Effects of Pre-Pregnancy Overweight/Obesity on the Pattern of Association of Hypertension Susceptibility Genes with Preeclampsia

The aim of this study was to explore the effects of pre-pregnancy overweight/obesity on the pattern of association of hypertension susceptibility genes with preeclampsia (PE). Ten single-nucleotide polymorphisms (SNPs) of the 10 genome-wide association studies (GWAS)-significant hypertension/blood pressure (BP) candidate genes were genotyped in 950 pregnant women divided into two cohorts according to their pre-pregnancy body mass index (preBMI): preBMI ≥ 25 (162 with PE and 159 control) and preBMI < 25 (290 with PE and 339 control). The PLINK software package was utilized to study the association (analyzed four genetic models using logistic regression). The functionality of PE-correlated loci was analyzed by performing an in silico database analysis. Two SNP hypertension/BP genes, rs805303 BAG6 (OR: 0.36–0.66) and rs167479 RGL3 (OR: 1.86), in subjects with preBMI ≥ 25 were associated with PE. No association between the studied SNPs and PE in the preBMI < 25 group was determined. Further analysis showed that two PE-associated SNPs are functional (have weighty eQTL, sQTL, regulatory, and missense values) and could be potentially implicated in PE development. In conclusion, this study was the first to discover the modifying influence of overweight/obesity on the pattern of association of GWAS-significant hypertension/BP susceptibility genes with PE: these genes are linked with PE in preBMI ≥ 25 pregnant women and are not PE-involved in the preBMI < 25 group.


Introduction
Preeclampsia (PE) is a multisystem disorder of pregnancy previously defined by the onset of hypertension accompanied by significant proteinuria after 20 weeks of gestation [1]. PE is a major cause of maternal and perinatal mortality and morbidity associated with a number of complications for both mother and fetus [1,2]. PE affects 5% to 7% of all pregnant women and is responsible for over 70,000 maternal deaths and 500,000 fetal deaths worldwide every year [2]. PE is associated with a number of short-(intrauterine fetal death, preterm birth, fetal growth restriction, low Apgar score, etc.) and long-term (cerebral palsy, hearing loss, visual impairment, insulin resistance, etc.) perinatal and postnatal complications, including death [1]. However, despite its prevalence, well-cataloged risk factors, and clinical characteristics, the exact pathophysiology of this disorder remains yet unknown [2]. This knowledge deficit has hampered the development of targeted therapies and limited treatment options for healthcare providers [3].
The epidemiology of PE reflects a broad range of risk factors as well as the heterogeneity and complexity of the disease [1,4]. The maternal pre-existing risk factors for PE A case-control study was conducted and included 950 women divided into two groups as follows (according to their pre-pregnancy BMI): Group I-women with preBMI ≥ 25 (n = 321), included 162 women with PE and 159 control; Group II-women with preBMI < 25 (n = 629), included 290 women with PE and 339 control. PE was defined according to the recommendations of the American College of Obstetricians and Gynecologists (the presence of systolic and/or diastolic blood pressure ≥ 140 mm Hg and/or ≥ 90 mm Hg, respectively, and proteinuria with excretion of 0.3 g or more of protein in a 24 h urine specimen) [29]. The control groups (preBMI ≥ 25 and <25) consisted of women without PE. A clinical examination of all pregnant women was conducted by an experienced obstetrician at the perinatal center of the St. Joasaph Belgorod Regional Clinical Hospital. Inclusion criteria were the following: singleton pregnancy, 37-40 gestation weeks [23,30], and born in Central Russia and self-reported Russian origin [31]; exclusion criteria were the following: pregnant with pathological placental location, uterine leiomyoma, malformation disease of female reproductive organs, hepatic/renal failure, diabetes mellitus, and isosensitization of blood group systems (ABO/Rh factor) [32,33]. Written informed consent was obtained from all participating individuals, and the present study was approved by the Local Ethical Committee of the Belgorod State University.
