What T cells see in WT-1

The Wilms tumor protein, WT-1, is a widely recognized tumor antigen that is aberrantly expressed in myeloid and lymphoid leukemia and in this issue of Blood, Doubrovina et al report the most extensive catalog heretofore of HLA-restricted immunogenic peptides derived from WT-1, which are recognized by CD8 and CD4 T cells.¹ 

The investigators used a pool of overlapping 15-mer peptides from WT-1 to screen for cytokine production from autologous T cells from the blood of 56 healthy donors, and to subsequently identify the peptide sequences and HLA restrictions of the immunogenic epitopes. The authors identified 42 peptide epitopes from WT-1 including 41 new epitopes restricted to class I and class II HLA molecules. Importantly, peptide-specific T-cell responses were evident in nearly 80% of the donors, and T-cell lines specific for 29 of the 42 epitopes induced specific lysis of WT-1–expressing leukemic blasts. These results show that a large number of potentially immunogenic peptides are naturally processed and presented from WT-1 on the surface of leukemic blasts. The results suggest that these epitopes could be useful in immunotherapy strategies that target WT-1–expressing malignancies, including leukemia.

Wilms tumor is the most common renal neoplasm of children and 50% of individuals carrying a germ line mutation predisposing to Wilms tumor develop the disease. The WT-1 protein is a zinc finger transcription factor that is important in embryonic kidney and early genitourinary development. It acts as a tumor suppressor or oncogene depending on the cell type and promoter context. The WT1 gene is composed of 10 exons that code for multiple isoforms of the WT-1 protein, with molecular weights of 45 to 49 kDa. Four major isoforms of WT-1 are the result of alternative splicing, while 8 minor isoforms result from different initiation sites. Some isoforms have a role in RNA processing rather than transcription regulation. Thus, WT-1 may be a critical antigen in the initiation or maintenance of a malignant phenotype.

WT-1 is a promising target antigen in many malignancies including leukemia because it is differentially expressed in a broad number of malignant cells. WT-1 also is a useful marker of residual myeloid leukemia, and it may be expressed in a subset of leukemia-initiating cells.^2,3  Until the current study, a modest number of WT-1 epitopes confined mostly to the HLA-A2, HLA-A24, and HLA-DR4 alleles had been identified as immunogenic.^4–6  Nevertheless, the results of clinical immunotherapy trials in leukemia patients have shown that T-cell responses can be induced against these epitopes and objective clinical responses have been observed in some patients. Therefore, it is reasonable to expect that if further epitopes that are restricted by additional HLA alleles could be identified, targeted immunotherapy could be extended to more patients, which may be clinically beneficial. Moreover, targeted immunotherapy approaches, such as those that target WT-1, are potentially less toxic than conventional treatments due to the specificity of the immune response and the absence of significant toxicity in the antigen-targeted therapy trials to date.

Nevertheless, there are caveats to the exciting potential for directly translating these findings to the clinic. While adaptive immunotherapy strategies necessarily target specific epitopes, identifying the relevant antigens is not the only, nor the most significant, obstacle to the development of new immunotherapy treatments for leukemia and other cancers. For example, even when we understand which epitopes are targeted on leukemic blasts, to eliminate leukemia the antigens must be expressed on the leukemia stem cell and the therapy must preferentially eliminate leukemia stem cells over normal hematopoietic stem cells. In addition, we must increase our understanding of how tolerance to these antigens is regulated. In particular, both central (thymic) and peripheral (lymph node) regulation of immunity to tumor-associated antigens such as WT-1, a self-antigen, must be understood. The role of immune checkpoint regulation, governed by interactions of surface molecules on antigen-presenting cells and T cells such as CD80/CTLA-4 and PD-L1/PD-1, are clinically highly relevant molecules that alter the threshold of T-cell activation, thereby favoring a milieu in which immunity to tumor antigens can develop. The role of regulatory T cells and B cells, and of myeloid-derived suppressor cells (MDSCs), is also critical for maintaining tolerance to tumor antigens. In addition, optimal strategies for therapeutically delivering the antigens (eg, antigen as peptide, protein, DNA, or cell-based system) and in what immunologic context (eg, modification of tolerance) and clinical context (eg, minimal residual disease state) still must be elucidated.

Over the past decade the search for more effective and less toxic therapies in oncology, and particularly in hematologic malignancies, has lead to an enormous emphasis on the discovery and development of novel targeted therapies, often involving small molecule inhibitors that are highly specific for target molecules aberrantly expressed in malignant cells. The promise of this strategy naturally grew from the highly successful discovery that treatment of chronic-phase chronic myelogenous leukemia (CML) with tyrosine kinase inhibitors produced lasting molecular remissions in the majority of patients, and with far fewer side effects compared with standard chemotherapy.⁷  It logically followed that these results might be duplicated in other cancers once similar “drugable” target molecules, especially those resulting from analogous “driver” mutations, could be identified in the malignant cells. This approach was emboldened by the revolution in low-cost, high-throughput genetic and epigenetic mapping strategies, which have been used to catalog lesions in primary and metastatic malignant cells. While we are still in the early days of discovering and cataloging these lesions, new drugs have already resulted from this approach, although we have yet to reproduce the stunning success of imatinib for CML in other cancers. In part, the reason for our limited success is due to the degeneracy of cell-signaling pathways in tumors, the absence of an “oncogene-addiction” to the mutated proteins in the malignant cells, and the overlap of targeted cellular pathways with pathways that are critical in normal, healthy cells.

Like the search for small molecule inhibitors of mutated oncogenes, the search for additional epitopes for immunotherapy is critical for developing future treatment strategies, but it must be taken in context. As the redundancy of oncogenes and signaling pathways in malignant and normal cells is critical for clinical successes of targeted small molecules, so, too, is the redundancy of target antigen recognition and immune regulation for the success of targeted immunotherapy. Strategies to transfer expanded populations of antigen-specific T cells derived from healthy donors in the allogeneic stem cell transplant setting might be one way to activate T cells ex vivo. Other promising strategies for addressing the potential problem of reaching activation threshold of T cells is by gene modifying normal polyclonal T cells to express T-cell receptors specific for tumor antigens,⁸  chimeric antigen receptors (CARs) comprising immunoglobulin single chain Fv fused to CD3 and CD28 endodomains,⁹  or to express a CD3 signaling domain fused to monoclonal T-cell receptors against cancer (ImmTACs) that are specific for tumor antigen–derived peptide/HLA surface antigens.¹⁰  Such adoptive T-cell therapy strategies rely on the successful identification of tumor target antigens, however, and the study by Doubrovina et al is an important first step in these and similar approaches to the treatment of leukemia and other malignancies.