DDX3Y encodes a class I MHC-restricted H-Y antigen that is expressed in leukemic stem cells

Abstract

The Y chromosome encodes male-specific minor histocompatibility (H-Y) antigens that stimulate T- and B-lymphocyte responses after sex-mismatched allogeneic hematopoietic cell transplantation (HCT). A CD8(+) cytotoxic T lymphocyte (CTL) clone that recognizes a novel HLA-B*2705-restricted H-Y antigen encoded by the DDX3Y gene was isolated from a male who had received a hematopoietic cell graft from his human leukocyte antigen (HLA)-identical sister. The antigenic peptide is a decamer that differs from the homologous DDX3X-encoded peptide at 4 positions. Expression of DDX3Y and of the H-Y epitope that it encodes was examined by quantitative polymerase chain reaction (PCR) and by CTL recognition assays. Expression of DDX3Y is detected in all myeloid and lymphoid leukemic cells that carry an intact Y chromosome. Moreover, the DDX3Y-encoded H-Y epitope is presented on the surface of both myeloid and lymphoid leukemic cells from male HLA-B*2705(+) patients. DDX3Y-specific CTLs prevent engraftment of human acute leukemia in nonobese diabetic/severe combined immune deficient mice, demonstrating that the DDX3Y-encoded H-Y antigen is also expressed in leukemic stem cells. These results demonstrate that CD8(+) T-cell responses against DDX3Y have the potential to contribute to graft-versus-leukemia (GVL) activity after female into male allogeneic HCT. This study is registered at https://ichgcp.net/clinical-trials-registry/NCT00107354" title="See in ClinicalTrials.gov">NCT00107354.

Figures

Figure 1

CTL 68H7–819 recognizes a male-specific minor histocompatibility (H-Y) antigen presented by HLA-B*2705. (A) 51Cr release assay at the indicated effector-target (E/T) ratios showing cytolytic activity of CTL 68H7-819 against donor- and recipient-derived EBV-LCLs, recipient-derived unfractionated PBMCs and PHA-stimulated T-cell blasts, and donor- and recipient-derived dermal fibroblasts. (B) 51Cr release assay at E/T 5:1 showing cytolytic activity of CTL 68H7-819 against a panel of EBV-LCLs derived from unrelated male and female individuals who shared a single MHC class I allele with the donor-recipient pair from which CTL 68H7-819 was isolated. The shared MHC class I allele and sex of the individual is indicated.

Figure 2

DDX3Y encodes the H-Y antigen recognized by CTL clone 68H7-819. (A) Localization by Y-chromosome deletion mapping of the gene encoding the H-Y antigen recognized by CTL 68H7-819. The CTLs were tested for recognition of EBV-LCLs derived from males carrying Y chromosomes with constitutional deletions in a 51Cr release assay at E/T 10:1. EBV-LCLs were infected the day before the assay with a recombinant vaccinia virus carrying an HLA-B*2705 transgene. EBV-LCL WHT2996 is derived from an individual with a deletion encompassing genes DDX3Y and USP9Y. WHT2780 is derived from an individual with a splice site deletion that results in 90% truncation of the USP9Y gene, indicated by an X. WHY26 is from an individual with a chromosomal break in UTY. Arrows indicate the intact and aberrant segments of the Y chromosome. Y-chromosome landmarks include the boundaries of deletion intervals 5C and 5D, and selected sequence-tagged sites. + indicates lysis of 35% or more, and − indicates lysis of 4% or less. (B) Plasmids encoding DDX3Y minigenes were cotransfected into COS-7 cells with a plasmid encoding HLA-B*2705. On the following day, CTL 68H7-819 was added to the COS-7 transfectants, and IFN-γ release was measured in the supernatants by ELISA after 20 hours of coculture. (C) Epitope reconstitution assay to determine CTL 68H7-819 recognition of donor EBV-LCLs that had been pulsed for 30 minutes with the indicated synthetic peptides over the indicated range of concentrations; 4-hour 51Cr release assay, E/T 5:1. (D) Partial sequence alignment of the DDX3Y and DDX3X proteins spanning the region that includes the epitope recognized by CTL 68H7-819. Asterisks indicate disparate residues.

