Associations between maternal matrix metalloproteinase gene polymorphisms and birth weight

Birth weight and length are key parameters for assessing neonatal status and pregnancy outcomes. They have substantial prognostic value and may serve as biological indicators of future health [1–4]. Compared with children with an average birth weight, neonates with high or low birth weight are at increased risk of developing various diseases throughout life, including mortality [1, 2, 5].

Normal human pregnancy is associated with substantial haemodynamic, vascular, and uteroplacental changes that ensure adequate placentation as well as sufficient blood flow and nutrient delivery to the developing fetus [6]. Inadequate trophoblast invasion and impaired remodelling of the uterine spiral arteries result in persistent placental ischaemia and hypoxia, substantially increasing the likelihood of fetal growth restriction (FGR) and other pregnancy complications [7–9]. Matrix metalloproteinases (MMPs), a family of zinc-dependent hydrolases that cleave extracellular matrix proteins and glycoproteins, play an important role in these processes [10]. Twenty-eight types of MMPs have been identified in vertebrates, 23 of which are expressed in human tissues [11]. Based on available evidence, MMPs are classified into groups according to the type of proteolytic substrate they act on and degrade: collagenases (MMP-1 and MMP-13), gelatinases (MMP-2 and MMP-9), stromelysins (MMP-3), and membrane-type MMPs [11]. Physiologically, MMPs play essential roles in various biological processes, including tissue repair and remodelling, cell proliferation and differentiation, angiogenesis, apoptosis, ovulation, endometrial proliferation, and embryogenesis [12, 13]. MMPs are expressed in various tissues and cells, including cytotrophoblasts, syncytiotrophoblasts, and the maternal decidua, thereby contributing to endometrial and maternal vascular remodelling during pregnancy and promoting trophoblast invasion of the spiral arteries [12, 13]. Abnormal MMP activity in trophoblasts or maternal cells substantially disrupts normal spiral artery remodelling and contributes to uteroplacental ischaemia, thereby increasing the risk of FGR and low birth weight [14–18].

However, the number of current studies investigating the association between MMP gene polymorphisms and neonatal birth weight or FGR is very limited [19, 20]. In the first study, Sengar A. et al. [19] investigated the effect of the fetal MMP-2 C1306T polymorphism on the risk of FGR. They found that carriage of the heterozygous CT genotype significantly increased the risk of delivering a low-birth-weight infant compared with the CC genotype. The study by Mărginean C.O. et al. [20] investigated the associations of two SNPs (rs17577 and rs17576) in the MMP-9 gene with birth weight in a Romanian population. The authors found that the rs17577 polymorphism in the maternal genome was significantly associated with lower birth weight. Contemporary studies investigating the role of MMP gene polymorphisms in pregnancy disorders have focused primarily on pre-eclampsia, pregnancy loss, and spontaneous abortion [21–25].

This study aimed to investigate the associations between maternal matrix metalloproteinase gene polymorphisms and birth weight.

Materials and methods

For this molecular genetic study, a sample of 317 pregnant women was consecutively recruited from the inpatient department of the Perinatal Center at the Belgorod Regional Clinical Hospital between 2008 and 2016. The inclusion criteria were: provision of written voluntary informed consent, self-reported Russian ethnicity, birth in the Central Black Earth region of the Russian Federation, and a singleton pregnancy at 24 to 41 weeks of gestation resulting in a live birth. Exclusion criteria included: refusal to provide informed consent, consanguinity among participants, non-Russian ethnicity, birth outside the Central Black Earth region of Russia, multiple pregnancy, congenital anomalies in the mother or fetus, and preterm delivery before 24 weeks of gestation.

Clinical, laboratory, and instrumental assessments of the pregnant women and neonates were conducted at the Perinatal Center of the Belgorod Regional Clinical Hospital. Obtaining informed consent from each participant was a mandatory requirement for study enrollment. The study was reviewed and approved by the Research Ethics Committee of the Medical Institute of Belgorod State University (Protocol No. 2, dated February 13, 2008), in accordance with the standard requirements established by the Ethics Committee of the Russian Federation.

Molecular genetic analyses were performed at the Department of Medical and Biological Disciplines, Medical Institute, National Research Belgorod State University. Venous blood samples (4–5 mL) were collected from the antecubital vein using EDTA-containing Vacutainer tubes. Genomic DNA was extracted from the blood samples using the phenol-chloroform method and stored at -80°C. For subsequent genotyping, DNA samples were diluted to a working concentration of 5–10 ng/μL, as quantified by a NanoDrop 2000c spectrophotometer (Thermo Scientific, USA).

To identify associations between candidate genes and neonatal birth weight, single nucleotide polymorphisms (SNPs) were selected based on the following three criteria: 1) gene polymorphisms whose protein products are crucial for embryonic and fetal intrauterine development, according to the KEGG PATHWAY [26] and Reactome Pathway [27] databases; 2) functional role, including associations with nonsynonymous substitutions, epigenetic effects on neighboring genes, and regulation of gene expression and alternative splicing (this assessment was performed in silico using the HaploReg v4.2 online resource [28]); and 3) a minor allele frequency of ≥5% in the European population. Consequently, five polymorphic loci of matrix metalloproteinase (MMP) genes were selected for molecular genetic analysis: rs1799750 (1G>2G) in MMP-1, rs243865 (C>T) in MMP-2, rs3025058 (6A>5A) in MMP-3, rs11568818 (T>C) in MMP-7, and rs17577 (G>A) in MMP-9.

