Abstract
XPO1 (Exportin-1/CRM1) is a nuclear export protein that is frequently overexpressed in cancer and functions as a driver of oncogenesis. Currently small molecules that target XPO1 are being used in the clinic as anticancer agents. We identify XPO1 as a target for natural killer (NK) cells. Using immunopeptidomics, we have identified a peptide derived from XPO1 that can be recognized by the activating NK cell receptor KIR2DS2 in the context of human leukocyte antigen–C. The peptide can be endogenously processed and presented to activate NK cells specifically through this receptor. Although high XPO1 expression in cancer is commonly associated with a poor prognosis, we show that the outcome of specific cancers, such as hepatocellular carcinoma, can be substantially improved if there is concomitant evidence of NK cell infiltration. We thus identify XPO1 as a bona fide tumor antigen recognized by NK cells that offers an opportunity for a personalized approach to NK cell therapy for solid tumors.
INTRODUCTION
Natural killer (NK) cells are becoming increasingly recognized for their anticancer activity, and their functions are tightly controlled by a diverse repertoire of cell surface receptors (1). One important family of NK cell receptors are the killer cell immunoglobulin-like receptors (KIRs), which form a polymorphic family of receptors with human leukocyte antigen (HLA) class I ligands (2). KIRs have been implicated in susceptibility to, and the outcome of, many different cancers and thus contribute to heterogeneity within the innate immune response to malignancy (3–9).
The KIR can be activating or inhibitory. The HLA class I ligand specificities of the inhibitory KIR are relatively well defined, while the ligand specificities of the activating KIR have been much harder to identify (10). Recent work has shown that activating KIR can have an HLA class I–restricted peptide specificity (11–15). While T cell receptors have a tight restriction on the peptide:HLA complexes that they bind, KIRs recognize families of peptide:HLA complexes in a motif-based manner allowing recognition of multiple different combination of peptide and HLA (16, 17). We have recently shown that KIR2DS2 recognizes highly conserved flavivirus and hepatitis C peptides with an alanine-threonine sequence at the C-terminal −1 and −2 positions of the peptide in the context of HLA-C (12, 18). We have also recently shown that targeting KIR2DS2 in vivo using a peptide:HLA-C–based strategy can increase NK cell activity against cancer (19). In these experiments, KIR2DS2-positive NK cells in the KIR transgenic mouse were activated using a DNA construct that encoded both HLA-C*0102 and the peptide IVDLMCHATF, which together provide a ligand for KIR2DS2.
Activating KIR have been associated with protective responses against a number of cancers, especially hematological malignancies. In genetic studies, KIR haplotypes containing activating KIR confer protection against relapse of hematological malignancies following bone marrow transplantation, and this has been mapped to the region of the KIR locus that contains KIR2DS2 (20–22). A recent study using sequence-based KIR typing has mapped this protection to either the KIR2DS2 or KIR2DL2 genes that are in tight linkage disequilibrium within the KIR locus (23). Furthermore, the benefit of cord blood transplantation is accentuated if the recipient of the transplant has group 1 HLA-C allotypes, the ligands for KIR2DS2 (24). KIR2DS2 has also been associated with protection against a number of solid tumors including cervical neoplasia, breast cancer, lung cancer, colorectal cancer, and hepatocellular carcinoma (HCC) (4, 5, 25–27). These genetic studies lack functional correlates, and although an in vitro study demonstrated recognition of cancer cell lines by KIR2DS2, this was not specific and also encompassed recognition by the inhibitory receptor KIR2DL3 (28). As KIR2DS2 recognizes the combination of HLA-C and peptide in a motif-based manner (12), this suggests that there is a potential for KIR2DS2 binding peptides to be tumor-associated antigens derived from endogenously expressed proteins that are up-regulated during tumorigenesis. However, to date, no cancer-associated peptides that bind activating KIR have been identified, and we therefore sought to investigate this in a proof-of-concept study.
RESULTS
XPO1 provides a peptide recognized by KIR2DS2
KIR2DS2 is a peptide:HLA-specific receptor, and our previous work had indicated that the human HCC cell line Huh7 transfected with HLA-C*01:02 (Huh7:C1) may be recognized by KIR2DS2 (19, 29). Parental Huh7 cells express HLA-A*11:01, but HLA-B and HLA-C are undetectable on the cell surface (30). HLA-A*11:01 binds peptides predominantly with C-terminal lysine or arginine residues and HLA-C*01:02 binds peptides with C-terminal hydrophobic residues. This distinction provided an opportunity for a robust distinction of peptides presented by the two different HLA alleles expressed on the transfected cell line. This allowed identification of potential KIR2DS2-binding peptides expressed on HLA-C*01:02. We sequenced HLA class I–bound peptides from Huh7 and Huh7:C1 cells following elution with the pan class I antibody W6/32. We identified ~5800 and ~7200 peptides from untransfected Huh7 and Huh7:C1 cells, respectively (table S1). As expected, the untransfected Huh7 immunopeptidome consisted mainly of peptides with C-terminal lysine or arginine residues, characteristic of the HLA-A*11:01 binding specificity (Fig. 1A). In contrast, epitopes identified from the Huh7:C1 cell line contained peptides that displayed two distinct sequence motifs, one with the previously observed HLA-A*11:01–binding motif and one with an HLA-C*01:02–binding motif (Fig. 1A). From this comprehensive repertoire of HLA class I binders, only one peptide, NAPLVHATL, was identified that conformed to both the HLA-C*01:02–binding motif (xxPxxxxxL) and the KIR2DS2-binding motif (xxxxxxATx). NAPLVHATL, which was identified by three independent, high-quality peptide-spectrum matches (Fig. 1B), is derived from the nuclear export protein XPO1 (Exportin-1/CRM1). This protein is up-regulated in hematological malignancies and solid tumors including lung cancer, breast cancer, and HCC. It is also associated with a poor prognosis in cancer and is targeted by selinexor, a Food and Drug Administration–approved drug for multiple myeloma and diffuse large B cell lymphoma (31–37). Data-independent acquisition (DIA) mass spectrometry was subsequently used to quantify NAPLVHATL and other epitopes between the transfected Huh7:C1 and the parental Huh7 cell line (table S2). The peptide NAPLVHATL was eluted at 15-fold greater levels in Huh7:C1 compared to Huh7 cells, consistent with it being presented by endogenous HLA-C*01:02 but at much lower levels. In addition, this peptide has been eluted from HLA class I on cirrhotic livers and a derivative APLVHATL identified from an HCC sample, indicating that this peptide is naturally presented and processed in vivo in the liver (38). Furthermore, we did not identify any additional XPO1-derived peptides expressing the HLA-A*1101 motif and, in netMHCpan analysis of the XPO1 protein, there were no predicted HLA-A*1101 peptides with the KIR2DS2-binding motif described by Liu et al. (39)….