The main exposure variable was preBMI based on maternal pre-pregnancy weight in kilograms and maternal height in centimeters from antenatal care visits. PreBMI was calculated as body weight in kilograms divided by height in meters squared (kg/m 2 ) [34]. PreBMI was categorized according to WHO definitions as underweight < 18.5, normal weight 18.5-24.9, overweight 25.0-29.9, and obese ≥ 30.

Statistical Analysis
Deviations from Hardy-Weinberg equilibrium (HWE) were assessed by the goodnessof-fit χ 2 test [42]. The SNPs' data were analyzed by logistic regression under dominant, additive, recessive, and allelic inheritance models [43] separately in two studied groups (preBMI ≥ 25 and <25). Baseline and clinical characteristics that could potentially influence the risk of PE separately in each study group of pregnant women (age, BMI, number of gravidity, spontaneous and induced abortions, presence of obesity, family history of PE for preBMI ≥ 25 group, and presence of family history of PE and tobacco consumption for preBMI < 25 group (Table 1)) were used as confounding factors in logistic regression models. To adjust for the multiple comparisons, an adaptive permutation testing was used [44,45]. The association analysis was performed using the PLINK package [46]. A p perm level less than or equal to 0.025 was regarded as significant (a Bonferroni correction according to the number of the groups compared, n = 2, was additionally performed).

Results
Baseline and clinical data of preeclamptic and normotensive pregnant women with preBMI ≥ 25 (162 case and 159 control) and women with preBMI < 25 (290 cases and 339 control) are presented in Table 1. As shown in Table 1, the women with preBMI ≥ 25, PE cases vs. non-PE, were more likely to be older (p = 0.001) and had higher pre-pregnancy BMI (p = 0.0001), obesity (p = 0.0001), family history of PE (p = 0.02), mean number of gravidity (p = 0.03), and spontaneous (p = 0.008) and induced (p = 0.003) abortions. Among the women with preBMI < 25, PE cases vs. non-PE, there was a higher percentage of family history of PE (p = 0.0008) and lower proportion of tobacco consumption (p = 0.05). According to the received data, we used the abovementioned "specific" parameters for each study group of pregnant women (preBMI ≥ 25 and preBMI < 25) as covariates in the association regression analysis.
Among women with preBMI < 25 (Supplementary Table S3) and preBMI ≥ 25 (Supplementary Table S4), genotype frequencies were in Hardy-Weinberg equilibrium in the PE and control groups for all considered loci (after the Bonferroni correction for 10 SNPs P bonf > 0.005).
The data association analysis in Table 2 showed that there were significant differences in the relationship between the genetic polymorphisms of the hypertension/BP genes and the PE in different preBMI groups. After adjustment for multiple comparisons (used Bonferroni correction), statistically significant associations (p perm ≤ 0.025) between studied SNPs and risk of PE were found only among women with preBMI ≥ 25 (Table 2). In this cohort of subjects, two SNPs of hypertension/BP genes (rs805303 of the BAG6 and rs167479 of the RGL3) were associated with PE. The polymorphic variant of SNP rs805303 (A allele) was negatively associated with PE using three genetic models: allelic (OR: 0. 66  There were no significant differences in the genotype/allele frequencies of the hypertension/BP gene polymorphism between preeclamptic patients and normotensive pregnant women with preBMI < 25 (p perm > 0.025) ( Table 2).
In Silico Data of Functional PE-Associated SNPs SNP rs1674769. Genetic variation caused by PE-associated SNP rs1674769, occurring in protein coding regions, alters the encoded amino acid at the mutated site and causes structural (Pro162His RGL3) and functional (predictive value is «deleterious» by SIFT and «probably damaging» by PolyPhen-2) changes in the mutated protein.
Based on STRING resource annotations, the interactomic networks of nine TFs associated with the PE risk allele at SNP rs167479 of RGL3 were visualized (

In Silico Data of Functional PE-Associated SNPs
SNP rs1674769. Genetic variation caused by PE-associated SNP rs1674769, occurring in protein coding regions, alters the encoded amino acid at the mutated site and causes structural (Pro162His RGL3) and functional (predictive value is «deleterious» by SIFT and «probably damaging» by PolyPhen-2) changes in the mutated protein.
