Molecular Analysis of HLA-G in Women with High-Risk Pregnancy and Their Partners with Regard to Possible Complications

The understanding of the molecular and biochemical characteristics of the human leukocyte antigen-G (HLA-G) is important because of the diverse influence of this antigen’s polymorphisms on the course of a pregnancy. The aim of our study was to assess how the variation of the HLA-G allele and the HLA-G 14-bp ins/del polymorphism influence predisposition to a complicated pregnancy. The clinical material consisted of parental pairs with complicated pregnancies (210 women; 190 men). The control group included parental pairs without complications during pregnancy (89 women; 86 men). The study involved isolation of genome DNA from peripheral blood leukocytes, sequencing, and analysis of the 14-bp ins/del polymorphism in the 3′-untranslated region (3′-UTR) of the HLA-G gene based on polymerase chain reaction (PCR). The most common HLA-G allele in the group of women with complicated pregnancies was the HLA-G 10101 allele. There were no statistically significant differences in the frequencies of the 14-bp ins/del polymorphism in the 3′UTR of the HLA-G gene between the groups. Our results suggest that the risk of complications in pregnancy is influenced by the HLA-G 10101, HLA-G 10108, and HLA-G 10106 alleles and is not influenced by the 14-bp ins/del polymorphism in the 3′UTR of the HLA-G gene.


Introduction
During embryo implantation and pregnancy, the maternal immune system comes into close contact with fetal trophoblast cells. To avoid rejection of a semiallogeneic fetus, the mechanisms modulating the maternal immune system must be initiated [1].
The understanding of the molecular and biochemical characteristics of the human leukocyte antigen-G (HLA-G) and its products is important due to the diverse influence of this antigen's polymorphisms on the course of the pregnancy. HLA-G was the first identified trophoblast HLA, located on the short arm of the 6p21.1-6p.21.3 chromosome. It is a highly polymorphic group, including HLA class Ia genes (HLA-A, -B, and -C) and HLA class Ib genes (HLA-E, -F, and -G) [2]. The roles of the majority of the HLA molecules have already been well documented. HLA class I molecules are located on the surface of all nucleated cells of the body, erythrocytes, and thrombocytes. The expression of MHC class II molecules-responsible for presenting antigens to CD4+ T-helper lymphocytes-is mainly observed on the surface of antigen-presenting cells (APC). This process belongs to the key mechanisms of immune response regulation that enable recognition and effective reaction to foreign antigens, but may also provoke pathological recognition of the body's own antigens [2,3].
The HLA-G polymorphism is the one that seems to be the most essential from the point of view of the course of a pregnancy. Belonging to the nonclassical HLA class I molecules, HLA-G has been, for the first time, detected on the surface of trophoblast cells in the early stage of pregnancy. It is believed to be a factor that influences the process of implantation of the fertilized ovum in the mucous membrane of the uterus [4]. HLA-G expression is observed in ova immediately after fertilization at the oocyte stage [5]. Under physiological conditions, the HLA-G antigen protects fetal tissues against natural killers (NKs) and cytotoxic lymphocytes through inhibiting their activity. Decreased expression of HLA-G has been observed under conditions posing a threat to pregnancy, such as preeclampsia and recurrent spontaneous abortion, which confirms its protective effect on pregnancy [6].
It is worth emphasizing that the molecular structure of HLA-G has many polymorphic variants, but most of them have no impact on the sequences of emerging amino acids. Nevertheless, there are also mutations contributing to the protein structure; a single base pair (bp) deletion at nucleotide 1597 causes frame-shift to the 130th amino acid. The cytosine deletion on codon 130 results in the creation of the null allele (called G * 0105N) that cannot encode the functional HLA-G isoform. It has been demonstrated that the null allele may be related to a higher risk of recurrent spontaneous abortion [7]. The mutation that is essential for the course of pregnancy is the HLA-G gene 14-bp insertion (ins)/deletion (del) (rs66554220) polymorphism in exon 8, which has an impact on the HLA-G mRNA stability and isoform splicing patterns [8]. HLA-G expression is regulated by some polymorphisms in the 5 -URR (upstream regulatory region) and also a 14-base pair (bp) ins/del in the 3 -untranslated region (3 -UTR). Researchers have reported a significant association between the single nucleotide polymorphism (SNP) at position 1754 in exon 8 in the 3 -UTR (untranslational region) of the HLA-G gene and the risk of preeclampsia [9].
The aim of this study was to assess how the variation of the HLA-G allele and the HLA-G 14-bp ins/del polymorphism in parental pairs influence their predisposition to pregnancy complicated by antiphospholipid syndrome (APS), preeclampsia (PE), intrauterine growth restriction (IUGR), and recurrent spontaneous abortion (RSA).

