Successful immunotherapies for childhood cancers typically target lineage-, rather than cancer-specific markers, B-cell specific CD19 being the prime example. These successes culminated in the recent FDA approval of bi-specific T-cell engagers and chimeric antigen receptor (CAR)-armed T-cells for B-cell acute lymphoblastic leukemia (B-ALL). However, relapses frequently occur in patients treated with CD19- directed immunotherapies, often due to epitope loss. While strategies based on dual antigen targeting are beginning to emerge, they are still based on targeting canonical B-cell markers, with the unavoidable side effect of total ablation of normal B-cells. The apparent paucity of tumor-specific targets in pediatric cancers (including leukemias) is likely to limit future immunotherapies. We hypothesize that alternative splicing could be both a mechanism of epitope loss and a rich source of a neo-antigens in B-ALL. Indeed, using computational and biochemical approaches, we have identified hundreds of local splicing variations (LSVs), mapping to transmembrane proteins (e.g., CD19 and CD22) with prominent extracellular domains (ectodomains). In fact, in our prior work we described a mechanism of acquired resistance to CART-19 based on selective loss of the CD19 ectodomain, primarily through exon 2 skipping. The two large Aims of this U01 are as follows.
Aim 1 : To identify programs and determinants of altered splicing of B-ALL cell surface antigens. Our overarching goal is to construct a dedicated ?splicing code? for leukemic B-cells. Such a code will predict cis-acting genetic variants as well as trans-acting factors involved in alternative splicing and allow us to identify all alternatively spliced ectodomains. Then in Aim 2, we will investigate the effects of alternative splicing on B-ALL immunotherapy. Using CD22 as just one example, we will determine how truncated protein isoforms confer resistance to CD22-targeting immunotherapeutics, including antibody- drug conjugates such as inotuzumab ozogamicin, which was recently approved by FDA to treat relapsed or refractory B-ALL. We will also raise antibodies against peptides spanning novel exon junctions, generate antibody-drug conjugates, and test their efficacy against B-ALL cell lines and patient-derived xenografts. In summary, this leukemia-based U01 will create new computational and conceptual frameworks, which would be highly synergistic with Pediatric Immunotherapy Discovery & Development Network (PI-DDN) overall goals, including identification of antigenic epitopes that are uniquely expressed on childhood cancers and of cancer cell-intrinsic mechanisms of immune evasion.
Pediatric cancers, especially hematologic malignancies, typically have low mutation burden, which could make them invisible to the immune system, trained to recognize non-self antigens. We hypothesize that in B-cell acute lymphoblastic leukemia (B-ALL) there exists another important source of neo-antigens, namely alternative splicing. By constructing and validating a dedicated B-ALL splicing code we will lay ground for the development of new immunotherapeutics that would target pediatric cancers with the specificity current modalities do not possess.