Association of Killer Immunoglobulin-Like Receptor and Human Leukocyte Antigen Class I Ligand with Recurrent Abortion in Saudi Women

Introduction

Natural killer (NK) cells are an important component of the innate immune response, and they are involved in adaptive immunity (Smyth et al., 2002). Their effector functions include cytolytic activity against target cells and secretion of an inflammatory and regulatory panel of cytokines (Guillerey et al., 2016). NK cells express a diverse family of receptors (Makrigiannis and Anderson, 2003). Prominent among these are the killer cell immunoglobulin-like receptor (KIR) family that includes members that recognize and interact with human leukocyte antigen class I (HLA).

A total of 17 KIR genes and pseudogenes have been identified and are clustered in chromosome 19q13.4 within the leukocyte receptor complex (Wende et al., 1999). The KIR gene complex exhibits substantial diversity that affects the presence or absence of individual KIR genes, forming the haplotype (Middleton and Gonzelez, 2010; Vendelbosch et al., 2015). Eight of these 17 Kir genes encode inhibitory KIRs (2DL1, 2DL2, 2DL3, 2DL5A, 2DL5B 3DL1, 3DL2, and 3DL3), 7 encode activating KIRs (2DL4, 2DS1, 2DS2, 2DS3, 2DS4, 2DS5, 2DS5, and 3DS1), and 2 are pseudogenes (2DP1 and 3DP1) (Long et al., 1997; Middleton and Gonzelez, 2010).

Substantial allelic polymorphism has been reported for different KIR genes (Middleton et al., 2007). Many different haplotypes have been reported, depending on the occurrence of KIR genes in an individual. These haplotypes are classified into the A and B haplotypes. Group A haplotypes contain a fixed number of genes encoding inhibitory genes and only one activating gene (KIR2DS4). Group B haplotype have variable gene contents and different activating KIRs, and the group is characterized by the presence of the KIR2DL2, KIR2DS1, KIR2DS2, KIR2DS3, and KIR2DS5 genes (Uhrberg et al., 2002; Uhrberg, 2005).

The ability of NK cells to recognize and eliminate foreign cells mainly results from the interaction of their KIRs with major histocompatibility complex class I molecules (MHC-I) expressed on the surface of target cells. Although >3000 alleles of HLA-C have been reported, they are recognized by KIRs as two distinct groups, HLA-C1 and HLA-C2, which are distinguished by the amino acid dimorphism at position 80. Several genetic studies have reported strong associations of KIRs and their ligands with the outcome of large number of infectious, autoimmune, or cancer diseases (Khakoo, 2010). In the Saudi population, associations between KIRs and some cancer and autoimmune diseases have recently been reported (Xiong et al., 2013; Al Omar et al., 2015; Osman et al., 2016; Alomar et al., 2017).

Recurrent spontaneous abortion (RSA) is a component of the range of reproductive failure, defined as two or more consecutive pregnancy losses before the 20th week of gestation in women of reproductive age (Rai and Regan, 2006; Branch et al., 2010). RSA is considered an important complication during pregnancy and represents a common frustrating problem in gynecology (Jauniaux et al., 2006). RSA has been reported in 1-2% of reproductive-aged women. The reasons for most cases of RSA (40-55%) remain unexplained (Li et al., 2002; Van den Berg et al., 2014). Different epidemiological studies have considered RSA as a multifactorial problem (Ferrara et al., 1996; Daher et al., 2012). Different etiologic factors have been implicated in this problem, such as endocrine regulation, autoimmune reactions, psychological characteristics, thrombophilia, viral infections, environmental factors, and genetic background (Li et al., 2002; Quenby et al., 2005).