Experimental Groups and a Control Group
The clinical material consisted of retrospectively assessed parental pairs (211 women aged 20-35 years and their partners-189 men aged 21-42 years) in which women had experienced an unsuccessful or complicated pregnancy. The control group included parental pairs-89 women and their 86 partners-having offspring without complications during pregnancy. All studied pregnant women were Caucasian Poles. Blood samples from the cubital vein were taken from all participants. Blood for laboratory analysis was collected from the women in the experimental and the control groups after the end of pregnancy.
All women involved in the study were patients of the Clinic of Maternal and Fetal Medicine, the Outpatient Clinic at the Clinic of Maternal and Fetal Medicine, and the Rheumatology Outpatient Clinic, Pomeranian Medical University in Szczecin (Poland). The protocol was approved by the Bioethical Commission of the [covered for blind review] (approval number BN-001/94/07). All participants gave voluntary and informed written consent to take part in the study.

Division into Groups
The women involved in the study and their partners were divided into a control group (C) and four experimental groups depending on the course of pregnancy, namely a group with antiphospholipid syndrome, a group with preeclampsia, a group with intrauterine growth restriction, and a group with recurrent spontaneous abortion. The patients were assigned to the groups according to the criteria shown in Table 1. The syndrome was diagnosed on the basis of clinical criteria (the patient's obstetric history, and/or experience of embolic/thrombotic complications, and/or autoimmune diseases), and laboratory criteria. The criteria for inclusion in the group were [10]: (1) Thrombosis-one or more episodes of arterial thrombosis, venous thrombosis (except for superficial venous thrombosis, SVT), or capillary thrombosis within any tissue or organ, confirmed by imaging, Doppler, or histological examination; (2) obstetric failure, i.e., at least one death of a morphologically normal fetus after the 10th week of pregnancy, or at least one premature birth of the morphologically normal fetus before the 34th week of pregnancy due to preeclampsia, eclampsia, or severe placental insufficiency, or at least three spontaneous miscarriages before the 10th week of pregnancy with the exclusion of anatomical causes and hormonal disorders in the mother, and chromosomal disorders in both parents; (3) laboratory criteria-the presence of lupus anticoagulant in plasma detected on at least two occasions minimum 12 weeks apart, using methods recommended by the International Society on Thrombosis and Haemostasis; mean or high levels of IgG or IgM class anticardiolipin antibodies (>40 GPL or MPL or <99th percentile) detected on at least two occasions minimum 12 weeks apart by a standardized ELISA method; anti-β2-glycoprotein I antibodies in serum or plasma (>99th percentile) detected on at least two occasions minimum 12 weeks apart.

2
The group with preeclampsia (PE) 48 women and their 43 partners The criteria for a diagnosis of severe preeclampsia were [11,12]: (1) Increased blood pressure and proteinuria after the end of the 20th week of pregnancy (with the exception of gestational trophoblastic disease (GTD) and multifetal pregnancy); (2) the lack of proteinuria if the following occurred for the first time after the 20th week of pregnancy: Thrombocytopenia (a blood platelet count <100,000/µL), liver disease (two-fold higher transaminase activity than normal), impaired renal function (level of creatinine >1.1 mg/dL or a two-fold increase in the creatinine level without kidney disease in the patient's medical history), pulmonary edema, central nervous system disorders, or vision disorders. The mean BMI of the pregnant women with preeclampsia was 28.1 (25.6-37.1); the mean birth weight of the infants was 2838.7 ± 514 g; 11 pregnant women were diagnosed with chronic hypertension. The patients after in vitro fertilization were not included in the study.

