both SBP and DBP than non-carriers (GG genotype). The minor allele C of TaqI SNP (rs73123) increased the SBP and DBP of women who carried them in the dominant model, and the minor allele G of ApaI SNP (rs7975232) had the opposite effect on SBP and DBP.
When the examined women were divided into subgroups, statistically significant differences we observed for diastolic blood pressure in the codominant model and BsmI (p adj. = 0.0431) and TaqI (p adj. = 0.0469) variants in the controls. In the group of early onset preeclampsia, the differences were noted for TaqI in the recessive (for SBP p adj. = 0.0301 and for DBP p adj. = 0.0033) and the codominant model for DBP (p adj. = 0.0130). For the late onset PE group, the FokI polymorphism and systolic blood pressure significant difference in recessive model was obtained (p adj. = 0.0182).

Linkage Disequilibrium and Haplotype Analysis of VDR SNPs
It was also investigated whether the four SNPs were in linkage disequilibrium. A schematic diagram of the LD pattern among genotyped SNPs is shown in Figure 1 (the range of the area surrounded by lines indicates that three SNPs of the VDR gene were contained in a haplotype and in a state of linkage disequilibrium). Block 1 comprised rs1544410 (BsmI), rs7975232 (ApaI) and rs731236 (TaqI), which were in strong pairwise LD (D' > 0.95; r 2 > 0.5). However, rs2228570 (FokI) was poorly correlated with the other three SNPs. The most common haplotypes of the study polymorphisms, calculated by Haploview 4.2, are summarized in Table 8. Eight four-marker haplotypes with a frequency of more than 1% were detected (TGGC 25.2%, TGGT 22.5%, CTAC 19.0%, CTAT 16.6%, TTGT 7.3%, TTGC 4.4%, TTAT 2.2%, and TTAC 1.6%). In the haplotype analysis, only the TTAT haplotype occurrence was more frequent in the study group compared to the control group (0.036 vs. 0.013, p = 0.0572). The other four-locus haplotypes did not show any difference in frequencies between the cases and controls. For three -marker haplotype analyses, CTA (p = 0.0241) and TAT (p = 0.0509) were more frequent in preeclamptic women, while the TGG haplotype was more common in the uncomplicated pregnancy group (0.510 vs. 0.425, p = 0.0376). In the permutation testing, none of the significant haplotypes was significantly associated with PE.  The most common haplotypes of the study polymorphisms, calculated by Haploview 4.2, are summarized in Table 8. Eight four-marker haplotypes with a frequency of more than 1% were detected (TGGC 25.2%, TGGT 22.5%, CTAC 19.0%, CTAT 16.6%, TTGT 7.3%, TTGC 4.4%, TTAT 2.2%, and TTAC 1.6%). In the haplotype analysis, only the TTAT haplotype occurrence was more frequent in the study group compared to the control group (0.036 vs. 0.013, p = 0.0572). The other four-locus haplotypes did not show any difference in frequencies between the cases and controls. For three -marker haplotype analyses, CTA (p = 0.0241) and TAT (p = 0.0509) were more frequent in preeclamptic women, while the TGG haplotype was more common in the uncomplicated pregnancy group (0.510 vs. 0.425, p = 0.0376). In the permutation testing, none of the significant haplotypes was significantly associated with PE. In a haplotype analysis the 3 -marker haplotype CTA (TaqI/ApaI/BsmI) was associated with a significantly higher systolic (p = 0.0075, 95%CI = 2.36-15.33) and diastolic (p = 0.0072, 95%CI = 1.50-9.56) blood pressure risk as compared with the most frequent haplotype TGG (Table 9). Table 9. Estimated frequency of haplotypes and association with risk of higher blood pressure.