In fact, the immunologic relationship between the mother and fetus is determined by fetal antigens and the maternal immune system. The inadequate recognition of fetal antigens might result in failed pregnancy (Vince and Johnson, 1995; Christiansen, 1996). NK cells have been reported to constitute the predominant leukocyte population present in the endometrium at the time of implantation and in early pregnancy (Hiby et al., 2004; Moffett et al., 2004; Male et al., 2011). Different studies have demonstrated the crucial role that NK cells play in the success versus failure of pregnancy during the first weeks. In the fetal-maternal interface, activity of NK cells is regulated through the interaction of the maternal KIRs with the paternal HLA-C ligand, which is the only polymorphic MHC molecule expressed by the invasive extravillous trophoblast (Hiby et al., 2010a; Male et al., 2011). Thus, the polymorphisms that characterize the KIRs and HLA-C ligand may affect the pregnancy outcome and could explain in part the reported relationship between the occurrence of some pregnancy disorders, including RSA, pre-eclampsia, and defective placentation, and some KIR-HLA-C combinations in different populations (Varla-Leftherioti et al., 2003; Wang et al., 2007; Hong et al., 2008; Nowak et al., 2009; Hiby et al., 2010a; Chazara et al., 2011; Faridi and Agrawal, 2011; Ozturk et al., 2012; Babor et al., 2014; Moffett and Colucci, 2015; Colucci, 2017; Wilczynska et al., 2019). In this study, we report the results of our investigation of the association between the maternal KIR and HLA-C genes and the occurrence of RSA in Saudi women.

Materials and Methods

Genomic DNA extraction and KIR and HLA-C typing

Genomic DNA was extracted from peripheral blood using a DNeasy Blood & Tissue Kit (Qiagen, Valencia, CA) according to the manufacturer's instructions. Genotyping of KIRs was performed using a KIR polymerase chain reaction (PCR)-sequence-specific primer commercial typing kit (Miltenyi Biotec, Inc., Cologne, Germany) according to the manufacturer's instructions. This kit consists of a panel of locus-specific oligonucleotide primers allowing individual amplification of 17 KIR genes (15 genes and 2 pseudogenes) and the common variants of 2DS4 (Fig. 1).

OPEN IN VIEWER

FIG. 1.

Gel electrophoresis of KIR genotyping using PCR-SSP method. An internal positive control appears at 400 bp. The primers used for 2DL5 all embrace two distinct, but closely related genes named 2DL5A and 2DL5B. Some alleles of 2DL5A are detected only with primers used for 2DL5 all. In this typed sample, individual is considered 2DL5A⁺ and 2DL5B⁻. KIR, killer immunoglobulin-like receptor; PCR-SSP, polymerase chain reaction-sequence-specific primers.

For HLA-C1 and HLA-C2 typing, the same primers were used as those reported by Tajik et al. (2010). For each reaction, PCR was performed as reported by Omar et al. (2016). All PCRs were run in a Techne thermocycler apparatus (Techne TC-Plus Satellite). The PCR products were analyzed by electrophoresis in 2% agarose gels stained with ethidium bromide and visualized on a UV transilluminator using a gel documentation system (BioRad Gel225 Doc™XR+) to check for the presence or absence of gene-specific amplicons.

Results

In this case-control study, we examined the distribution of KIR genotypes and the HLA-C alotype ligands among women with at least three unexplained RSAs and healthy women with at least two live births and without a history of RSA. Clinical and demographic characteristics of the RSA and control women are shown in Table 1.

Table 2 presents the comparative distribution of the percentages of the 17 KIR genes for the two groups. In this study, all KIR genes were detected in both the RSA and control groups. The framework genes were present in all of the healthy control women. Among the RSA group, KIR2DL4 genes was absent in three individuals. 3DL3 and 3DP1 were recorded in all individuals. Significantly lower frequencies for the 2DL5A, 2DS2, and 2DS3 genes were recorded in the RSA group compared with the control group (OR = 0.16, 0.178, and 0.41, respectively, with p-values <0.05). However, for the 2DS3 gene, the difference between the two groups was no longer significant after Bonferroni correction. The frequency of 3DS1 was also lower in the RSA group, but the difference did not reach statistical significance (p = 0.059). Regarding the AA and BB KIR haplotypes, no significant difference was observed in their distribution between the RSA and control groups. Moreover, no significant difference was observed in the frequencies of the two centromeric (Cen-A and Cen-B) and telomeric (Tel-A and Tel-B) subhaplotypes (Table 2).