3
The group with intrauterine growth restriction (IUGR)

women and their 34 partners
In this group, a fetal weight was below the 10th percentile for the gestational age according to the first ultrasound performed in pregnancy, and the causes of intrauterine growth restriction-such as genetic determinants, diabetes, hypertension, preeclampsia, infections, renal disease, fetal developmental malformations, smoking, alcohol consumption, uterine abnormalities, taking medicines, autoimmune disease, etc.-were excluded.

4
The group with recurrent spontaneous abortion (RSA) 58 women and their 58 partners The women in this group had at least three spontaneous miscarriages in the first trimester of pregnancy; other causes of miscarriages were excluded. In all pairs, a normal karyotype was confirmed and health problems, such as diabetes, thyroid and adrenal gland disease, anatomic abnormalities, TORCH infections (toxoplasmosis, other (syphilis, HIV, varicella), rubella, cytomegaloviral disease, herpes), other infections, and autoimmune disease (the presence of anticardiolipin antibodies, lupus anticoagulant, and antinuclear antibodies) were excluded.

5
The control group (C)

healthy women and their 86 partners
The women in this group had no significant perinatal history, had given birth at least twice, and their pregnancies were uncomplicated. In these women, embolic/thrombotic complications and concomitant autoimmune diseases were excluded.

Isolation of Genome DNA from Peripheral Blood Leukocytes
DNA was isolated using the detergent method. 10-mL peripheral blood samples were collected from the patients, each to be used with 1 mL of 10% edetate disodium. 20 mL of TKM buffer (10 mM Tris-HC1, 10 mM KCI, 2 mM EDTA, 4 mM Mgc1 2) and the detergent IGEPAL (Sigma) were added to degrade cell membranes. Next, the samples were spun for 10 min (3400 rotations per minute, 12 • C). 30 mL and 20 mL of TKM were added one by one to the obtained leukocyte sediment, and spun again. To break the sediment and obtain homogeneous suspension, 2 mL of TKM was added. 0.5 mL of 10% SDS (sodium dodecyl sulphate) was added to the obtained pure leukocyte sediment. Then, it was incubated in a water bath at 60 • C for 7 min until the cell membranes and protein junctions were broken. The protein was salted out with 1 mL of 5 M NaCl solution. The mixture was shaken until homogeneous emulsion was achieved and then spun (9500 rotations per minute, 12 • C). 5 mL of 96% ethanol was added to the obtained supernatant to precipitate the DNA. As the final step, the DNA was rinsed out with 1 mL of ethanol, and dried in a vacuum centrifuge for 10 min. The last stage involved adding 3-4 drops of the TE buffer (10 mM Tris-HCl, 1 mM EDTA). The isolated DNA was diluted to the concentration of 50 ng/µL.
The amplicons were put on a 100× Microcon column (Amicon), which was placed in an Eppendorf test tube. 400 µL of distilled water was added and the whole thing was spun (1850 G, 15 min). The filtrate was poured out, and the reaction products left on the filter were poured over with 400 µL of water and spun again. This cycle was repeated thrice. To retrieve purified PCR products, the column was put into a new test tube and spun (9000 G, 10 min). The product was diluted with 20 µL of water. Asymmetric PCR was performed in an automatic thermocycler Gene Amp PCR System 9600 (Perkin Elmer) in a reaction mixture containing 4 µL of a purified PCR product, 30  The amplicons were precipitated with 96% ethanol, rinsed with 70% ethanol, and dried in a vacuum apparatus Eppendorf Concentrator 5301. Next, they were dissolved in 4 µL of loading buffer (150 µL of deionized formamide, 50 µL of 50 mM EDTA, 0.05% Dextran Blue). The samples were denatured for 4 min at 94 • C, put into an ice water bath, and placed on a denaturing polyacrylamide sequence gel (6% acrylamide-19:1, 1 × TBE, 8 M urea). Electrophoretic separation was performed using a 373A DNA Sequencer (Applied Biosystems). Electrophoresis data collection and analysis were performed using the 373 Data Collection and the 373 Analysis software (Applied Biosystems).
The HLA-G alleles observed in our study were determined by comparison of the base pair sequences in exons 2, 3, and 4 and the base pair sequences in the HLA-G alleles published on the website of Nolan Institute [14].

