HLA-G is a nonclassical HLA class I molecule that plays a pivotal role in immune regulation, particularly in pregnancy. HLA-G exerts its regulatory function by engaging inhibitory receptors such as ILT2, ILT4, and KIR2DL4 present on NK cells, T cells, and antigen-presenting cells. These interactions dampen cytotoxic responses, influence cytokine release, and foster a tolerogenic immune environment critical for maternal–fetal tolerance. Unlike classical HLA class I molecules (HLA-A, -B, -C), HLA-G exhibits limited polymorphism, reflecting its conserved immunological function. As of September 2025, 194 HLA-G alleles have been identified, of which 61 encode functional proteins and 6 are null alleles (Robinson et al., 2024). Structurally, the HLA-G gene includes eight exons and seven introns. Exons 2–4 encode the extracellular domains involved in antigen presentation and receptor binding, while exon 8 forms the 3′untranslated region (3′UTR), crucial for regulating mRNA stability and translation. HLA-G typically forms a non-covalent complex with β2-microglobulin, essential for its stability and function.

At the maternal–fetal interface, HLA-G binds to inhibitory receptors on immune cells ILT2 and ILT4 (monocytes, dendritic cells, T cells), and KIR2DL4 (uterine NK cells) (Figure 1). This interaction suppresses cytotoxicity and promotes immune tolerance. Additionally, HLA-G stimulates anti-inflammatory cytokines (e.g., IL-10, TGF-β) and angiogenic factors (e.g., Vascular Endothelial Growth Factor (VEGF)), facilitating placental vascularization and spiral artery remodeling. Notably, HLA-G has been linked to senescence-like changes in trophoblast and endothelial cells, aiding early pregnancy adaptations (LeMaoult and Yan, 2021). Single-cell analyses of decidual tissue reveal dynamic regulation of HLA-G across trophoblast subtypes. Reduced HLA-G expression is associated with adverse outcomes, including PE, RIF, and miscarriage (Zhou et al., 2019). Thus, HLA-G serves as a key mediator of immune tolerance, vascular remodeling, and successful implantation.

Soluble forms of HLA-G (sHLA-G), mainly the HLA-G5 and HLA-G6 isoforms, are secreted into extracellular fluids and play critical roles in early embryo–maternal communication. These isoforms are detectable in follicular fluid, embryo culture media, maternal serum, and uterine secretions, and are believed to promote implantation and maternal immune adaptation during assisted reproduction.

Pioneering studies by Jurisicova and colleagues (1996) (Jurisicova et al., 1996) demonstrated sHLA-G expression from the 2-cell to blastocyst stage. Elevated sHLA-G levels in embryo culture media have since been linked to increased implantation rates and pregnancy success in IVF (Fuzzi et al., 2002). Clinical studies have shown that low maternal sHLA-G correlates with early pregnancy loss (Pfeiffer et al., 2000), while higher levels are associated with ongoing pregnancies. A meta-analysis of over 6,000 IVF cases confirmed that embryos secreting sHLA-G were significantly more likely to implant and result in clinical pregnancy (Niu et al., 2017). Furthermore, combining sHLA-G quantification with embryo morphology scores enhances predictive accuracy (Kotze et al., 2010; Kotze et al., 2013). sHLA-G predominantly localizes to the trophectoderm, supporting its role in embryo–endometrium interaction (Shaikly et al., 2008).

Despite these encouraging findings, the clinical utility of sHLA-G remains contested. Several studies investigating RIF failed to confirm consistent associations between sHLA-G levels and reproductive outcomes (Vani et al., 2021; Hu et al., 2022). These discrepancies may stem from differences in assay platforms, thresholds used, and sampling windows across studies. These disparities underscore a critical need for harmonized protocols before sHLA-G can be considered a reliable biomarker.

Importantly, the absence of sHLA-G may reflect intrinsic embryonic abnormalities rather than maternal immune defects (Criscuoli et al., 2005). New technologies, including Luminex-based assays and AI-driven embryo selection tools, are increasingly integrating sHLA-G data to enhance IVF personalization (Tang et al., 2024). Although promising, these innovations require international harmonization and large-scale validation to ensure their clinical utility. Despite ongoing challenges, sHLA-G remains a promising non-invasive biomarker for embryo competence and IVF success, pending further standardization and validation. Future studies should focus on validating optimal sHLA-G thresholds in diverse populations and IVF settings to support its implementation in routine clinical decision-making.

