An archaic HLA class I receptor allele diversifies natural killer cell-driven immunity in First Nations peoples of Oceania

Abstract

Genetic variation in host immunity impacts the disproportionate burden of infectious diseases that can be experienced by First Nations peoples. Polymorphic human leukocyte antigen (HLA) class I and killer cell immunoglobulin-like receptors (KIRs) are key regulators of natural killer (NK) cells, which mediate early infection control. How this variation impacts their responses across populations is unclear. We show that HLA-A^∗24:02 became the dominant ligand for inhibitory KIR3DL1 in First Nations peoples across Oceania, through positive natural selection. We identify KIR3DL1^∗114, widespread across and unique to Oceania, as an allele lineage derived from archaic humans. KIR3DL1^∗114⁺NK cells from First Nations Australian donors are inhibited through binding HLA-A^∗24:02. The KIR3DL1^∗114 lineage is defined by phenylalanine at residue 166. Structural and binding studies show phenylalanine 166 forms multiple unique contacts with HLA-peptide complexes, increasing both affinity and specificity. Accordingly, assessing immunogenetic variation and the functional implications for immunity are fundamental toward understanding population-based disease associations.

Keywords: First Nations Australians; HLA; KIR; NK cells; Oceania; immunogenetic diversity; influenza virus; introgression.

Figure 1.. High combinatorial diversity of HLA and KIR in First Nations Australians.

(A) Shown is the number of distinct HLA-A, -B, and -C alleles of 75 random individuals for each indicated population. Error bars represent mean ± standard deviation from 1,000 resamples with replacement. Unpaired two-sample Wilcoxon tests with Bonferroni multiple test correction were performed between First Nations Australians (shaded red) and each reference population (p for each comparison is < 0.001, except with Han for HLA-C, which is not significant, ns). (B) The ten most frequent HLA class I haplotypes detected in First Nations Australians. The KIR ligands are coloured; A3/11 (yellow), Bw4 (green), C1 (red), C2 (blue). At the right are shown their frequencies in First Nations Australians (AUS), Papua New Guineans (PNG), Europeans (EUR), and East Asians (EAS) (key is underneath). † - haplotype of likely European origin (Figure S1 and Table S2). (C) The distribution of KIR ligands among HLA class I allotypes. The combined frequencies of each subset are shown. (D) Shows the total allele frequencies of Bw4⁺ HLA-B (left) and Bw4⁺ HLA-A (right) across the representative populations. Bw4⁺ allotypes that preferentially interact with 005 lineage KIR3DL1 are shown in light green; those that interact with 005- and 015-lineage in dark green. (E) The number of distinct KIR alleles detected in 75 individuals from each of the populations indicated, obtained as described for Figure 1A. (F) Frequency spectra of KIR alleles encoding inhibitory receptors specific for HLA class I (coloured by cognate ligand from panel B). Each segment corresponds to a distinct KIR allotype. (blue) activating KIR3DS1, (purple) allotypes that are not expressed on the cell surface, and (grey) gene absence. (G) Ten most frequent haplotypes considering inhibitory KIR specific for HLA class I. (H) The number of distinct HLA class I and KIR genotypes observed in 80 First Nations Australians. “KIR + ligand” excludes non-expressed allotypes and non-functional interactions. An explanation of HLA and KIR nomenclature is given in Methods.

Figure 2.. KIR3DL1*114 has archaic human origin and coevolved with HLA-A*24:02 in Oceania

(A) Neighbour-joining phylogenetic analyses of external-domain coding sequences of the KIR3DL1/S1 allotypes observed in First Nations Australians. Colours represent the three major KIR3DL1/S1 lineages, 005 (blue), 015 (red), 3DS1 (green), and recombinant 005/015 alleles (purple). KIR3DL1*114 clusters in the 015 lineage. Bootstrap values for nodes are given when >50%. Lower right pie chart shows relative allele frequencies of the three lineages, according to the same colour scheme, grey indicates gene absence. (B) Polymorphic amino acid residues distinguishing the lineages of KIR3DL1/S1, alongside KIR3DL1*086 and *114. KIR3DL1*015 is used as the reference sequence; residues identical to *015 are indicated by dashes. Residue 166 is shown in bold. (C) Geographic distribution of KIR3DL1*114/*086, HLA-A*24:02 and KIR3DL1 005-lineage alleles in Oceania and Southeast Asia (Table S3). (D). KIR haplotypes carrying KIR3DL1*114 or *086 and the populations from where they were characterized. (n.d. -not determined) (E) D-statistics for testing Denisovan introgression of KIR3DL1*114 (n=4), *005 (n=6), and *015 (n=6). KIR3DL1*114 only includes homozygous individuals. W – haplotypes contain indicated allele, Z – Z-score, D – D score. See Methods and Table S3. (F) Nucleotide diversity (π) across distinct regions of the genome in HLA-A*24:02, HLA-A*34:01, and HLA-A*11:01 homozygous Papua New Guineans. π in 100bp windows; HLA-A and -BC are flanking 500kbp on each side. (*** - p<0.001, compared to HLA-A; Wilcoxon test (Table S3)). (G) Correlation of HLA-A*24:02 with KIR3DL1*114+005-lineage allele frequencies. Each dot represents a population, coloured by region (SEA – Southeast Asia). Only populations with less than 50% Polynesian-related ancestry (by RFMix) are included (others are shown in grey).

