Think of the bone marrow as the body’s factory for blood and immune cells. In myelodysplastic syndromes (MDS), that factory breaks down—producing too few cells, and the ones that do roll off the line often don’t work properly. This ineffective production leaves patients vulnerable to anemia, infections, and bleeding, and in some cases the disease progresses to acute myeloid leukemia (AML). The major culprits behind this breakdown are acquired mutations in spliceosome genes—the cellular machinery that processes RNA. These mutations take hold in the stem and progenitor cells that supply the marrow’s factory floor, setting the stage for myeloid malignancies. Unlike many cancer mutations that differ from patient to patient, changes in the spliceosome factor subunit B1 (SF3B1) gene show up in more than 20% of MDS cases at conserved hotspots.
One particular alteration, SF3B1K700E, is especially common and has become an attractive therapeutic target. Because this exact mutation arises in a large subset of MDS patients, it is an ideal candidate for developing a "shared" immunotherapy that could benefit many patients. The mutation creates what scientists call a "neoantigen" – a new protein fragment that doesn't exist in healthy cells but appears in cancer cells. T cells, the immune system's highly skilled surveillance agents, can potentially recognize these fragments as foreign and eliminate the cancer cells that carry them. To explore whether this vulnerability could be harnessed therapeutically, Jessica Lok, Dr. Melinda Biernacki, Dr. Marie Bleakley and their colleagues in the Bleakley lab conducted a series of investigations, with their results recently reported in Cancer Immunology Research.
“One of the challenges in treating myeloid malignancies like AML and MDS is that they are very heterogeneous, genetically as well as clinically. So, while neoantigens avoid the problem of "on-target, off-tumor" toxicity that could lead to myeloablation, the genetic heterogeneity of AML and MDS means that there's no single neoantigen target that will apply to all or even most patients,” shared Biernacki. “The goal is to build a “toolbox” of TCRs so that there is a TCR-T therapy for every patient. Spliceosome genes are recurrently mutated in AML and MDS, and our identification of this SF3B1K700E neoantigen, along with our prior identification of a U2AF1 neoantigen, suggests that TCR-T could be used for patients with these types of mutations.”
Biernacki, Lok and colleagues used a systematic approach to identify and validate their target. They first employed computational algorithms to predict which fragments of the mutated SF3B1 protein might be presented on the surface of cancer cells in a way that T cells could recognize. The team identified a candidate 9 amino acid neoantigen, QEVRTISAL (single letter amino acid code, mutated from QKVRTISAL), from the prevalent SF3B1K700E variant using in silico predictions of epitope processing and presentation. From these predictions, the researchers isolated a CD8+ T-cell clone, called D1.C24, that recognized the SF3B1K700E-derived peptide with remarkable sensitivity. Even when the peptide was present at vanishingly low levels, D1.C24 could detect and kill the target cells. Equally important, the clone was highly specific: it ignored the normal, unmutated version of the protein as well as related peptide sequences, responding only to the cancer-associated fragment when presented by the HLA-B*40:01 molecule, which acts as the “display case” that presents the peptide to the T cell.
The team next asked whether this recognition happened in a real disease setting, not just in prediction models or engineered cells. D1.C24 successfully eliminated leukemia cell lines carrying both the SF3B1K700E mutation and HLA-B*40:01, and it responded to malignant cells taken directly from a patient sample. These experiments confirmed that the mutation produces a bona fide neoantigen—a genuine marker that is naturally processed and displayed by cancer cells in MDS and AML.
While patients themselves did not appear to mount strong natural immune responses against SF3B1K700E, the researchers overcame this limitation by cloning the D1.C24 T-cell receptor and introducing it into healthy donor T cells. These engineered TCR-T cells retained the ability to specifically recognize and kill SF3B1K700E-expressing cancer cells in laboratory assays and showed antileukemic activity in a mouse model. Together, these findings highlight SF3B1K700E as a promising shared immunotherapy target and demonstrate the potential of TCR-based therapies to expand treatment options for patients with MDS and AML.
Explaining the significance of their work, Lok said, "Our work outlines a multi-step approach that can be applied to identifying neoantigens created from other recurrent mutations to expand the “toolbox”. We hope to see this method applied to other neoantigens in the future.” Biernacki added that the implications could extend beyond blood cancers. “In addition to treating AML and MDS, our SF3B1K700E-specific TCR-T could also help patients with other kinds of cancer that have the same mutation, including chronic lymphocytic leukemia, pancreatic cancer, renal cancer, and breast cancer."
Fred Hutch/University of Washington/Seattle Children’s Cancer Consortium Members Drs. Melinda Biernacki, Vivian Oehler, Derek Stirewalt, David Wu, Sergei Doulatov, and Marie Bleakley contributed to this research.
The spotlighted research was funded by the National Institutes of Health, MPN Research Foundation, the Leukemia and Lymphoma Society, the Be The Match Foundation, and the Gerdin Family.
Biernacki MA, Lok J, Foster KA, Cummings C, Busch S, Black RG, Ray S, Galvis LB, Monahan T, Oh ST, Oehler VG, Stirewalt DL, Wu D, Deeg HJ, Doulatov S, Bleakley M. 2025. SF3B1K700E Neoantigen Is a CD8+ T-cell Target Shared across Human Myeloid Neoplasms. Cancer Immunology Research. https://doi.org/10.1158/2326-6066.CIR-24-0091.