Analysis of the 14-bp ins/del Polymorphism in the 3 -Untranslated Region of the HLA-G Gene
The polymorphic region was amplified by polymerase chain reaction (PCR). The PCR amplicons were resolved on 4% agarose gel. The insertion allele was 14 nucleotides longer than the deletion allele. The PCR reaction was performed in a thermocycler (GeneAmp PCR System 9700, Applied Biosystems). The following primers were used in the PCR reaction:

Statistical Analysis
Statistical analysis was performed using Statistica v. 17.0. The chi-square test was employed to verify whether the genotype frequencies fit the Hardy-Weinberg equilibrium (H-W). The discrete variables were assessed by Pearson's chi-square test. The logistic regression model was used to estimate the risk of antiphospholipid syndrome, preeclampsia, intrauterine growth restriction, and recurrent spontaneous abortion for particular alleles of the 14-bp ins/del polymorphism in the 3 -UTR of the HLA-G gene. The results were described in terms of a relative odds ratio (ROR) and 95% confidence interval (CI); the probability was calculated using Fisher's two-sided exact test. The level of significance was set at p ≤ 0.05.

Results
In our study, 16 different allele variants of the HLA-G gene were observed. In the studied population, the most common alleles were 10101 (47.48%) and 10102 (25.30%). The least common alleles in the Polish population were 10104, 10109, and 10402 (0.09% each) ( Table 3). We analyzed the HLA-G allele frequencies in the parental pairs with antiphospholipid syndrome, preeclampsia, intrauterine growth restriction, and recurrent spontaneous abortion and in the control group (Table 4). The most common HLA-G allele in the groups of women with antiphospholipid syndrome, preeclampsia, intrauterine growth restriction, and recurrent spontaneous abortion was the HLA-G 10101 allele found in 35.71%, 44.79%, 57.35%, and 52.59% of the women, respectively. In the control group, this allele was observed in 48.31% of the women.
Analysis of the HLA-G allele frequencies in the men-partners of the women with pregnancy complicated by antiphospholipid syndrome, preeclampsia, intrauterine growth restriction, and recurrent spontaneous abortion (Table 3)-demonstrated that the HLA-G 10101 allele was the most common and was found in 45.37%, 47.67%, 47.14%, and 47.41% of the men, respectively, and in 51.74% of the men in the control group.
We also calculated an odds ratio (OR) for a complicated pregnancy depending on the HLA-G allele ( Table 5). A significantly higher risk of the antiphospholipid syndrome in the women was associated with the HLA-G 10108 allele (OR = 3.24; p = 0.042), while the HLA-G 10101 (OR = 0.59; p = 0.030) and the HLA-G 10106 (OR = 0.21; p = 0.037) alleles reduced that risk. The women with the HLA-G 10108 allele (OR = 5.91; p = 0.001) had almost a six-fold higher risk of preeclampsia than the control group. The HLA-G alleles in the women did not increase the risk of intrauterine growth restriction or recurrent spontaneous abortion. There was no relationship between the presence of the studied alleles in men and the risk of antiphospholipid syndrome, preeclampsia, intrauterine growth restriction, or recurrent spontaneous abortion in their partners.
The frequency of the HLA-G 14-bp ins/del polymorphism (women p > 0.101, men p > 0.336) in the control group met the Hardy-Weinberg equilibrium criteria. Table 6 shows the frequencies of the 14-bp ins/del polymorphism in the 3 -UTR of the HLA-G gene in women with antiphospholipid syndrome, preeclampsia, intrauterine growth restriction, and recurrent spontaneous abortion and their partners and in the control group. There were no statistically significant differences in the frequencies of the 14-bp ins/del polymorphism in the 3 -UTR of the HLA-G gene between the groups (p > 0.05) (Fisher's two-sided exact test). Table 4. The HLA-G allele frequencies in the women and men with antiphospholipid syndrome (APS), preeclampsia (PE), intrauterine growth restriction (IUGR), recurrent spontaneous abortion (RSA), and the control group (C).  Total  140  100  108  100  96  100  86  100  68  100  70  100  116  100  116  100  178  100  172  100  Table 5. The HLA-G allele frequencies in the women with antiphospholipid syndrome (APS), preeclampsia (PE), intrauterine growth restriction (IUGR), recurrent spontaneous abortion (RSA), and their partners. The risk of APS, PE, IUGR, and RSA in these women and in the control group (C).  Table 6. The frequencies of the 14-bp ins/del polymorphism in the 3 -UTR of the HLA-G gene in the women with antiphospholipid syndrome, preeclampsia, intrauterine growth restriction, recurrent spontaneous abortion and their partners and in the control group. The lack of statistical significance (p > 0.05) was also noted in the analysis of the frequencies of the common HLA-G alleles and the common alleles of the 14-bp ins/del polymorphism in the 3 -UTR of the HLA-G gene in the parental pairs with a complicated pregnancy (Table 7). Table 7. The frequencies of the HLA-G common alleles and the 14-bp ins/del polymorphism in the 3 -UTR of the HLA-G gene in the parental pairs with a pregnancy complicated by antiphospholipid syndrome, preeclampsia, intrauterine growth restriction, or recurrent spontaneous abortion and in the control group.