Figure 3.. KIR3DL1*114 inhibits NK cells upon recognition of HLA-A*24:02.

(A) Shows KIR3DL1 expression levels on the cell surface of primary NK cells isolated from individuals having KIR3DL1*114 (n=3), *001 (n=3), *005 (n=4) or *015-lineage (n=4) allotypes (Supplementary Figure S3A-B and Table S4). The geometric Mean Fluorescence Intensity (gMFI) of KIR3DL1 expression for each allotype was normalised within each donor to KIR3DL1^neg NK cells. **p<0.01, ***p<0.001; Kruskal-Wallis test. Error bars indicate the standard deviation of the mean. (B) Shows CD107a (left) and IFNγ (right) expression by ex vivo NK cells following incubation with 721.221 cells expressing HLA-A*24:02. NK cells were isolated from donors expressing KIR3DL1*114 (n=3), *001 (n=3), *005 (n=6) or *015-like (n=7). For each donor, values are normalized (% max) to those obtained from KIR3DL1⁺ NK cells incubated with the parental, HLA class I^neg 721.221 cell line. Each dot represents one donor. CD107a p=0.07, IFNγ p=0.02; Kruskal-Wallis test. (C) As panel B, except NK cells were differentiated into those expressing only KIR3DL1, KIR2DL1/S1 or KIR2DL2/3/S2. Each circle represents one donor. Responses were normalised to the maximal obtained with the parental 221 cell line and compared using a one-way ANOVA with Tukey’s multiple comparison test (**p<0.01, ***p<0.001). (D) Assays using NK cells isolated from ten donors and expanded in vitro using IL-2. Each donor was homozygous for one of four expressed KIR3DL1 allotypes. HLA allotype of target cells is indicated on the x-axis (allotypes present in the cohort shown in bold). CD107a expression of KIR3DL1⁺ NK cells was normalised to the maximal expression in response to parental 721.221 cells. Background CD107a expression by NK cells was subtracted. For each donor, two-three replicates were performed, and the mean of all replicates from the donors is shown as the bar graph with the standard deviation plotted. ***p<0.001, **p<0.01, *p<0.05; Two-way ANOVA with Bonferroni's multiple comparisons. For each KIR3DL1 allotype, triangles, circles and squares represent the donors used, whose genotypes are given in Table S4. (E-G) Shows the result of in vitro activation (CD69 upregulation) of reporter cells expressing specific KIR3DL1 allotypes. Jurkat cells were transduced with a given natural (panel E) or mutated (panels F and G) KIR3DL1 allotype that was fused to the intracellular domain of CD3ζ. Colours represent the KIR3DL1 allotypes, as shown in the key. The reporter cells were incubated with a panel of 721.221 target cells for eight hours and the subsequent upregulation of CD69 assessed by flow cytometry (Supplementary Figure S3D-E). Three independent experiments were performed, each in triplicate, and statistics performed on the average of the replicates. *p<0.05, **p<0.001, ***p<0.001; 2-way ANOVA with Tukey’s multiple comparisons test.

Figure 4.. Phenylalanine at residue 166 of KIR3DL1*114 enhances interaction with HLA ligands.

(A) Ribbon diagram of the crystal structures solved for KIR3DL1*001 (left, cyan) and KIR3DL1*114 (right, blue) bound to the HLA-A*24:02-TW9 peptide (Influenza virus A polymerase basic protein 2 residues 549–557) complex. HLA-A*24:02 in grey, peptide in mauve and β2M in orange. KIR3DL1 and HLA domains are labelled on the left, positions of the peptide and residue 166 are indicated on the right. (B) Orthogonal view of KIR3DL1 residue 166 interactions with α1 helix residues of the HLA molecule. (i) KIR3DL1*001 (L166), (ii) KIR3DL1*114 (F166). Dotted lines indicate Van der Waals interactions. (C) (i) Side view; cut away to show the interactions of KIR3DL1 residue 166 with peptide and HLA molecules. Leucine 166 is represented shaded cyan and Phenylalanine dark blue. (ii) Aerial view of peptide showing contacts with L166 and (iii) F166. (D) Comparison of interactions of KIR3DL1 residue 166 from allotypes (left to right) KIR3DL1*001, *086 and *114 with HLA α1 helix residues of (top) HLA-A*24:02-NEF and (bottom) -B*57:03-AW10. NEF is derived from HIV-1, AW10 from human Catenin.

Figure 5.. KIR3DL1 allotypes having Phenylalanine at residue 166 bind HLA-A*24:02-peptide complexes with greater affinity than do other allotypes.

(A) (Left to right) HLA class I allotype, and name, sequence and origin of bound peptide. Residues in bold correspond to P8 of the 9-mer peptide. (B-C) Shown are the affinity curves of soluble KIR3DL1 allotypes binding to the immobilized HLA-B*57:03-AW10 (B) or HLA-A*24:02-TW9 complex (C). The KIR3DL1 allotypes are *001, *005, *015, *086, and *114, and mutations *086-F166L, *114-F166L. R.U. – Response Units. Mean from two independent experiments shown. Underneath are shown the mean K_D values; +/− values and remaining HLA-A*24:02-peptide binding curves are given in Figure S5. (D) Comparison of K_D values across HLA-A*24:02-peptide complexes. Data are representative of two independent experiments. Error bars are SEM. (E) Comparison of K_D values obtained by the KIR3DL1 allotypes across the four HLA-A*24:02 complexes. T-test, Bonferroni corrected (Pc).