Large Scale Identification and Validation of 85 New HLA Class I-Restricted Minor Histocompatibility Antigens for Immune Monitoring and Modulation after Allogeneic Stem Cell Transplantation

Background: Allogeneic stem cell transplantation (alloSCT) is a curative treatment for hematological malignancies, but patients may experience relapse of their disease or Graft-versus-Host Disease (GvHD). After transplantation, GvHD as well as Graft-versus-Leukemia (GvL) reactivity are caused by donor T cells recognizing minor histocompatibility antigens (MiHAs) on patient cells. MiHAs are HLA binding peptides that are encoded by genetic differences in single nucleotide polymorphisms (SNPs) between patient and donor. Dependent on whether the antigen is presented on tumor cells or healthy non-hematopoietic tissues of the patient, donor T cells may induce the favorable GvL effect or GvHD, respectively. MiHAs are fundamental to understand and modulate immune responses in patients after transplantation. However, only 65 in vivo immunogenic MiHAs have been identified in the last 30 years, which is insufficient for immune monitoring and modulation.

Aims: To identify and characterize the dominant repertoire of MiHAs in seven common HLA class I molecules (HLA-A*01:01, A*02:01, A*03:01, B*07:02, B*08:01, C*07:01, C*07:02) for immune monitoring and modulation after alloSCT.

Methods: Activated donor derived CD8+ T cells were single cell sorted from 40 patients who developed GvL reactivity in the absence or presence of GvHD after donor lymphocyte infusion following T cell-depleted, HLA-matched alloSCT. Growing T cell clones were tested for reactivity against patient and donor cells, and patient-specific T cell clones were tested against an optimized panel of 191 selected EBV transformed B cell lines, which have been sequenced as part of the 1000 Genomes Project. This panel enables identification of MiHAs presented by seven common HLA class I molecules (HLA-A*01:01, A*02:01, A*03:01, B*07:02, B*08:01, C*07:01, C*07:02) by genome wide association study (GWAS) extending SNP coverage to 11 million (MAF > 0.01). SNPs that strongly associated with T cell recognition patterns were subsequently validated to encode MiHAs.

Results: We identified 85 new MiHAs presented by seven common HLA class I molecules with our GWAS method, thereby more than doubling the previously known repertoire of 65 MiHAs. We now show that 111 (74.0%) of the total 150 MiHAs are derived from annotated ORFs of protein-coding genes, whereas 39 (26.0%) of MiHAs are cryptic out-of-frame antigens or antigens translated from SNPs in 5’ UTR or non-coding regions (Figure 1A). This demonstrates that one quarter of total antigens targeted in effective anti-cancer immune responses are cryptic antigens that are often ignored in immune monitoring studies.

Our large scale identification also shows that antigens targeted in immune responses after transplantation are often shared, i.e. recurrently found in multiple patients. In total, 177 distinct T cell clones for 109 different MiHAs were isolated in 40 patients. Strikingly, 44 (40.4%) MiHAs were shared and recognized by 112 (63.3%) of the 177 T cell clones (Figure 1B). We also measured SNP mismatches for all 150 MiHAs by whole exome sequencing and showed that that number of SNP mismatches for known MiHAs in patients with limited (n = 7) or severe GvHD (n = 20) were in the same range as in patient without GvHD (n = 13). However, the proportion of SNP mismatches for known MiHAs that were demonstrated to be targeted by T cells were significantly higher in patients with GvHD.

Finally, we investigated tissue distribution and SNP disparities of all 150 MiHAs to identify potential targets for immunotherapy based on whether they are selectively expressed by hematopoietic cells and frequently mismatched. In total 12 hematopoietic-restricted MiHAs were found of which 6 antigens are novel.

Summary/Conclusion: In conclusion, despite many SNP mismatches between patients and donors, our data demonstrate that the repertoire of MiHAs is often shared, confined and unconventional. Our large collection of MiHAs is fundamental to predict, follow or manipulate immune responses after allogeneic stem cell transplantation to improve clinical outcome of transplanted patients.

Griffioen:Miltenyi Biotec B.V & Co. KG: Research Funding.