Variation in the HLA-F gene locus with functional impact is associated with pregnancy success and time-to-pregnancy after fertility treatment

Hum Reprod. 2020 Mar 27;35(3):705-717. doi: 10.1093/humrep/dez276.


Study question: The aim of this study was to investigate a possible influence of three single nucleotide polymorphisms (SNPs) in the HLA-F gene locus on time-to-pregnancy and pregnancy success after fertility treatment.

Summary answer: HLA-F SNP genotypes and HLA-F diplotypes are associated with the number of fertility treatment cycles needed to achieve pregnancy and live birth.

What is known already: HLA class Ib molecules, including HLA-F, which are known to be expressed by extra-villous trophoblast cells have immunomodulatory properties and play a role at the feto-maternal interface. However, a few recent studies suggest that HLA-F expressed in the mid-luteal endometrium may play a part in the establishment of pregnancy as well. Three genetic polymorphisms in the HLA-F gene locus influence the expression of HLA-F in the mid-luteal endometrium and are associated with time-to-pregnancy in healthy women.

Study design, size, duration: The current study included 102 female patients and 91 male patients attending for ART treatment and recruited between 2009 and 2014 at fertility clinics in a University Hospital setting, and 78 fertile female controls recruited in 2017 and 2018 at a department of Obstetrics and Gynaecology in a University Hospital. All women in the control group conceived naturally, and no other clinical data for the controls were retrieved.

Participants/materials, setting, methods: Genotyping of genomic DNA from blood samples was performed with Sanger sequencing for the three SNPs of interest in the HLA-F gene locus: rs1362126 (G/A), rs2523405 (T/G) and rs2523393 (A/G). Furthermore, clinical data were collected for the couples in fertility treatment.

Main results and the role of chance: There were no significant differences in the distributions of the three HLA-F SNP genotypes and alleles between the female fertile control group and the female infertility group. We considered if the number of treatment cycles was related to the HLA-F SNP genotypes and HLA-F diplotypes in a discrete time to event analyses. A significant association with longer time-to-pregnancy, measured as number of fertility treatment cycles, was observed for women in the ART group who carried the HLA-F genotypes that are associated with a lower amount of HLA-F mRNA expressed in mid-luteal endometrium. For the rs1362126 AA genotype relative to the GG genotype, the odds ratio (OR) was 0.30 (95% CI = 0.10-0.87, P = 0.02); for the rs2523405 GG genotype relative to the TT genotype, the OR was 0.40 (95% CI = 0.15-1.04, P = 0.06); and for the rs2523393 GG genotype relative to the AA genotype, the OR was 0.27 (95% CI = 0.09-0.78, P = 0.01). In addition to comparing the HLA-F genotypes by a standard likelihood-ratio test, a trend test based on the number of G or A alleles were also performed. The HLA-F genotypes associated with longer time-to-pregnancy in these tests were as follows: number of A alleles at rs1362126 (P = 0.01), the OR was 0.56 per A allele (95% CI = 0.35-0.89); number of G alleles at rs2523405 (P = 0.05), OR was 0.65 per G allele (95% CI = 0.42-1.00); and number of G alleles at rs2523393 (P = 0.01), OR was 0.56 per G allele (95% CI = 0.36-0.86). On average, for the rs1362126 SNP, 2.1 more treatment cycles for a woman who carried the AA genotype were needed to achieve pregnancy within the first eight treatment cycles compared with a woman who carried the GG genotype. Likewise, for the rs2523405 SNP, 1.8 more cycles for the GG genotype compared with the TT genotype were needed, and for the rs2523393 SNP, 2.2 more treatment cycles for a woman who carried the GG genotype compared with a woman who carried the AA genotype were needed. Adjustments for the covariates BMI, female age, IVF (yes/no for each cycle), ICSI (yes/no for each cycle), female factor (yes/no) and male factor (yes/no), were also performed modeling the cycle-specific probabilities and the genotypes remained significant and almost unchanged.

Limitations, reasons for caution: Specific types of ART will be chosen from the start of treatment, which means that the chances of achieving pregnancy could differ between the women solely due to their first line of treatment. However, multivariate analyses are performed to adjust for type of ART treatment.

Wider implications of the findings: To our knowledge, this is the first study that shows associations between, and implications of, HLA-F gene locus variation and time-to-pregnancy and pregnancy success in a clinical setting for fertility treatment/ART.

Study funding/competing interest(s): Supported by the Region Zealand Health Sciences Research foundation and by Zealand University Hospital through the ReproHealth Research Consortium ZUH. The authors declare no conflict of interest.

Keywords: HLA-F genotypes; ART; infertility; natural killer cells; pregnancy; reproductive immunology.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Female
  • Fertilization in Vitro
  • Genotype
  • Histocompatibility Antigens Class I / genetics*
  • Humans
  • Infertility, Female*
  • Live Birth
  • Male
  • Pregnancy
  • Pregnancy Rate
  • Time-to-Pregnancy*


  • HLA-F antigens
  • Histocompatibility Antigens Class I