Genetic variation within HLA-G, particularly in the 3′UTR, critically affects its expression and reproductive outcomes. The most studied variant is the 14-base pair (bp) insertion/deletion (ins/del) polymorphism (rs371194629). The insertion allele is associated with reduced mRNA stability and lower sHLA-G levels, contributing to impaired immune tolerance and increased risk of implantation failure and miscarriage (Hviid et al., 2004; Costa et al., 2016). Studies across populations have linked the ins allele to recurrent IVF failure (Sipak-Szmigiel et al., 2009; Fan et al., 2017). Conversely, the del allele correlates with enhanced HLA-G expression and better reproductive outcomes. This suggests that host genetic background could be a major determinant of IVF success, yet allele frequency differences across populations may partly explain variability in findings.

Additional polymorphisms, such as rs1063320 (+3142C>G) and rs1632947 (−964G>A), influence mRNA decay and transcription, respectively. These variants, along with others like rs1233334 and rs9380142, have been variably linked to IVF success (Hu et al., 2022), though findings remain inconsistent across ethnicities and study designs. Such heterogeneity underscores the need for larger, ethnically diverse cohorts and standardized genotyping methods to draw robust conclusions. Notably, some polymorphisms initially associated with infertility (e.g., HLA-G*01:03) have shown different trends in broader populations, suggesting complex gene environment interactions (Costa et al., 2016). Interestingly, endometrial sHLA-G levels do not consistently differ by 14-bp genotype or RIF status (Papúchová et al., 2022), highlighting the dominant role of embryo-derived HLA-G. However, higher endometrial expression during the implantation window may support pregnancy success (Kofod et al., 2017). This observation implies that maternal genotype may not fully reflect the local immune-tolerant environment, reinforcing the importance of embryo competence in reproductive immunology. HLA-G haplotypes, such as UTR-4, which combines favorable polymorphisms, have been linked to shorter time to pregnancy (Nilsson et al., 2020). Still, prospective studies are needed to assess whether haplotype-based screening can meaningfully inform IVF strategies. Next-generation sequencing is advancing our understanding of HLA-G’s role in IVF, though clinical testing for polymorphisms remains investigational. Translating these genetic insights into clinical practice will require rigorous validation and cost-effectiveness evaluation.

KIR genes, highly polymorphic and categorized into A (inhibitory) and B (activating) haplotypes, significantly affect reproductive success.

For women with the KIR AA genotype, immunomodulatory interventions offer new therapeutic options in IVF.

HLA-E is a nonclassical HLA class I molecule first characterized by Koller et al., in 1988. Unlike the highly polymorphic classical HLA-Ia molecules (HLA-A, -B, -C), HLA-E exhibits limited genetic variability and a highly specialized immunological role. Structurally, HLA-E forms a heterodimer composed of a membrane-bound α heavy chain (∼45 kDa) and a β2-microglobulin light chain. Its gene consists of seven exons: exons 1–4 encode the extracellular domains (α1, α2, α3), exon 5 the transmembrane region, and exons six and 7 the cytoplasmic tail. Despite this genomic structure, exons 7 and 8 are typically excluded from the mature mRNA transcript, resulting in a shortened cytoplasmic tail (Arnaiz-Villena et al., 2022) (Figure 1).

HLA-E expression is widespread across tissues and tightly regulated at the cell surface, requiring peptides derived from the leader sequences of other HLA class I molecules, especially HLA-G, with which it shares strong functional synergy. These leader peptides stabilize HLA-E at the surface and facilitate its interactions with immune receptors. Due to its restricted peptide repertoire and conserved anchor residues, HLA-E’s function is more immunoregulatory than antigen-presenting (Walpole et al., 2010).

As of September 2025, the IPD-IMGT/HLA Database lists 385 HLA-E alleles, including 144 protein-coding and 10 null variants (Robinson et al., 2024). Among these, HLA-E*01:01 and HLA-E*01:03 are the two most prevalent alleles, found at roughly equal frequencies across ethnic groups, suggesting balancing selection that preserves both high- and low-expression variants (Grimsley and Ober, 1997; Tamouza et al., 2007). Immunologically, HLA-E exerts dual roles in innate and adaptive immunity.

In innate immunity, HLA-E binds to CD94/NKG2A inhibitory receptors on NK cells, suppressing cytotoxic activity and promoting immune tolerance, especially in immune-privileged sites like the maternal–fetal interface.

