KIR3DL01 recognition of Bw4 ligands in the rhesus macaque: maintenance of Bw4 specificity since the divergence of apes and Old World monkeys

Abstract

The identification of MHC class I ligands for rhesus macaque killer cell Ig-like receptors (KIRs) is fundamental to our basic understanding of KIR and MHC class I coevolution and to the study of NK cell responses in this nonhuman primate model for AIDS and other viral diseases. In this study, we show that Mamu-KIR3DL01, which is expressed by ∼90% of rhesus macaques, recognizes MHC class I molecules with a Bw4 motif. Primary NK cells expressing Mamu-KIR3DL01 were identified by staining with a mAb which, in this study, was shown to bind Mamu-KIR3DL01 allotypes with an aspartic acid at position 233. The cytolytic activity of Mamu-KIR3DL01(+) NK cells was suppressed by cell lines expressing the Bw4 molecules Mamu-B*007:01, -B*041:01, -B*058:02, and -B*065:01. The Bw4 motif was necessary for Mamu-KIR3DL01 recognition because substitutions in this region abrogated Mamu-KIR3DL01(+) NK cell inhibition. However, the presence of a Bw4 motif was not sufficient for recognition because another Bw4 molecule, Mamu-B*017:01, failed to suppress the cytolytic activity of these NK cells. Replacement of three residues in Mamu-B*017:01, predicted to be KIR contacts based on the three-dimensional structure of the human KIR3DL1-HLA-Bw4 complex, with the corresponding residues at these positions for the other Mamu-Bw4 ligands restored Mamu-KIR3DL01(+) NK cell inhibition. These results define the ligand specificity of one of the most polymorphic and commonly expressed KIRs in the rhesus macaque and reveal similarities in Bw4 recognition by Mamu-KIR3DL01 and human KIR3DL1, despite the absence of an orthologous relationship between these two KIRs or conservation of surface residues predicted to interact with MHC class I ligands.

FIGURE 1

NKVFS1 stains Mamu-KIR3DL01*001. Jurkat cells were electroporated with HA-tagged KIR expression constructs that co-express GFP and were stained with the NKVFS1 Ab and an HA-specific mAb. Samples were gated on GFP⁺ cells and analyzed for NKVFS1 versus anti-HA staining. Data shown is representative of results obtained in three independent experiments.

FIGURE 2

Residue 233D is essential for NKVFS1 binding to Mamu-KIR3DL01*001. (A) An amino acid alignment shows residues that differ between Mamu-KIR3DL01*001 and -KIR3DL01*002 in D0, D1, and D2. (B) Jurkat cells were electroporated with constructs expressing HA-tagged KIR with reciprocal substitutions at positions that differ between Mamu-KIR3DL01*001 and -KIR3DL01*002, or (C) constructs expressing five different Mamu-KIR3DL01 alleles. Staining of GFP⁺ cells for HA and NKVFS1 is shown. Data shown is representative of results obtained in three independent experiments.

FIGURE 3

Mamu-B*007:01, Mamu-B*041:01, Mamu-B*058:02, and Mamu-B*065:01 are ligands for Mamu-KIR3DL01. (A) PBMC were stimulated overnight with parental 721.221 cells, or 721.221 cells expressing the indicated MHC class I molecules at a 5:1 ratio in the presence of a mAb to CD107a, and stained with NKVFS1 and antibodies to CD3 and CD8. After gating on CD3⁻CD8⁺ lymphocytes, the upregulation of CD107a on NKVFS1⁺ versus NKVFS1⁻ NK cell subsets was assessed. (B) The frequency of CD107a upregulation for Mamu-KIR3DL01⁺ and -KIR3DL01⁻ populations is summarized for four different animals where error bars indicate + 1 SD. Asterisks indicate a significant difference (*p<0.05, **p<0.01, ***p<0.005 by two-way ANOVA with Dunnett’s test) between Mamu-KIR3DL01⁺ and -KIR3DL01^– populations coincubated with a given 721.221 cell line. (C) Mamu-KIR3DL01⁺ and -KIR3DL01⁻ NK cells from the same animal were coincubated with parental 721.221 cells or 721.221 cells expressing the indicated rhesus macaque MHC class I molecules at the indicated E:T ratios. Killing of target cells was evaluated by release of CAM from target cells into the culture supernatant. % Specific Lysis is defined as (test release – spontaneous release) / (maximum release – spontaneous release). Results are representative of those obtained with expanded cells from three different animals and the compiled results for Mamu-KIR3DL01⁺ NK cells from these animals are presented in (D). Error bars indicate + 1 SD. Asterisks indicate a significant difference (*p<0.05, ****p<0.001 by two-way ANOVA with Dunnett’s test) between coincubation with 721.221 parental cells and coincubation with the indicated cell line at all E:T ratios shown.