However, when considering the 2DS2 and 2DL5A genes, the occurrence of both genes in the B haplotypes contributed highly significantly to protection against RSA (OR = 0.091; p < 0.001). Inversely, the lack of both 2DS2 and 2LA5A was associated with an increase in the occurrence of spontaneous abortion within the Saudi population (OR = 5.72; p = 0.003). Different comparisons were performed for the distribution of the number of activating genes between the RSA group and control group, and we did not observe differences between the groups in the study population (Table 3). However, the presence of 2DS2 conferred significant protection, regardless of the number of activating genes.

The frequencies of the HLA-C1 and HLA-C2 allotypes and genotypes were not different between the RSA group and healthy control group (Table 4).

The results of the analysis of the combined association of the KIR genes and HLA-C ligand with RSA are reported in Table 5. The 2DL1 and 2DS2 genes were found to be associated for some combinations with the HLA-C ligand. Thus, a protective effect was observed for 2DL1 in the absence of its C2 ligand (OR = 0.085, p = 0.008). However, 2DS2 exhibited a protective effect when associated with its HLA-C1 ligand in the homozygote (OR = 0.27, p = 0.028) and heterozygote (HLA-C1C2) forms (OR = 0.38, p = 0.03). The distribution of 2DL2/3 and the HLA-C1 ligand did not show any statistically significant difference.

Discussion

This study is the first investigation on the association between KIR gene polymorphism and recurrent abortion in Arabian ethnic group. We confirm, through these results, the relationship between some individual KIRs and some of its combinations with the HLA-C maternal ligand and the success versus failure of pregnancy in a Saudi population. We showed herein the frequent appearance of the three KIR genes, 2DL5A, 2DS2, and 2DS3, among the healthy control group compared with the RSA group. The highest, and significant, association with protection against failure of pregnancy was observed with the 2DL5A and 2DS2 genes. These two genes are considered to belong to the B haplotype and have different positions in the KIR B haplotypes. The 2DL5A gene has a telomeric position, whereas 2DS2 is positioned at a centromeric region. Interestingly, the 2DS2 molecule interacts with the C1 ligand. The presence of both genes correlates with reduced RSA (10-fold lower rate of occurrence) possibly by supporting successful pregnancy. Inversely, a lack of both genes is associated with a more than fivefold higher rate of RSA occurrence. This result indicates a possible dose effect of the KIR genes on the occurrence of RSA.

It is worthy to notice that the biological significance of KIRs in vivo depends on whether these receptors are present in individuals simultaneously with their ligands. A combinatory analysis confirmed that the KIR2DS2 gene has a strong protective effect with its specific HLA-C1 allele both in homozygote and heterozygote forms. Moreover, this protection was no longer effective in the absence of its ligand. In this context of the KIR-HLA-C interaction effect, the activating KIR2DL1 exhibited a protective effect when its HLA-C2 ligand was missing.

Until now, few studies had been performed on the association between KIRs and spontaneous abortion. Except for the study by Witt et al. (2004), which did not report any significant association of KIR genes or genotypes with spontaneous abortion in 51 Brazilian Caucasian women, most of the studies reported conflicting results. A protective association of some activating KIRs of the telomeric B haplotype region, particularly 2DS1 as well as 3DS1, 2DS5, and 2DL5, with RSA was reported in an English Caucasian population composed of 95 RSA cases and 269 controls (Moffett and Colucci, 2015). Moreover, in the same study population, the AA haplotype was found to be positively associated with recurrent miscarriage (Hiby et al., 2010b; Moffett and Colucci, 2015). Faridi et al. (2009) reported a protective effect of the KIR AA haplotype and inhibitory genes in an Asian Indian population composed of 205 women with a history of RSA and 224 controls belonging to different ethnic groups. In this study, the authors reported a probable protective effect of the inhibitory genes 2DL2 and 3DL1 and a predisposition effect of the activating genes 2DS2 and 2DS3. Other conflicting results were also reported by Hong et al. (2008), showing an association of 2DL2 (a B haplotype gene) with RSA in a very small group of Chinese women composed of 16 women with a history of RSA and 24 controls and without correction of the results. This result was not confirmed by another Chinese study that included higher numbers of individuals (73 RSA couples and 68 control couples), as it showed an increase in the activating KIR genes 2DS1 and 2DS5 and an association of RSA with higher numbers of activating KIRs (Wang et al., 2007). Similarly, in 68 Brazilian Caucasian RSA patients, genotypes with five or six activating KIR genes were significantly more frequent than in 68 controls (Vargas et al., 2009). Although no single KIR gene reached significance for its frequency distribution in this study Varla-Leftherioti et al. (2003) reported a lack of appropriate inhibitory KIRs in RM women from Greece in a study that included 15 patients with a history of RSA. Likewise, the AA genotype was reported to have a high frequency compared with a low frequency of B genotypes in 40 Anatolian women of European descent with a history of RSA who settled on the Mediterranean coast (Ozturk et al., 2012). More recently, a study conducted by Ay et al. (2019) in a women population from Turkey having unexplained recurrent pregnancy loss (URPL: before or after birth) reported strong positive association of eight KIR genes (2DL1, 2DL2, 2DL3, 2DL4, 2DS1, 2DS2, 2DS4, and 2DS5) with loss of pregnancy. Surprisingly in this study, the framework 2DL4 gene is not frequent in control group (21.4%) versus case group (68%), which was never reported in any studied population. In this study, authors did not consider the 2DL5 gene and failed to amplify the 2DS3 genes in both case and groups (Ay et al., 2019). These conflicting results are a subject of controversy and debate regarding the method of typing and the matching of patients and controls according to ethnic origin as well as exclusion criteria/inclusion in the choice of patients and controls (Faridi and Agrawal, 2009; Hiby et al., 2010b). Indeed, in almost all studies, the patients and controls were not selected according to the same criteria regarding the number of spontaneous abortions in the patient group and the number of pregnancies in the control group.

