Search for a command to run...
Abstract The anti-apoptotic B-cell lymphoma 2 (BCL-2) protein, a key regulator of cell death, is frequently overexpressed in hematological malignancies, including >80% of acute myeloid leukemia (AML) cases (Fowler-Shorten Blood Rev 2024). Intensive chemotherapy and allogeneic hematopoietic stem cell transplantation (HCT) remain the potentially curative standard of care. However, relapse after HCT is common and older or fragile individuals are often not eligible for harsh regimens.As an emerging alternative, Venetoclax - a selective BCL-2 inhibitor - restores apoptotic signaling in leukemic cells. Due to reduced toxicity, Venetoclax-based regimens have shown strong clinical efficacy for frail patients with an otherwise dismal prognosis (DiNardo N Engl J Med 2020), as well as in post-HCT relapse prevention. However, it frequently displays dose-limiting hematotoxicity.Here we explored whether engineering BCL-2 inhibitor resistance into transplanted hematopoietic stem and progenitor cells (HSPCs) could overcome this limitation. We previously showed that precise epitope engineering enabled the selective eradication of tumor cells while preserving healthy cells from targeted antigen-specific therapies, such as antibody drug-conjugates or CAR T cells (Garaude Nature 2024). Similarly, known BCL-2 mutations conferring Venetoclax resistance (Blombery Cancer Discov 2019; Blombery Blood 2020) support the hypothesis that specific BCL-2 variants can be developed to evade Venetoclax binding. Healthy, BCL-2 engineered cells would maintain effective hematopoiesis, allowing the use of higher and prolonged doses of Venetoclax, potentially enhancing therapeutic efficacy while reducing dose-dependent hematotoxicity.We engineered 31 BCL-2 variants with point mutations in 7 residues lining the Venetoclax binding pocket, aiming to reduce Venetoclax binding, while preserving BCL-2's apoptotic regulatory function. Mutations were selected based on structural analysis of BCL-2 bound to Venetoclax (Birkinshaw Nat Commun 2019) or its pro-apoptotic interaction partners, BAX (Ku Cell Res 2011), as well as known Venetoclax resistant variants (such as G101V, D103E and F104L). Residues critical for binding the natural pro-apoptotic ligands (BAK, BAX and BIM) were excluded. All variants were systematically screened for maintained binding to the pro-apoptotic ligands and for loss of Venetoclax binding. Several variants showed over 80-fold reduced Venetoclax binding (IC50) while retaining binding to pro-apoptotic proteins and wild-type thermal stability, indicating structural integrity. Our engineered variants provided greater Venetoclax resistance than known mutations and the best candidate, BCL-2Var4.5, was chosen for further experiments.Although base editing (BE) can install the known Venetoclax-resistance variants, it is not suitable for introducing the preferred BCL-2Var4.5. Therefore, we employed prime editing (PE) due to its high versatility and precision. Using a semi-high throughput plasmid-based platform in HEK293T cells, we designed and screened 92 engineered prime editing guide RNA (epegRNA) encoding BCL-2Var4.5, varying reverse transcriptase template (RTT) and primer binding site (PBS) lengths and using either PEmax or PE7 prime editors to find the most efficient combination. The screen identified 4 epegRNAs that efficiently introduced the preferred variant, each exceeding 10% editing efficiency measured by next generation sequencing (NGS) without generation of indels. These 4 epegRNAs were used to engineer clinically relevant human CD34+ HSPCs mobilized from peripheral blood. RNA-based nucleofection of HSPCs yielded editing efficiencies comparable to the plasmid screening, achieving >20% NGS edited reads. We are currently further improving BCL-2 editing efficacy. In parallel, we are evaluating the ability of the BCL-2Var4.5 edited cells to resist Venetoclax treatment across multiple cell lines and primary HSPCs using in vitro cytotoxicity and cell viability assays. Finally, we are also assessing enrichment of engineered BCL-2Var4.5 cells upon Venetoclax treatment in vitro and in vivo.In summary, we identified and successfully engineered BCL-2 variants that will effectively protect cells from Venetoclax therapy. We anticipate that engineering Venetoclax resistance will overcome treatment-related hematotoxicity yet enable increased therapeutic efficacy. In the future this could offer patients expanded options to treat hematological malignancies.