Based on STRING resource annotations, the interactomic networks of nine TFs associated with the PE risk allele at SNP rs167479 of RGL3 were visualized (Figure 1   Using the public GTEx database, the expression quantitative effect of rs1674769 was observed: individuals carrying the PE risk G allele showed lower levels of expression of CTC-510F12 in the pituitary (Supplementary Table S5). SNP rs805303. The effects of SNP rs805303 on chromatin structure and allele-specific transcription factor binding were identified using HaploReg. The intronic annotation of rs805303 BAG6 indicated that it affected DNA motifs such as CACD (PE protective allele A decreases affinity to HIM, ∆LOD: −2.2) and that there was a direct effect on enhancer (hESC-derived CD56+ mesoderm cultured cells, H9-derived neuronal progenitor cultured cells, and Primary B and T cells (regulatory, effector/memory enriched, helper, etc.) from peripheral blood and brain (hippocampus middle, anterior caudate, dorsolateral and prefrontal cortex, etc., and male fetal brain, fetal adrenal gland, fetal muscle trunk, etc.)) and promoter (brain germinal matrix, adipose-derived mesenchymal stem cell cultured cells, adipose nuclei, etc.) histone markers.
In the Blood eQTL database, the minor allele of the genetic variant rs805303 was associated with decreased whole-blood mRNA expression of four genes: LY6G5C (Z-score: Based on in silico data of the GTEx tool, the pronounced tissue-specific gene expression effects of rs805303 were found (Supplementary Table S5). It was inferred from GTEx that rs805303 is associated with the expression of 35 genes in 44 tissues/organs (Supplementary  Table S5). This locus regulates the expression level of many genes in the PE pathophysiology important organs (tissues), such as the brain (cortex, substantia nigra, basal ganglia,  Table S5).

Discussion
This study was the first to identify a BMI-specific association of GWAS-significant hypertension/BP susceptibility genes with PE: rs805303 of BAG6 (protective allele: A, OR: 0.36-0.66) and rs167479 of RGL3 (risk allele: G, OR: 1.86) were associated with PE in preBMI ≥ 25 pregnant women and not associated with disorder in the preBMI < 25 group. Pronounced pleiotropic tissue-specific regulatory/expression/splicing effects of the PE-associated SNPs (rs1674769 and rs805303) were also documented. The most significant functionality (multi-expression and splicing patterns) was registered for rs805303, which affected 41 various genes.
Literature data overwhelmingly support that high BMI is a major risk factor for PE [4,6,10]. Overweight and obese women have, respectively, an increased risk of developing PE with severe features at ≥34 wks of gestation (overweight, OR = 1.4; obese, OR = 2.0) [15]. While numerous epidemiological studies have demonstrated that obesity/overweight increases the risk of PE, these mechanisms have yet to be fully elucidated [9]. It may be that placental, adipose tissue, and underlying endothelial dysfunction by metabolic factors such as leptin, fasting insulin, insulin resistance, etc., mediate the impact of obesity on increasing the risk for PE [10,11,[58][59][60]. Spradley et al. [9] hypothesized that obesity-related metabolic factors increase the risk for developing PE by impacting Figure 3. The protein-protein interaction networks of the candidate genes associated with rs805303 (eQTL/sQTL/regulatory effects this SNP) inferred using STRING (https://string-db.org/ (accessed on 16 June 2022)).

Discussion
This study was the first to identify a BMI-specific association of GWAS-significant hypertension/BP susceptibility genes with PE: rs805303 of BAG6 (protective allele: A, OR: 0.36-0.66) and rs167479 of RGL3 (risk allele: G, OR: 1.86) were associated with PE in preBMI ≥ 25 pregnant women and not associated with disorder in the preBMI < 25 group. Pronounced pleiotropic tissue-specific regulatory/expression/splicing effects of the PEassociated SNPs (rs1674769 and rs805303) were also documented. The most significant functionality (multi-expression and splicing patterns) was registered for rs805303, which affected 41 various genes.