Regarding the maternal HLA-C1 and HLA-C2 groups, as reported in almost all previous studies, we did not observe significant differences between the RSA and control groups. However, controversial associations have been reported between paternal C1 and C2 allotypes of the fetus and recurrent abortion in some studied populations. The paternal C2 allotype, which is the cognate ligand of the 2DS1 gene, but not the maternal C2 allotype, was positively associated with recurrent abortion in English Caucasian affected women, especially when the mother carried the A haplotype (Moffett and Colucci, 2015). In an Indian population, in contrast to our results and those reported by Faridi and Agrawal (2011) and Hiby et al. (2010b) reported that KIR2DL1 was significantly higher in control couples when both parents were homozygous for HLA-C2 and that the activating KIR2DS2 was significantly higher in patients when both parents were homozygous for HLA C1. These authors hypothesized that RSA was related to excessive activation or the lack of appropriate inhibition of NK cells. However, recent functional analyses have provided evidence that uterine NK (uNK) cells are phenotypically and functionally different from peripheral NK (pNK) cells (Nishikawa et al., 1991; Park et al., 2010). Thus, it is obvious that the primary function of the immune system and decidual NK cells in pregnancy is not killing but angiogenesis and ensuring good vascularization of the placenta by allowing good blood flow, which is necessary to maintain placental growth (Xie et al., 2005; Arcuri et al., 2006; Kumar and Medhi, 2008; Hofmann et al., 2014; Fraser et al., 2015). These uNK are characterized by their high capacity to produce a wide range of growth factors, cytokines, and chemokines, which are beneficial for trophoblast migration and the trophoblast-mediated remodeling of uterine vessels. Recently, functional studies have shown the important role played by activating genes such as 2DS1 and 2DS4 in the success of pregnancy. These two genes as well as the inhibitory 2DL1 gene were reported to be overexpressed in decidual NK cells at the beginning of pregnancy (Xiong et al., 2013). In the absence of KIR2DS1, the insufficient activation of decidual NK cells results in poor trophoblast invasion, placental stress, and growth restriction of the fetus (Chazara et al., 2011; Xiong et al., 2013).

These functional studies support the fact that signals delivered by the inhibitory or the activating KIRs in decidua are not always relatively cytotoxic as they are in their action against foreign transplants but also induce the secretion of cytokines and chemokines necessary for the development of the placenta and fetus. Although Xiong et al. (2013) reported a lack of expression of KIR2DS1 in uNK cells, no study has been performed on the KIR2DS2 gene. In conclusion, we believe that further investigations involving subjects with different ethnic origins would aid in understanding this relationship between the uNk phenotype and the evolution of gestations by taking into account genetic diversity.