Additionally, HLA-E interacts with a range of immunomodulatory receptors, including ILT2 and ILT4, as well as activating NKG2C and NKG2E on NK and T cells (Allan et al., 2002; Kaiser et al., 2005). These receptor engagements modulate cellular responses in contexts such as infection, cancer, transplantation, and pregnancy, positioning HLA-E as a critical mediator of immune homeostasis.

Although less extensively studied than HLA-G, HLA-E has emerged as a relevant immunogenetic factor in reproductive medicine, particularly in the context of IVF and RIF.

HLA-F is a nonclassical HLA class I molecule first identified by Geraghty et al., in 1990. Structurally similar to classical HLA class I molecules, HLA-F comprises a ∼40–41 kDa α heavy chain paired with β2-microglobulin. However, HLA-F is distinct due to its alternative splicing, which often excludes exon 7, leading to a truncated cytoplasmic tail. This structural variation may affect intracellular trafficking and receptor signaling (Geraghty et al., 1990; Boyle et al., 2006) (Figure 1).

As of September 2025, the IPD-IMGT/HLA Database reports 126 HLA-F alleles, including 28 protein-coding and three non-functional (null) variants (Robinson et al., 2024), underscoring the relatively low polymorphism characteristic of nonclassical HLA genes.

HLA-F is expressed in a variety of immune cells including B cells, T cells, NK cells, and peripheral blood lymphocytes—and is inducible in activated immune environments. Its expression is predominantly intracellular under resting conditions, but surface expression is observed upon cell activation, suggesting immune-state–dependent regulation (Lee et al., 2010). In the reproductive tract, Shobu et al. (2006) found minimal surface expression of HLA-F on extravillous trophoblasts (EVTs) during the first trimester (Shobu et al., 2006), though Hackmon et al. (2017) later demonstrated increased HLA-F expression on migrating and invasive EVTs, pointing to a functional role during trophoblast invasion and early placentation (Hackmon et al., 2017).

One of the unique immunological features of HLA-F is its ability to exist in an open conformer state lacking a bound peptide, which enables it to interact with various immune receptors independently of antigen presentation. It binds a wide spectrum of inhibitory and activating receptors, including.

ILT2 (LILRB1) and ILT4 (LILRB2)

These receptor-ligand interactions enable HLA-F to modulate NK cell function, antigen-presenting cell activity, and overall immune tolerance at the maternal–fetal interface. HLA-F engagement with LILRB1 and LILRB2 on decidual macrophages and dendritic cells promotes a tolerogenic immune profile, facilitating trophoblast invasion, vascular remodeling, and immune quiescence, all of which are crucial during early pregnancy (Gardiner, 2008; Ishigami et al., 2015).

Recent studies have begun to shed light on the functional relevance of HLA-F in reproductive immunology and its potential contribution to IVF outcomes. While not as extensively studied as HLA-G or HLA-E, HLA-F is increasingly recognized as a key regulator of endometrial receptivity and immune tolerance.

MICA (MHC class I polypeptide-related sequence A) and MICB are nonclassical MHC class I molecules that exhibit extensive genetic diversity. As of September 2025, the IPD-IMGT/HLA Database lists 589 MICA alleles, including 29 protein-coding and 9 null variants, and 312 MICB alleles, with 53 protein-coding and 25 null variants (Robinson et al., 2024) (Figure 1). Unlike classical MHC class I molecules, MICA/B do not associate with β2-microglobulin, nor do they present peptides to T cells. Instead, they serve as stress-induced ligands for activating immune receptors.

Structurally, MICA/B are transmembrane glycoproteins composed of three extracellular α domains (α1, α2, α3), a transmembrane segment, and a cytoplasmic tail. Their promoter regions contain heat shock elements, making their expression highly inducible under cellular stress, including heat shock, oxidative damage, infection, and inflammation (Groh et al., 1999).

Under normal physiological conditions, MICA/B are expressed at low levels, predominantly in gut epithelium, thymus, endothelial cells, and fibroblasts. However, their expression is markedly upregulated during tumorigenesis, viral infection, autoimmune disease, and pregnancy-associated stress (Diefenbach and Raulet, 2002; Tieng et al., 2002). This dynamic regulation suggests a role for MIC molecules in immune surveillance and tissue adaptation during pregnancy.