Literature data overwhelmingly support that high BMI is a major risk factor for PE [4,6,10]. Overweight and obese women have, respectively, an increased risk of developing PE with severe features at ≥34 wks of gestation (overweight, OR = 1.4; obese, OR = 2.0) [15]. While numerous epidemiological studies have demonstrated that obesity/overweight increases the risk of PE, these mechanisms have yet to be fully elucidated [9]. It may be that placental, adipose tissue, and underlying endothelial dysfunction by metabolic factors such as leptin, fasting insulin, insulin resistance, etc., mediate the impact of obesity on increasing the risk for PE [10,11,[58][59][60]. Spradley et al. [9] hypothesized that obesity-related metabolic factors increase the risk for developing PE by impacting various stages in the pathogenesis of PE (cytotrophoblast migration and placental ischemia; release of soluble placental factors into the maternal circulation; maternal endothelial and vascular dysfunction). Authors put forth the concept that obesity and metabolic factors such as lipids, insulin, glucose, and leptin affect placental function and increase the risk of developing hypertension in pregnancy by reducing placental perfusion, enhancing placental release of soluble factors, and by increasing the sensitivity of the maternal vasculature to placental ischemia-induced soluble factors [9]. BMI-related dyslipidemia and elevated C-reactive protein increase the risk of PE [61,62]. It should be noted that BMI may modify the association of C-reactive protein with PE [61].
Mendelian randomization has previously indicated causal associations of genetically predicted BMI and visceral adipose tissue (VAT) mass with PE (OR = 2.09 per 1 SD increase in obesity trait and OR = 3.08 per 1 kg increase in predicted VAT mass, respectively) (analyzed data of 257,193 women of European ancestry in UK Biobank and publicly available genome-wide association studies) [14]. Furthermore, the work [14] showed that on the one hand, leptin and insulin influence the risk of PE independently of obesity, but on the other hand, leptin, fasting insulin, and insulin resistance each mediated between 20% and 50% of the total genetically predicted association of obesity with PE.
Obesity/overweight and PE are diseases that result from multiple genetic and environmental factors [9]. These two conditions share many pathophysiological mechanisms, however, only 10% of obese women will develop PE [6,63]. Important here may be "additional" risk factors, including genetic ones, which increase the probability of developing PE for obese women. These genetic risk factor developments of PE for obese/overweight women can be GWAS-significant hypertension/BP susceptibility genes (BAG6 and RGL3), which, according to the results of our work, are involved in the PE development in preBMI ≥ 25 pregnant women and not associated with disorder in the preBMI < 25 group. These genes may be part of shared genetic components of etiological relationships of overweight/obesity with PE.
According to data from earlier GWAS, rs805303 of BAG6 was associated with levels of both systolic and diastolic BP and hypertension [64,65], and rs167479 of RGL3 was associated with BP (systolic, diastolic and pulse, and mean arterial pressure) and hypertension [65][66][67][68][69][70][71]. Moreover, the materials we obtained on the protective effect of allele A rs805303 of BAG6 (OR: 0.36-0.66) and the risk role of allele G rs167479 of RGL3 (OR: 1.86) for PE in preBMI ≥ 25 pregnant women are fully consistent with the effects of these alleles, established in the above GWAS: allele A rs805303 BAG6 was associated with low BP and decreased risk of hypertension and allele G rs167479 RGL3 was correlated with high BP and elevated risk of hypertension.