FIGURE 4

Mamu-KIR3DL01 recognition of Mamu-B*065 is dependent upon Bw4 residues N77, T80, and R83. Mutations were introduced into the Bw4 region of Mamu-B*065:01 to make it Bw6-like at positions which vary between Bw4 and Bw6. (A) Mamu-KIR3DL01⁺ and -KIR3DL01⁻ NK cells from the same animal were coincubated with parental 721.221 cells or 721.221 cells expressing the indicated Mamu-B*065:01 mutants at the indicated E:T ratios. Killing of target cells was evaluated by release of CAM from target cells into the culture supernatant. Results are representative of expanded NK cells from three different animals, and the compiled results for Mamu-KIR3DL01⁺ NK cells from these animals are presented in (B) wherein error bars indicate + 1 SD. Asterisks indicate a significant difference (*p<0.05, ****p<0.001 by two-way ANOVA with Dunnett’s test) between coincubation with 721.221-Mamu-B*065 cells and coincubation with 721.221 cells expressing a given Mamu-B*065 mutant at all E:T ratios shown.

FIGURE 5

Residues G76, F142, and A149 are determinants of recognition by Mamu-KIR3DL01. (A) An amino acid alignment of the MHC class I α1 and α2 domains of Mamu-B*065:01, -B*007:01, and -B017:01 with the predicted KIR contact residues highlighted based on the HLA-B*57:KIR3DL1*001 crystal structure (36). Reciprocal substitutions were introduced into Mamu-B*065:01 and -B*017:01 at each of the positions indicated with asterisks that are the same for Mamu-B*065:01 and -B*007:01, but differ for Mamu-B*017:01. (B) Mamu-KIR3DL01⁺ and -KIR3DL01⁻ NK cells from the same animal were coincubated with parental 721.221 cells or 721.221 cells expressing the indicated rhesus macaque MHC class I mutants at E:T ratios ranging from 10 to 0.625. Cytotoxicity was measured by the release of CAM from target cells into the culture supernatant. Results are representative of those obtained with expanded cells from three different animals, and the compiled results for Mamu-KIR3DL01⁺ NK cells from these animals are presented in (C), wherein error bars indicate + 1 SD. Asterisks indicate a significant difference (*p<0.05, **p<0.01, ****p<0.001 by two-way ANOVA with Dunnett’s test) between coincubation with 721.221-Mamu-B*065:01 or -B*017:01 cells and coincubation with 721.221 cells expressing a given Mamu-B*065:01 or -B*017:01 mutant at all E:T ratios shown.

FIGURE 6

KIR contact residues of rhesus macaque Bw4 molecules. (A) An amino acid alignment of the MHC class I α1 and α2 domains of HLA-B*5701 and Mamu-B*065:01, -B*007:01, -B*017:01, -B*022:01, B*041:01, B*056:01, and B*058:02 with the predicted KIR contact residues shaded based on the HLA-B*57:KIR3DL1*001 crystal structure (36). (B) Crystal structure of HLA-B*57 (yellow) and bound peptide (cyan). Highlighted in red are positions important for recognition by Mamu-KIR3DL01 identified in this study: residues 76, 77, 80, 83, 142, and 149.

FIGURE 7

Phylogenetic analysis of human and rhesus macaque KIRs. (A) Midpoint-rooted phylogenetic tree generated from full-length amino acid sequences using Neighbor-Joining (NJ) method. Bootstraps were calculated with 500 iterations. (B) An amino acid alignment of the D0, D1, and D2 domains of human KIR3DL1 and allotypes of Mamu-KIR3DL01 with the predicted MHC ligand contact residues shaded based on the HLA-B*57:KIR3DL1*001 crystal structure (36). (C) Crystal structure of human KIR3DL1 with HLA-B*57 contact residues highlighted. Residues highlighted in red are conserved between KIR3DL1 and Mamu-KIR3DL01*001, whereas residues highlighted in yellow differ between the two KIRs.