Allele
Allele

Discussion
HLA-G is critical for effective implantation in pregnancy. Increasingly numerous scientific reports show that HLA-G plays an essential role in fetal tolerance through the mechanism of inhibition of the cytotoxic function of T and NK cells. The literature provides evidence that the polymorphic variants of the HLA-G antigen involve complications in pregnancy, such as recurrent spontaneous abortion and preeclampsia [15][16][17][18]. Nevertheless, the available studies compare various molecular changes in different populations and the reported data are often ambiguous.
In the pathogenesis of preeclampsia, the contribution of a wide range of genes interacting with environmental factors is analyzed. A high BMI, African American and South Asian races, chronic hypertension, pregnancy after ovulation induction therapy, a family or obstetric history of preeclampsia, as well as chronic hypertension and diabetes preceding pregnancy are factors that predispose to preeclampsia. It is suggested that genes associated with the function of the vascular endothelium, the function of the placenta, as well as blood coagulation and fibrinolysis may play a role [19,20].
Our investigation indicates that the risk of preeclampsia in female carriers of the HLA-G 10108 allele is almost six-fold higher than in the control group. However, our analysis of the frequencies of the 14-bp ins/del polymorphism in the 3 -UTR of the HLA-G gene in particular groups have not demonstrated statistically significant differences compared with the control group. It is worth mentioning that in their study, Sipak-Szmigiel et al. [21] observed that an increased risk of failure in pregnancy (OR = 5.86; 95% CI; 0.93-21.24; p = 0.09) in the experimental group was associated with the frequency of the HLA-G 10108 allele. Similar research was conducted by Abbas et al. [22], who analyzed the HLA-G polymorphisms in a group of 120 women with recurrent spontaneous abortion and 120 women with uncomplicated pregnancy in India. As the authors reported, the HLA-G 010103 allele was more common in the women with recurrent spontaneous abortion and the HLA-G 10105 allele was present in 1.7% of them. The HLA-G 010108 allele was observed in 0.4% of the women with recurrent spontaneous abortion and was not found at all in the women with uncomplicated pregnancy. The frequency of the HLA-G 0105N allele in the Indian population was quite high (13.8%) compared with other worldwide populations. A similar study was carried out by Durmanova et al. [23], who analyzed 123 women with preeclampsia and 102 women with an uncomplicated pregnancy to assess potential links between the selected HLA-G polymorphisms and the risk of preeclampsia. Their results did not reveal any association between the HLA-G 14-bp polymorphism and the risk of preeclampsia. A similar observation was made by Rokhafrooz et al. [24], who performed genotyping of the 14-bp ins/del polymorphism in exon 8 of the HLA-G gene in 150 healthy pregnant women and 150 women with preeclampsia. In the study of Ferreira et al. [ [26], which was later confirmed in Caucasians in the study by Hylenius et al. [27]. Moreau et al. [28] and Tan et al. [29] also reported a significant association between the HLA-G*0106 allele and the risk of preeclampsia. Interestingly, the HLA-G*0106 allele includes the 14-bp HLA-G polymorphism. Yie et al. have reported a significant association between the SNP at position 1754 in exon 8 in the 3 -UTR of the HLA-G gene and the risk of preeclampsia [30]. In our study, the HLA-G 106 allele (1.04%) was rare among women with previous episodes of preeclampsia and their risk of preeclampsia was low and statistically insignificant.
Our study did not confirm an increased risk of recurrent spontaneous abortion in the parental pairs with the HLA-G alleles. Kalotra et al. [8] analyzed the HLA-G 14-bp ins/del polymorphism in over 200 women with recurrent miscarriages (two or more miscarriages) and their partners, comparing it with the control group. They did not note significant differences in the frequencies of the HLA-G 14-bp ins/del polymorphism and the studied genotypes between the groups. The study of Hashemi et al. [31] provided evidence that the HLA-G 14-bp ins/del polymorphisms may contribute to the risk of recurrent spontaneous abortion. Nonetheless, our study did not confirm this relationship in the Polish population.
Another look at this issue was presented by Koc et al. in 2018 [32], who assessed the connection between the risk of miscarriage and the HLA-G polymorphism in the fetus. Similarly to us, they found that the most common HLA-G type, both in the study group (30.3%) and the control group (47%), was HLA-G 10101. They also reported that the HLA-G 10104 allele was visibly related with miscarriage (p = 0.007), and the 14-bp deletion in the 3 -UTR of the HLA-G gene was more frequent in the group with miscarriages. The difference was not statistically significant.
Also, SNP mutations in the HLA-G promoter region are important from the point of view of maintaining pregnancy. Research in this field was conducted by Yazdani et al. [33], who investigated the -1573T> C and -1746C> A SNP mutations in the HLA-G promoter region, associated with recurrent spontaneous abortion. Their haplotype analysis revealed a significant link between recurrent miscarriages and the H1 (ATCCAGGTACGCAA) and H2 (CTTCGAGAACGCAG) haplotypes.
In our study, the higher risk of antiphospholipid syndrome for the women was accompanied by the presence of the HLA-G 10108 allele, while the lower risk was related to the HLA-G 10101 allele. The literature of the subject offers few publications describing the relationship between the HLA-G polymorphisms and antiphospholipid syndrome. Research on the contribution of the HLA-G 10101, HLA-G 10108, and HLA-G 10106 alleles to the risk of complications in pregnancy in the Polish population requires further investigation.