Genotyping of the polymorphic loci was performed using real-time polymerase chain reaction (PCR). TaqMan probes were used for the detection of reaction products, and analyses were conducted on a CFX96 instrument (Bio-Rad, USA).

Statistical analysis

In this study, the distributions of alleles and genotypes for all investigated single-nucleotide polymorphisms (SNPs) were compared for consistency with Hardy–Weinberg equilibrium. Associations between polymorphisms and neonatal birth weight were assessed using gPLINK software (Java version 2.050) by log-linear regression analysis under four genetic models: allelic, additive, dominant, and recessive. To reduce the likelihood of false-positive findings, a multiple-comparison adjustment accounting for the number of analyzed SNPs was applied using an adaptive permutation test to calculate permutation p values (p_perm). Permutation testing effectively addresses the problem of multiple comparisons, including in genome-wide association studies [29]. Transformed neonatal birth-weight values were generated in R because the distribution of birth weight in the study sample of pregnant women deviated from normality, as assessed using the Shapiro–Wilk test. Median and 25th/75th percentiles (Me [Q1; Q3]) were also used to describe the observed (untransformed) birth-weight values.

At the initial stage of the study, we assessed the effects of maternal factors—maternal age, pre-pregnancy body-mass index, age at menarche, history of sexually transmitted infections and pre-eclampsia, smoking, and alcohol consumption—on neonatal birth weight. For this purpose, Kruskal–Wallis analysis of variance (Kruskal–Wallis ANOVA) was performed across the study subgroups of women, which were defined according to the aforementioned maternal factors. When significant differences were identified, post-hoc pairwise analyses were conducted to determine the subgroups between which statistically significant differences were present, using the Mann–Whitney U test.

The direction of the identified associations between SNPs and neonatal birth weight was determined using the regression coefficient (β). A positive β value indicated an association between the minor allele and increased neonatal birth weight. Conversely, a negative β value indicated lower neonatal birth weight in the presence of the minor allele. A p_perm value of 0.05 or less was considered statistically significant.

The functional relevance of polymorphic loci associated with neonatal birth weight and SNPs in strong linkage disequilibrium with these loci (at a correlation coefficient of r²≥0.80) was evaluated. For this purpose, we used the widely applied in-silico methodology and six contemporary bioinformatics tools: HaploReg [28] – epigenetic changes; SIFT (Sorting Tolerant From Intolerant) [31] and PolyPhen 2 (Polymorphism Phenotyping v2) (http://genetics.bwh.harvard.edu/pph2/) – missense mutations; GTExportal [32] and the Blood eQTL Browser (http://www.genenetwork.nl/bloodeqtlbrowser) to assess the effects of SNPs on gene expression and alternative splicing in different tissues.

Results and discussion

The primary demographic, clinical, and anamnestic characteristics of the study cohort of pregnant women are presented in Table 1.

For all investigated polymorphic loci, the genotype and allele frequency distributions conformed to Hardy–Weinberg equilibrium expectations. Additionally, the minor allele frequency for all loci exceeded 5%.

Analysis of single nucleotide polymorphism (SNP) associations with birth weight identified two loci associated with this phenotype: the 5A allele of rs3025058 in MMP-3 [allelic model (β=-0.112, p=0.002, p_perm=0.003), additive model (β=-0.125, p=0.001, p_perm=0.001), dominant model (β=-0.155, p=0.009, p_perm=0.013), recessive model (β=-0.194, p=0.002, p_perm=0.002)] and the 2G allele of rs1799750 in MMP-1 [recessive model (β=0.153, p=0.015, p_perm=0.018)] (Table 2).

An evaluation of functional/functional-genomic effects for loci associated with birth weight revealed that rs3025058 in MMP-3 is in linkage disequilibrium with rs679620 in MMP-3, which causes a missense substitution of lysine to glutamate at position 45 of the MMP-3 protein (also known as progelatinase). This variant is classified as benign, with a risk score of 0.00. The sensitivity and specificity estimate for this variant are 1.00 and 0.00, respectively.

The investigated rs3025058 polymorphism of MMP-3 is located in the promoter region and represents an insertion/deletion variant characterized by five or six adenine residues, with the 5A allele exhibiting higher promoter activity than the 6A allele [33]. Furthermore, the rs3025058 polymorphism of MMP-3 is in linkage disequilibrium (r²>0.8) with 39 different SNPs, 35 of which affect the expression of MMP-1 in the thyroid gland, visceral adipose tissue, and cultured fibroblasts (p_FDR ≤ 0.05).