Figure 3

DDX3Y is transcribed outside the testis and is universally expressed in myeloid and lymphoid leukemia cells that carry a Y chromosome. Relative expression of DDX3Y in a panel of normal human tissues (A), normal blood cell fractions (B), and primary male ALL, CLL, and AML samples (C). Quantitative real-time PCR using SYBR green was carried out as described in “Methods.” Analysis of GAPDH expression was used to standardize samples for RNA quality and quantity, and the relative DDX3Y expression in the KG-1a AML cell line was arbitrarily defined as 1. The asterisk in panel C indicates a primary male ALL sample that had clonal loss of the Y chromosome by cytogenetic analysis and no detectable DDX3Y transcript. Five of 14 female samples are shown, all of which were negative.

Figure 4

The DDX3Y-encoded H-Y antigen recognized by CTL 68H7-819 is expressed on the surface of HLA-B*2705+ leukemic cells. CTL 68H7-819 recognition of primary leukemic samples from males carrying the HLA-B*2705 allele was tested by IFN-γ ELISA after overnight coculture with CTL at a CTL/leukemia ratio of 1:5. The 6 samples consisted of BMMCs derived from the hematopoietic cell transplant recipient from whom CTL 68H7-819 was derived (Recip ALL), PBMCs from one patient each with primary refractory AML and with CML in T-lymphoid blast crisis (CML-BC), and 3 PBMC samples from CLL patients. Negative controls included a CLL sample that did not carry the HLA-B*2705 allele, and CTL 68H7-819 alone.

Figure 5

CTL 68H7-819 targets the leukemic stem cell in T-lymphoid blast phase CML (CML-BC). (A) Recognition of CML-BC cells by CTL 68H7-819 in a 51Cr release assay at the indicated E/T ratios. (B) Survival analysis of mice injected with CML-BC cells cultured overnight in medium alone or with CTL 68H7-819. Each group was composed of 5 mice. (C) Flow cytometric analysis using human PE-conjugated antihuman CD34 and FITC-conjugated anti–HLA-B27 antibodies of BMMCs from representative mice injected with CML-BC cells that had been cultured overnight in medium alone or with CTL 68H7-819. Uninjected CML-BC cells were used as a positive control. (D) Human Y-chromosome PCR analysis of genomic DNA extracted from mouse BMMCs to detect human male leukemic cells. Flow cytometric and PCR analysis was performed on mice that survived at least 15 days after injection.

Figure 6

CTL 68H7-819 targets the leukemic stem cell in AML. Flow cytometric analysis using human PE-conjugated antihuman CD45 and FITC-conjugated anti–HLA-B27 antibodies of BMMCs from representative mice injected with PBS (sham), or with AML cells that had either been cultured overnight in medium alone, with CTL 68H7-819, or with an irrelevant CD8+ CTL clone that did not recognize the AML in vitro. Top panels show human Y-chromosome–specific PCR analysis of genomic DNA isolated from BMMCs (BM), peripheral blood (PB), spleen (Sp), thymus (Thy), lung (Lg), liver (Lv), and kidney (Kid). The flow cytometry and PCR data from 2 representative mice from each group are shown, and are designated as A or B.

Figure 7

DDX3Y peptide–specific T-cell responses in a F→M hematopoietic cell transplant recipient experiencing GVL. PBMCs obtained from UPN 21234 (HLA-A*01, -A*02, -B*07, -B*15, -DRB1*0401/0404, and -DQB1*03) on day +127 after transplantation were stimulated in vitro with a pool of 36 overlapping DDX3Y-derived pentadecapeptides pulsed onto donor-derived EBV-LCLs and subsequently analyzed by IFN-γ ELISpot for DDX3Y peptide–specific responses (Tables 1, 2). (A) ELISpot analysis of day +127 PBMCs that had been stimulated in vitro with the entire pool of 36 DDX3Y-derived pentadecapeptides. Asterisks indicate the 4 peptide pools that stimulated spot formation above background. (B) An aliquot of the T-cell line analyzed in panel A was restimulated with pentadecapeptide no. 17, and subsequently analyzed by ELISpot for reactivity with the SRDSRGKPGY decamer, encompassed within the sequence of pentadecapeptide no. 17.