It should be noted that our data on significant correlations between genetic predisposition to hypertension/BP and PE in preBMI ≥ 25 pregnant women are consistent with the results of previous studies on this issue [26][27][28]. According to the multi-ethnic maternal PE GWAS data from Gray et al., the disorder-associated locus rs9478812 PLEKHG1 [28] has previously been implicated in GWAS of BP and BMI [72,73]. In the GWAS, Steinthorsdottir et al. found five variants (rs259983 ZNF831 and rs1421085 FTO were associated on genomewide significance, p ≤ 4 × 10 −9 ; rs16998073 FGF5, rs3184504 SH2B3, and rs419076 MECOM were associated on significance 4 × 10 −9 < p < 5.6 × 10 −5 ) associating with PE through the maternal genome [27]. All of them have previously been GWAS-associated with BP [74][75][76], and one SNP, rs1421085 FTO, has also been GWAS-associated with BMI [77]. In addition, Steinthorsdottir et al. showed a positive genetic correlation between systolic and diastolic BP, hypertension, and PE (r g = 0.3-0.4) and uncovered an association of the polygenic risk score for hypertension with PE [27]. Previously published GWAS completed by Gray et al. demonstrated that risk alleles for elevated DBP and increased BMI were most strongly associated with PE risk [28].
The specific genetic mechanisms by which known clinical risk factors contribute to PE development, including hypertension and obesity/overweight, have not been fully elucidated [28]. Gray et al. express the following proposed genetic mechanisms for these relationships [28]. Firstly, in previously published hypertension GWAS, risk loci are enriched for regulatory elements affecting gene expression in vascular endothelial cells and are associated with end organ damage in the heart, cerebral vessels, carotid artery, etc. [78]. As PE is characterized by diffuse endothelial dysfunction [2,3,7], women with genetic predisposition to altered vascular endothelial cell function are likely to be at high risk of PE. Secondly, in a published obesity GWAS, identified risk loci are highly associated with brain regions important for appetite regulation, learning, emotion, memory, insulin utilization, energy/lipid metabolism, and adipogenesis [79]. These pathways mediated through BMI-related effects may directly or indirectly contribute to PE pathophysiology, and in the opinion of Gray et al., it is an important question for future investigations [28].
One of the parts of the possible answer to the question about the mechanisms of BMI-related effects and their correlation with genetic predisposition to hypertension/BP with PE may be our data on the pronounced functional effects on GWAS-significant SNPs of the hypertension/BP susceptibility genes associated with PE. Regarding the specific genetic locus tagged by the study-significant SNP, rs805303, this SNP lies within an intronic region of BAG6 in a region predicted to have an "active" promoter (marked by H3K9ac_Pro modified histone) in adipose-related cultured cells (mesenchymal-stem-cell-derived adipocyte, adipose-derived mesenchymal stem cell, and adipose nuclei). This genetic variant demonstrated the regulatory role for mRNA expression/splicing patterns of 35/17 genes in more than 35 tissues/organs, including 10/5 genes in visceral and subcutaneous adipose (BAG6, CYP21A1P, CYP21A2, HLA-DRB1, HLA-DRB5, HLA-DRB6, etc.), which enrich the regulation of immune system processes and organelle membrane components. The locus rs167479 of RGL3 is situated in the enhancer and promoter in placenta amnion. This SNP causes structural (Pro162His) and functional («deleterious» predictive potential) changes in the RGL3 protein. The T to G change at rs167479 (G, effect allele; T, other allele) is predicted to alter the binding sites of nine transcription factors, AP-1, CCNT2, Rad21, SETDB1, SP1, TR4, WT1, ZNF219 and Zic, which are involved in the regulation of the gene transcription process and estrogen/androgen receptor signaling pathways. Thus, the regions tagged by rs167479 and rs805303 are likely to be functionally important and, consequently, may be the specific regions responsible for increasing/decreasing PE risk. It should be noted that it is interesting that one of the genes whose expression in nerve tissue is determined by rs805303 POU5F1 (data obtained in silico in this work), together with genes ESRRG and ZNF554 (due to gene transcription regulation), has the highest number of significant correlations with predominantly placenta-expressed genes, and, hence, may be deemed as a hub factor for PE development [80]. Than et al. pointed to the possible involvement of these genes in the dysregulation of trophoblast differentiation in preterm PE [80].

Conclusions
This study demonstrated a BMI-specific association of GWAS-significant hypertension/BP susceptibility genes with PE: a significant association was found between these genes and PE in preBMI ≥ 25 pregnant women, and their absence was found in the preBMI < 25 group. It was shown that genetic predisposition to hypertension/BP is an important risk factor for PE in overweight/obese women.