Activating Ly-49D and inhibitory Ly-49A natural killer cell receptors demonstrate distinct requirements for interaction with H2-D(d)

Abstract

The activating Ly-49D receptor and the inhibitory Ly-49A receptor mediate opposing effects on natural killer (NK) cell cytotoxicity after interaction with the same major histocompatibility complex ligand, H2-D(d). To compare Ly-49D and Ly-49A interactions with H2-D(d), we created mutations in H2-D(d) and examined the functional ability of these mutants to activate lysis through Ly-49D or to inhibit lysis through Ly-49A. Specific single amino acid changes in either the H2-D(d) alpha(1) helix or the alpha(2) helix abrogated Ly-49D-mediated cytotoxicity, but these changes had no significant effect on Ly-49A-dependent inhibition. Each of three alpha(2) domain mutations in the floor of the peptide binding groove reduced functional recognition by either Ly-49D or Ly-49A, but all three were required to fully abrogate inhibition by Ly-49A. Our studies indicate that Ly-49D/H2-D(d) interactions require distinct determinants compared with Ly-49A/H2-D(d) interactions. These differences have important implications for the integration of activating and inhibitory signals in NK cells.

Figure 1

Functional recognition by Ly-49D requires both the α₁ and α₂ domains of H2-D^d. Cytotoxicity by RNK.Ly-49D (left) and RNK.Ly-49A (right) effectors was tested in the presence of blocking 12A8 F(ab′)₂ fragments (♦, anti-Ly-49A/Ly-49D), isotype-matched control F(ab′)₂ fragments (•, 2C7), or media (□). Targets were YB2/0 cells transfected with H2-D^d or H2-D^d/b chimeric molecules as indicated.

Figure 2

Location of mutated amino acid residues in H2-D^d. The molecular graphics image was created by H. Houtkooper, University of California, San Francisco, Computer Graphics Laboratory, with the MidasPlus modeling system. The image is based on Protein Data Bank coordinates 1bii 26.

Figure 3

Recognition of H2-D^d by Ly-49D but not by Ly-49A is highly sensitive to specific single amino acid changes in either the α₁ domain or the α₂ domain of H2-D^d. RNK.Ly-49D (right) and RNK.Ly-49A (left) effectors were tested for lysis of YB2/0 targets transfected with intact H2-D^d, H2-D^b, or H2-D^d mutated to the corresponding residue of H2-D^b as indicated in the left-hand column. Lysis of target cells is shown at an E/T ratio of 50:1. Lysis by RNK.Ly-49A effectors was assessed in the presence of anti–Ly-49A (A1, black bars) or no blocking mAb (spotted bars). Lysis by RNK.Ly-49D effectors was assessed in the presence of anti–Ly-49D F(ab′)₂ fragments (12A8, hatched bars) or no antibody (white bars).

Figure 4

Mutations in the floor of the peptide binding groove of H2-D^d abrogate functional recognition of H2-D^d by both Ly-49D and Ly-49A. Lysis by RNK.Ly-49D (right) and RNK.Ly-49A (left) effectors were tested in the presence of blocking 12A8 F(ab′)₂ fragments (♦, anti-Ly-49A/Ly-49D) or no antibody (□) against YB2/0 targets expressing H2-D^d or mutated H2-D^d as indicated in the left-hand column. In panels E and J, H2-D^d is mutated at all three sites (97, 99, and 114).

Figure 5

Mutations in H2-D^d demonstrate that functional recognition by Ly-49A and Ly-49D does not strictly correlate with binding by the mAb 34-5-8S. YB2/0 tumor target cells expressing H2-D^d or H2-D^d mutated at the indicated residues were stained with an anti–H2-D^d mAb (34-2-12S [α₃] or 34-5-8S [α_1,2]). Overlay (black line) shows staining of YB2/0 cells transfected with H2-D^b as background which was similar to staining of untransfected YB2/0 cells (not shown). Transfectants were initially selected for equivalent expression as assessed by 34-2-12S binding (A, C, E, G, I, K, M, and O, gray histogram) then assessed for 34-5-8S binding (B, D, F, H, J, L, N, P, gray histogram). For comparison, relative inhibition by Ly-49A and activation by Ly-49D are shown at the right.