Discussion
Polymorphic variants of vitamin D are involved in many biological processes and in the etiology of several disorders, including cardiovascular diseases, diabetes, osteoporosis, and cancer. These observations were supported by epidemiological studies. The association of genetic VDR polymorphisms with these disorders were indicated by recent studies for BsmI, ApaI, and TaqI nonfunctional variants. These variants probably could be in linkage disequilibrium (LD) with other functional polymorphisms, which explains the results observed recently. The explanation of this relationship may be helpful in defining the group of patients predisposed to development of the disorders as well as in creating tailored therapy [37,48].
It has often been indicated that some VDR gene variants could be associated with an increased risk of chronic hypertension [49,50]. Thus far, several publications have analyzed the relationship of VDR gene polymorphisms with the occurrence of certain complications in pregnancy [27]. It has been suggested that VDR polymorphism could be associated with the occurrence of gestational diabetes [51][52][53], preterm labor [54], recurrent miscarriages [55,56], gestational cholestasis [40], and intrauterine fetal restriction that often coexisted with PE and placental insufficiency [27]. Thus far, however, no unequivocal results have been obtained determining the importance of VDR gene variants in the pathways influencing the placenta function, regulating implantation, hormone secretion, or immune modulation.
According to our best knowledge, there have been no studies evaluating the frequency of VDR gene polymorphisms in the Polish population of pregnant women with preeclampsia that have been published. In our study, the most interesting observation was the demonstration that the VDR BsmI AA homozygous genotype was statistically significantly more frequent in preeclamptic women compared to the control group (p = 0.0263), which was also associated with a 2-fold increased risk of PE (OR = 2.06, p = 0.012). Additionally, the frequency of the VDR BsmI A allele was significantly higher in the PE group compared to controls (p = 0.005). The association between PE and VDR BsmI polymorphism occurrence persisted even after adjustment for the pre-pregnancy BMI value as a predisposing factor for hypertension in pregnancy. Moreover, the stratification analysis, after dividing the PE group into early onset and late-onset forms, showed a relationship between the EOPE form and VDR BsmI A allele carriage in both dominant (p = 0.0451) and recessive (p adj. = 0.0441) models, as well as between the LOPE form and VDR BsmI A allele carriage in the recessive model (p = 0.0165).
Additionally, we noted a higher frequency of the homozygous VDR TaqI CC genotype (p = 0.117) and the mutated VDR TaqI C allele (p = 0.038) in the PE group than in the controls. However, stratified analysis did not show any significant association between VDR TaqI polymorphism and EOPE/LOPE. Our analysis did not show a statistically significant difference for either the VDR ApaI or FokI polymorphism between the whole PE group and the controls. However, for these polymorphisms we noted an association between possession of the VDR ApaI T allele and EOPE in the codominant model (p = 0.0431) and in the recessive model (p = 0.0275), as well as the VDR FokI T allele in the recessive model after BMI adjustment (p adj. = 0.0402).
The studies considering the participation of VDR gene variants in populations of pregnant women with preeclampsia in various regions of the world show inconclusive results. Farajian-Mashhadi et al., in a group of pregnant women from Iran (152 PE/160 healthy pregnant women), observed a lower frequency of the heterozygous VDR FokI Ff genotype in women with PE, both in maternal serum (p = 0.02) and in the placenta (p = 0.06). The VDR FokI polymorphism was also correlated with a lower risk of PE in the dominant model (Ff + ff vs. FF) both in the mother and in the placenta, indicating that both genotypes, heterozygous Ff (OR = 0.5, p = 0.007) and homozygous ff (OR = 0.5, p = 0.02), may reduce the risk of PE. No relationship was found between the presence of PE and VDR BsmI, TaqI, and ApaI polymorphisms in maternal serum and placenta. Additionally, no relationship was detected between the studied VDR polymorphisms and the severity of PE [57].
