Towards understanding of NK cell antigenic specificity
Ustiuzhanina, M. O.; Shagina, I. A.; Nikitin, E.; Klimuk, E.; Britanova, O.; Ventura-Carmenate, Y.; Kovalenko, E.; Chudakov, D. M.
Show abstract
NK cells can form clonal populations demonstrating features of adaptive immunity, including long-term memory and at least partial antigenic specificity. Given the limited individual diversity of activating receptors, the nature of NK cell antigenic specificity remains elusive. To explore this riddle, we combined scRNA-Seq of ex vivo FACS-sorted NK cell subsets expressing specific KIR receptors, single-cell cloning and bulk RNA-Seq of in vitro cultured KIR2DS4 NK cell clones, transcriptomic profiling of antigen-stimulated NK cells, and in silico modeling of glycosylated KIR2DS4-peptide-HLA complexes. scRNA-Seq resolved 12-15 clusters per KIR subset with highly heterogeneous KIR, KLRC and NCR expression patterns, consistent with clonal lineages. Notably, those clusters demonstrated over 30 differentially expressed glycosyltransferase genes, potentially involved in post-translational modification of NK cell receptors. Single-cell-derived KIR2DS4 cultures exhibited clone-specific cytotoxic, chemokine and KIR receptor genes, and transcriptional differences in > 40 glycosyltransferases. In peptide culturing autologous assays, SARS-CoV-2 (KTFPPTEPK) and EBV (CRAKFKHLL) peptides elicited NK cell proliferation and distinct transcriptional programs linking cytotoxicity genes, KIR2DS4 and glycosyltransferases. Structural modeling revealed that N-linked glycosyl residues in specific regions of KIR2DS4 may alter its contacts and interaction with MHCI and the presented peptide. We conclude that KIR human NK cells comprise clonally imprinted populations with distinct glycosyltransferase expression profiles, and site-specific KIR2DS4 glycosylation may modulate interaction with peptide-MHCI complexes, suggesting a post-translational layer of clonal NK cell diversification as a clue to their antigenic specificity.
Matching journals
The top 5 journals account for 50% of the predicted probability mass.