The MMP-3 gene encodes MMP-3 (stromelysin-1), an enzyme directly involved in vascular remodeling [34]. MMP-3 expression has been detected in amniotic epithelial cells, syncytiotrophoblasts, and villous trophoblasts across all stages of gestation [18]. The enzymatic action of MMP-3 involves the degradation of collagen, proteoglycans, elastin, fibronectin, laminin, and gelatin, conferring its role in tissue remodeling, trophoblast motility, and angiogenesis, and playing a vital role in physiological remodeling and trophoblast invasion [13]. Research evaluating the role of MMP-3 in pregnancy complications has primarily focused on preeclampsia or preeclampsia complicated by FGR [18, 35]. Laskowska M. et al. [35] demonstrated elevated serum MMP-3 levels in women with early-onset preeclampsia, whereas enzyme concentrations in late-onset preeclampsia did not correlate with the disease. In a subsequent study [18], the same group reported higher maternal serum MMP-3 concentrations in patients with FGR complicating severe preeclampsia compared with women with severe preeclampsia without FGR.

Our findings indicate that the 5A allele of rs3025058 in MMP-3 is a risk factor for lower birth weight. It is noteworthy that the direct involvement of MMP-3 in vascular remodeling processes has led molecular genetic research to focus predominantly on the association between rs3025058 in MMP-3 and various cardiovascular diseases. However, findings across these studies remain inconsistent. For instance, an earlier study by Huang X.Y. et al. [36] demonstrated an association between rs3025058 in MMP-3 and an increased risk of ischemic stroke, whereas two subsequent meta-analyses failed to detect such associations [33, 37]. Association studies of rs3025058 in MMP-3 have also investigated susceptibility to hypertension [38, 39] and coronary artery disease [40] in diverse populations.

The 2G allele of rs1799750 in MMP-1, which was associated with higher  birth weight in our study, is correlated with reduced expression of MMP-1 in multiple organs and tissues relevant to fetal growth regulation: cultured fibroblasts (β=-0.66, p=9.6×10^-84, p_FDR≤0.05), thyroid gland (β=-0.52, p=1.3×10^-25, p_FDR≤0.05), lungs (β=-0.42, p=1.9×10^-25, p_FDR≤0.05), visceral adipose tissue (β=-0.45, p=1.7×10-¹⁸, p_FDR≤0.05), and subcutaneous adipose tissue (β=-0.28, p=2.4×10^-8, p_FDR≤0.05).

The MMP-1 gene encodes matrix metalloproteinase-1 (MMP-1; collagenase-1), a member of the collagenase family that cleaves several extracellular matrix proteins, though its primary function is the degradation of various types of fibrillar collagen [13]. Collagenase-1 is expressed in the placental cytotrophoblast, syncytiotrophoblast, and decidua, suggesting a potential role in trophoblast invasion [11]. Several studies have shown that pregnant women with preeclampsia exhibit lower MMP-1 concentrations in umbilical cord blood, placenta, and decidua compared with women with uncomplicated pregnancies, with the lowest levels recorded in those with severe preeclampsia [41]. Reduced MMP-1 concentrations/activity and elevated levels of its inhibitor were also detected in umbilical artery tissues of women with preeclampsia [11]. Furthermore, in cultured human decidual endothelial cells, basal and stimulated MMP-1 secretion was decreased in women with preeclampsia compared with those with normal pregnancies. Decreased MMP-1 expression in maternal decidual endothelial cells may suppress cytotrophoblast invasion and impair the capacity of the trophoblast to penetrate maternal decidual blood vessels [12]. Conversely, reduced MMP-1 concentrations/activity inhibit collagen degradation; collagen accumulation can decrease arterial wall compliance and reduce fetal blood flow [11]. Notably, however, vascular expression and plasma concentrations of MMP-1 were elevated in women with preeclampsia [41]. In addition, MMP-1 induced vasoconstriction and enhanced responsiveness to angiotensin II in isolated omental vessels from pregnant women [42]. Thus, elevated MMP-1 levels in the tunica media of uterine arteries may increase vascular tone and reactivity to vasoconstrictors, exacerbating uteroplacental ischemia and potentially contributing to the pathogenesis of preeclampsia and fetal growth restriction.

Molecular genetic studies of MMP-1 rs1799750 primarily focus on its association with the risk of pregnancy complications, such as preeclampsia and pregnancy loss [21, 23, 25]. Sakowicz A. et al. [21] investigated the association of five single-nucleotide polymorphisms (SNPs) in the MMP-1, MMP-2, MMP-3, and MMP-9 genes – in both mothers and neonates – with the risk of preeclampsia in a Polish population. Maternal or neonatal carriage of the MMP-1 rs1799750 1G/1G genotype was found to significantly increase the risk of this condition. However, a subsequent study in a Polish cohort evaluating the potential role of MMP-1 rs1799750 in the pathogenesis of preeclampsia revealed no statistically significant associations [25]. Another study in a Polish population evaluated associations between five SNPs in MMP genes and the risk of pregnancy loss [23]. Similarly, no significant association was observed for MMP-1 rs1799750.

Conclusion

Thus, the MMP-3 rs3025058 5A allele was associated with lower birth weight, whereas the MMP-1 rs1799750 2G allele was associated with higher birth weight. Both loci were also associated with MMP-1 expression in visceral and subcutaneous adipose tissue, thyroid and lung tissue, and cultured fibroblasts.