The influence of FokI, TaqI, and BsmI polymorphisms of the VDR receptor and the level of vitamin D (25(OH)-D) on the risk of PE and blood pressure were analyzed in the study by Rezavand et al. (100 PE/100 healthy pregnant women). The mean level of vitamin D in women with PE was significantly lower (16.6 vs. 19.6 ng/mL in healthy pregnant women, p < 0.001). A significantly higher frequency of the VDR FokI C allele was observed in women with PE (83%) compared to the control group (74%), which was associated with a 1.72-fold higher risk of PE. In the whole group of women, systolic and diastolic blood pressure values were significantly higher in carriers of the VDR FokI CC genotype compared to carriers of the VDR FokI TC and TT + TC genotypes. Moreover, VDR TaqI and VDR BsmI variants were not associated with the risk of PE. The study showed a relationship between the VDR FokI polymorphism, as well as insufficient level of vitamin 25(OH)-D in the mother's serum, and the risk of developing PE. Additionally, the influence of an insufficient level of vitamin 25(OH)-D and the VDR FokI polymorphism on increased blood pressure in female carriers of variants of this polymorphism was also demonstrated [58].
The study by Zhan et al. analyzed a group of pregnant women from the Chinese population (Han province, 402 PE/554 healthy pregnant women in the third trimester). The relationship between three polymorphisms of the VDR gene (VDR FokI rs2228570, Cdx2 rs11568820 and BsmI rs1544410) and the risk of PE appearance was analyzed. A statistically significant difference was observed in the frequency of the VDR FokI (G/A) polymorphism genotypes between the PE group and the control group (p = 0.001). The VDR G allele was a risk factor for the development of PE (p = 0.002, OR = 1.137). No relationship was found between the Cdx2 and BsmI polymorphisms and the risk of PE (p < 0.05) [59].
Contrary to the results obtained in our research, the above analyses did not show a direct relationship between BsmI polymorphism and the increased risk of PE.
In the study of Baca et al., a population of women from the USA (744 preeclampsia cases and 2411 controls) was analyzed. The polymorphisms of three genes-VDR, GC, and CYP27B1-were analyzed. Minor alleles of the rs12831006 polymorphism in the noncoding region of the VDR gene were significantly correlated with the risk of PE (rs12831006: OR = 1.5, p < 0.0001). The study also showed a relationship between the polymorphic variants of the GC gene: one in the intron (rs843010: OR = 1.4, p < 0.05) and two variants in the flanking region (rs842991: OR = 1.5, p < 0.05; rs16846876: OR = 0.75, p < 0.05). An association of the CYP27B1 gene polymorphism with the risk of PE was not found [60].
Some available analyses also suggest no influence of VDR gene polymorphisms on hypertensive complications in pregnancy. In the study by Rezende et al., the relationship between VDR polymorphisms (FokI, ApaI, and BsmI) and the occurrence of PE and GH (154 with GH, 162 with PE, and 213 healthy pregnant women) was analyzed. The frequency of all three VDR polymorphisms in the PE, GH, and healthy pregnant groups was similar (all p > 0.05) [61].
In our study we also performed haplotype analysis. Notably, each haplotype with significantly or borderline higher frequency in the PE group compared to controls contains the mutated BsmI A allele, which is correlated with PE development in our study (CTA p = 0.0241, TAT p = 0.0509, TTAT p = 0.0572).
Haplotype analysis has been performed in a few studies. Farajian-Mashhadi et al. reported that maternal and placental TABf haplotype may lead to decreased risk of PE development, and the placental TABF haplotype was associated with higher risk of PE [57].
Interesting results were also obtained in the analysis by Ghorbani et al., which indicated that in carriers of CYP T, VDR T, and RXR A (TTA) haplotypes compared to the GTG haplotype, the risk of PE was 6.71 times higher (p = 0.044). At the same time, in carriers of the heterozygous VDR ApaI GT genotype, an increased risk of PE (OR = 2.55, p = 0.04) was observed, while the presence of VDR ApaI GT + TT genotypes correlated with higher BMI and systolic blood pressure and a lower level of 25 (OH)-D3 [61].
In contrast to the above-mentioned literature, in a study by Rezende et al. haplotype frequency of VDR genes was similar in PE and GH groups and healthy pregnant women (all p > 0.05). These results did not show any association between VDR polymorphism or haplotypes and PE or GH development [60].