Conclusions
Our study allows the conclusion that the HLA-G 10108 allele may predispose women to antiphospholipid syndrome and preeclampsia. The HLA-G 10101 and the HLA-G 10106 alleles, on the other hand, may reduce the risk of antiphospholipid syndrome. The 14-bp ins/del polymorphism in the 3 -UTR of the HLA-G gene has most probably no effect on the risk of complications in pregnancy. The presence of the same HLA-G alleles in both partners probably does not modify the risk of complicated pregnancy.

Limitations
One of the limitations of our study is that we did not control for risk factors of the outcome. However, we decided it was not necessary as our aim was to estimate the magnitude of causal effects. We would only have needed to control for risk factors if there was a systematic causal reason for the risk factor to be associated with the exposure variable in the underlying target population, which was not the case here. The risk factors for the pregnancy outcomes in our study were not a "cause" of the genotype. Those risk factors that might be unbalanced in exposure groups within the "sample" will be accounted for in the confidence intervals. They produce random error, not systematic error. A similar strategy has been described in the study of Hernán et al. [34], who claimed that analyses of the effects of prenatal exposures frequently adjust for variables, such as maternal weight gain during pregnancy, gestational age, or birth weight, that are likely to be affected by either the exposure or the outcome. The decision to adjust is usually based on statistical criteria only.
We also need to say that the almost six-fold increased odds of preeclampsia in female carriers is estimated with very low precision. Our 95% CI for that estimate is 2.1-16.3. As such, our standard errors are quite large. However, an odds ratio of <2.5 in the underlying "target" population is consistent with our findings. A large OR results from the lack of the studied alleles in particular groups.