Complex interaction of HLA class I and innate immune receptors: in silico study and implication in viral pathologies

HLA class I and II molecules are important not only in activation but also in the maintenance of the adaptive immune response. However HLA class I has an important role also in the activation of innate immune response through the interaction with innate immune receptors such as KIRs and LILRs expressed in different cell subpopulation, mainly NK cells. KIR3DL1 interacts with polymorphisms located on Human Leukocyte Antigen (HLA)-A and HLA-B molecules that have been widely associated with the control of several infections. LILRB1 is a receptor with inhibitory function able to recognize both conventional and non-conventional HLA class I molecules binding the conserved α3-domain of HLA class I molecules. However, the interaction between HLA molecules, loading peptides and KIRs/LILRs receptors in relation to viral immune responses it’s still unclear. Aim of this project has been to evaluate the capability of HLA to both modulate adaptive and innate immune responses against viral infections using HIV as a viral model of infection. The project has been divided in immunogenetic and immunoinformatic analyses. The immunogenetic analysis has been focused on HLA amino acid position polymorphisms leading to different HIV progression. To minimize the effects of genetic background due to population coming from different geographic area and viral strain variability in the cohort, the analysis has been performed in a cohort of children with a monophyletic HIV-1 infection (CRF02_AG) during an outbreak in Libya. The immunoinformatic analysis has selected a strategy able to evaluate the role of the HLA-B amino acid polymorphisms in HLA-B/KIR3DL1 and HLA-B/LILRB1 interactions. The study has taken into consideration different HLA-B, KIRs and LILRBs alleles or haplotypes previously associated with HIV disease progression or non progression together with the contribution of different loading peptides. In the immunogenetic analysis, HLA-B locus and in particular amino acid positions 80-83 belonging to Bw4/Bw6 epitope together 6 with amino acid position 194 resulted the most interesting. Bw4 epitope resulted associated with LTNP (phenotype-frequency: LTNP=88.46%, FP=47.83%, OR=8.36; p<0.006), as well as the presence of Val 194 (Val/Val: LTNP = 19.23% FP = 8.7% OR = 2.21; Val/Ile: LTNP = 53.85% FP = 26.09% OR = 2.06; Ile/Ile: LTNP = 26.92% FP = 65.22% OR = 0.41; P < 0.03 all comparison). The immunoinformatic analysis on HLA/KIR3DL1 interaction show that the presence of HLA-B Bw4 epitope leads to a stronger interaction with all the different KIR3DL1 alleles. Among the different KIR3DL1 alleles, KIR3DL1*001 shows the stronger interaction in terms of general number of contacts together with the high expressive alleles *009 and *015 while less number of contacts are observed in presence of the low expressive *005 allele. Contribution of the peptide seems to be the key point able to disrupt HLA-B/KIR3DL1 interaction, in particular in presence of negative charged amino acid residues in peptide positions P8. This is in according to previously observed studies focused on KIR3DL1 interaction with HLA-B*27:05 and HLA-B*57:01 alleles due to the conserved amino acid residue Glu282 of KIR3DL1 receptor (Kollnberger et al., 2007; O'Connor et al., 2014; Peruzzi et al., 1996a; Peruzzi et al., 1996b; Stewart-Jones et al., 2005; Vivian et al., 2011). In the context of HLA/LILRB1 immunoinformatic analysis, HLA-B alleles carrying Ile 194 polymorphism show a higher strength of interaction respect to HLA-B alleles carrying Val 194, in all the LILRB1 haplotypes analyzed. Considering the most frequent LILRB1 haplotypes, a stronger interaction of LILRB1.01 haplotypes with HLA-B alleles is shown respect to LILRB1.02 and LILRB1.03 haplotypes in according with the affinity constant evaluated between LILRB1 haplotypes in the case of HLA-B*35:01 interaction study (Kuroki et al., 2005). Interestingly, for the first time we propose a strong contribution of amino acid polymorphisms located in peptide binding pocket 1 able to modulate HLA-B/LILRB1 interaction. Overall, the strong influence of the loaded peptide in the innate immune response point out similarities between T-cell and NK cell immune responses. In conclusion, immune response against viral infection might be in part determined by the binding capability of HLA-B and in part by the interaction between HLA-B molecules and NK’s inhibitory receptor such as KIR3DL1 and LILRB1. This might shed new light on the mechanisms behind the association of genetic susceptibility to HLA viral infections and represent a powerful tool for a more effective strategy in vaccine and diagnostics design ensuring wider coverage of the populations, including genetically susceptible subjects.