The primary immunological role of MICA and MICB is their interaction with the NKG2D receptor, expressed on NK cells, CD8⁺ cytotoxic T lymphocytes (CTLs), and γδ T cells (Bauer et al., 2018). Unlike classical HLA-TCR recognition, the MIC–NKG2D axis enables MHC-unrestricted cytotoxicity, allowing immune cells to detect and eliminate stressed, infected, or transformed cells without antigen presentation (Diefenbach and Raulet, 2001).

MIC proteins exist in both membrane-bound and soluble forms. Soluble MICA/B (sMIC) are generated through proteolytic shedding by metalloproteinases, such as ADAM10 and ADAM17 (Waldhauer and Steinle, 2006). While membrane-bound MIC molecules activate immune effector functions, soluble MICs paradoxically inhibit immunity by downregulating NKG2D surface expression, leading to impaired NK and CD8⁺ T cell cytotoxicity (Groh et al., 2002; Doubrovina et al., 2003).

In the context of pregnancy, placental trophoblasts produce both membrane-bound and microvesicle-associated forms of MICA/B, implicating these proteins in maternal–fetal immune regulation. Studies suggest that trophoblast-derived sMIC contributes to local immune tolerance, modulating NK cell activity to support fetal survival (Mincheva-Nilsson et al., 2006).

Growing evidence links aberrant MIC expression and sMIC shedding with adverse reproductive outcomes, particularly in ART such as IVF. A study by Porcu-Buisson et al. (2007) found elevated serum sMIC levels in women undergoing IVF, which were significantly associated with implantation failure and early miscarriage (Porcu-Buisson et al., 2007). In a subsequent prospective analysis, Haumonte et al. (2014) reported that 38% of infertile women exhibited detectable sMIC levels, which correlated negatively with clinical pregnancy rates (Haumonte et al., 2014). These findings support the notion that sMIC may serve as a noninvasive biomarker of endometrial receptivity and immune dysfunction in IVF.

Moreover, altered metalloproteinase activity a key driver of sMIC shedding has been observed in women with RIF, potentially contributing to overproduction of sMIC and inhibition of NKG2D-mediated cytotoxic responses (Shibahara et al., 2005). This imbalance may suppress IFN-γ secretion, compromising trophoblast invasion and spiral artery remodeling, both essential for early placental development. MIC shedding has also been implicated in autoimmune and alloimmune disorders, which frequently co-occur with infertility. Elevated sMIC levels have been reported in women with celiac disease, autoimmune thyroiditis, and systemic lupus erythematosus conditions associated with increased risk of miscarriage and poor IVF outcomes (Meloni et al., 1999; Meresse et al., 2004). These findings suggest that sMIC levels reflect not only local uterine immunoregulation but also systemic immune activation.

In IVF patients undergoing preimplantation genetic testing (PGT), sMIC levels were found to be elevated in those who later developed intrauterine growth restriction (IUGR) or PE pregnancy complications often linked to vascular dysfunction and immune dysregulation (Basu et al., 2008; Raulet et al., 2013). These observations further implicate MIC molecules in vascular remodeling, decidualization, and placental homeostasis. Overall, evidence links elevated sMIC with impaired implantation and placental development, but inconsistencies in assay methods and patient selection highlight the need for standardized protocols and longitudinal studies to confirm its predictive value in IVF.

Collectively, the MICA/B–NKG2D axis represents a pivotal immune checkpoint in reproductive immunology. MIC proteins, by serving as stress-inducible ligands for activating immune receptors, participate in both immune surveillance and immune evasion, depending on their form (membrane-bound vs. soluble) and regulatory context.

In the setting of IVF, dysregulated MIC expression whether through genetic variation, inflammatory stress, or aberrant shedding may impair NK and T cell function, contributing to implantation failure, miscarriage, and placental pathologies such as PE and IUGR. Key Areas for Future Investigation Include.

Mechanisms of sMIC generation and regulation in early pregnancy and ART settings.

As research advances, targeting the MIC–NKG2D pathway may offer novel therapeutic strategies. For example, metalloproteinase inhibitors could prevent excessive shedding of MIC molecules from the cell surface, thereby preserving their interaction with activating immune receptors. NKG2D agonists may enhance NK cell and cytotoxic T cell recognition of abnormal cells, while MIC-stabilizing agents could prolong surface expression and maintain effective immune surveillance may help rebalance immune tolerance and improve ART success. Integrating MIC profiling into personalized IVF protocols could enhance embryo selection, predict implantation potential, and improve outcomes for patients with unexplained infertility or repeated IVF failure.