Caccamo et al. [62] analyzed FokI and BsmI polymorphisms of VDR gene in 116 women with gestational hypertension (GH) and 69 normotensive pregnant women. The authors reveal that VDR FF/bB haplotype is linked to GH development (increasing two folds GH risk). The interesting results show that insufficiency of vitamin D was detected in 92% GH women carrying of FF/bB haplotype. This study underlines that analysis of polymorphic variants could personalize vitamin D supplementation strategy in prevention of hypertensive disorders in pregnancy [62].
Additionally, in our study the correlation of VDR BsmI polymorphism with systolic and diastolic blood hypertension was noted. In carriers of the BsmI A allele, higher blood pressure values were observed. The results in our study reveal that the VDR BsmI polymorphism is closely associated with predisposition to higher hypertension. The results were corroborated by the research performed by Wang et al., in which the authors found associations of VDR BsmI and FokI variants with hypertension risk in Japanese men [49]. Additionally, a meta-analysis performed by Zhu et al. revealed a significant correlation between the VDR BsmI variant and susceptibility to hypertension [50]. The very exciting result in our study was the observation that 3-marker haplotype CTA (TaqI, ApaI, BsmI) was associated with significantly higher systolic (p = 0.0075) and diastolic (p = 0.0072) blood pressure values. This result indicates the combined influence of different polymorphisms on blood pressure modulation.
Summarizing the research carried out thus far, overall, it indicates the significant role of the endogenous vitamin D placental system in the development of PE. Furthermore, it was shown that VDR receptor is present in the decidua and syncytiotrophoblast, probably modulating the immune response and endothelial function. From this point of view, VDR polymorphic variants may play a crucial role in PE development. The results of our study provide additional evidence for the biological role of polymorphisms and haplotypes of the VDR gene in the etiology of PE. At this point, it should be emphasized that the analysis of polymorphic variants may lead to a personalized vitamin D supplementation strategy in the prevention of hypertensive disorders in pregnancy, which is a serious challenge in medical care.
Certainly, this study has a few limitations. First, a relatively small sized control group and low number of cases of early onset preeclampsia and late-onset preeclampsia. However, the analyzed group was selected very carefully with respect to age of patients and week of the end of pregnancy, allowing for a valuable case-control study analysis. On the other hand, the study covers the four most studied VDR polymorphisms. The frequency of the occurrence of particular genotypes and alleles was analyzed, and simultaneously haplotype analysis of the VDR gene polymorphisms was performed. Additionally, we carefully assessed the influence the VDR genetic variants on clinical parameters such as the blood pressure and body mass index of preeclamptic women. Another limitation is primiparity, which may be associated with elevated risk of preeclampsia. Moreover, in this study we did not perform measurements of the level of vitamin D in the study subjects. It should be emphasized that this kind of study indicating the importance of polymorphic variants of the VDR gene in PE etiology was performed for the first time in a population of preeclamptic Polish women. It makes this study valuable in the discussion about the pivotal role of VDR genetic variants in pathogenesis of hypertension in pregnancy in a particular European population and undoubtedly constitutes the strength of the study.

Conclusions
Association and haplotype analysis indicated that the VDR BsmI A allele could play a significant role in the PE pathomechanism and hence could be a risk factor for PE development. Additionally, the VDR BsmI polymorphism and CTA haplotype (TaqI/ApaI/BsmI), containing the VDR BsmI A allele, could be closely associated with higher values of systolic and diastolic blood pressure in the Polish population of pregnant women. However, the stratification analysis, after dividing PE group into early onset and late-onset form, showed also the influence of VDR TaqI and FokI polymorphisms on changing of blood pressure in pregnant women. Together the results of our study indicate the importance of the VDR BsmI polymorphism and reveal that this variant is closely associated with higher risk of PE development, thus these interesting results merit future studies.