Supporters of pre-implantation HLA immunogenetic screening posit that identifying maternal and paternal polymorphisms in non-classical HLA genes may reveal immunogenetic mismatches that predispose patients to implantation failure, PE, or recurrent pregnancy loss (RPL).

One well-studied example involves maternal KIR haplotypes and their interaction with fetal HLA-C ligands. Several studies have shown that the KIR AA haplotype, which lacks activating KIRs, is associated with higher risk of early miscarriage, placental insufficiency, and lower live birth rates, particularly in oocyte donation cycles where maternal–fetal HLA compatibility may be further reduced (Alecsandru et al., 2014; Zhang and Wei, 2021). While promising, KIR/HLA-C testing remains limited by inconsistent predictive power across ethnic groups, underscoring the need for population-specific validation before routine clinical use.

Moreover, soluble HLA-G (sHLA-G) levels along with specific HLA-G 3′UTR polymorphisms (e.g., the 14-bp ins/del variant) have been linked to implantation potential and pregnancy maintenance. A study by Alecsandru et al. (2020) suggests that low sHLA-G expression in embryo culture media, especially in combination with specific maternal genotypes, may correlate with poorer IVF outcomes (Alecsandru et al., 2020).

Emerging research has also identified associations between HLA-F SNPs (e.g., rs2523393) and improved endometrial receptivity, as well as HLA-E*01:03 homozygosity and impaired male fertility, suggesting a broader role for non-classical HLA polymorphisms across both partners in influencing reproductive success (Gelmini et al., 2016; Mika et al., 2018; Langkilde et al., 2020). Yet, most findings come from small-scale or single-center studies, making replication in larger, multi-ethnic cohorts essential to determine true clinical utility.

Recent advances in machine learning and bioinformatics have significantly enhanced our ability to predict the functional consequences of non-classical HLA polymorphisms. For instance, DeepHLAPred, a deep learning-based tool, can accurately forecast ligand-binding affinity for both classical and non-classical HLA variants, offering insights into immune receptor engagement and signaling potential (Huang et al., 2023). Such tools may soon enable in silico screening of HLA genotypes prior to embryo transfer, identifying high-risk immunogenetic profiles and guiding immunomodulatory interventions.

In parallel, innovations in single-cell RNA sequencing, decidual immune profiling, and AI-assisted embryo selection are facilitating a more integrated approach to evaluating maternal–fetal immunological compatibility. These technologies promise to shift IVF from a one-size-fits-all approach toward precision reproductive immunology.

Despite these advancements, the routine clinical application of non-classical HLA genotyping in IVF remains controversial for several reasons.

While the concept of non-classical HLA genotyping prior to IVF is scientifically compelling, current evidence is insufficient to recommend routine clinical implementation. Nonetheless, ongoing research is rapidly expanding our understanding of how HLA-G, HLA-E, HLA-F, and MICA/B shape maternal immune responses during early pregnancy. Future large-scale, ethnically diverse, and functionally validated studies are essential to determine.

Which polymorphisms or haplotypes are truly predictive of implantation failure or adverse pregnancy outcomes.

If these questions are answered affirmatively, non-classical HLA testing may become a cornerstone of personalized IVF protocols, guiding embryo transfer strategies, immune therapies, and patient counseling in cases of unexplained infertility or RIF.

In particular, MICA and MICB stress-inducible ligands that bind to the NKG2D receptor demonstrate a dual immunological function. Membrane-bound MICA/B enhance immune activation in response to cellular stress, while soluble forms (sMIC), generated by proteolytic shedding, may downregulate NKG2D and suppress cytotoxic immune responses. This balance is crucial during implantation, where excessive immune activation or suppression can impair endometrial receptivity. Recent research has linked elevated sMIC levels to.

Impaired implantation due to decreased NKG2D expression on uNK cells and CD8⁺ T cells.

Further, polymorphisms such as MICA*008, promoter variants, and regulatory SNPs have been associated with differences in sMIC shedding, NKG2D affinity, and reproductive outcomes. However, the clinical application of MIC genotyping remains limited due to population variability, lack of assay standardization, and incomplete mechanistic data.

To realize the full clinical value of non-classical HLA and MIC molecules in reproductive medicine, future research should focus on the following critical areas.