Abstract NG05: Epitope editing enables targeted immunotherapy of acute myeloid leukemia

Gabriele Casirati,Andrea Cosentino,Adele Mucci,Mohammed S. Mahmoud,Iratxe Ugarte Zabala,Jing Zeng,Scott B. Ficarro,Denise Klatt,Christian Brendel,Alessandro Rambaldi,Jerome Ritz,Jarrod A. Marto,Danilo Pellin,Daniel E. Bauer,Scott A. Armstrong,Pietro Genovese
DOI: https://doi.org/10.1158/1538-7445.am2024-ng05
IF: 11.2
2024-04-07
Cancer Research
Abstract:Introduction. Chimeric antigen receptor T cells (CAR-T), bispecific and antibody-drug conjugates are promising adoptive immunotherapies that can overcome the limitations of conventional cancer treatments and have demonstrated striking efficacy when targeting dispensable lineage antigens (Ag), e.g. CD19 for B-ALL. Nonetheless, the absence of safely actionable tumor-restricted markers hampers their application to other hematological malignancies, such as acute myeloid leukemia (AML). Since AML shares most surface markers with normal hematopoietic stem/progenitor cells (HSPC) or differentiated myeloid cells, on-target/off-tumor toxicities would result in myeloid aplasia and impairment of hematopoietic reconstitution. Furthermore, due to AML intra-tumoral heterogeneity, targeting more than one Ag may be required, exacerbating the risk of overlapping toxicity. Despite this, AML immunotherapies are currently under development, but their role will likely be restricted to bridge treatment before allogeneic HSPC transplantation (HSCT), decreasing the chances of AML eradication. Removal of targeted Ag through CRISPR-Cas KO from donor HSPCs used in HSCT has recently been proposed, but this can only be applied to genes dispensable for hematopoietic function. However, targeting irrelevant genes may facilitate tumor escape through Ag downregulation. Here, we show that precise editing of the targeted epitope within FLT3, KIT (CD117) and IL3RA (CD123) in HSPCs results in loss of Ab binding without KO, preserving physiologic protein expression, regulation, and intracellular signaling. Critically, this strategy enables targeting one or more genes fundamental for leukemia survival, resulting in potent anti-leukemia efficacy with minimal on-target/off-tumor toxicity. Methods. Through epitope-mapping, we identified substitutions in the FLT3, KIT and CD123 extracellular-domains that avoid detection by therapeutic Abs. We validated the functionality of mutated receptors (ligand affinity, western-blot, proliferation, RNAseq, phospho-proteomic MS) and their resistance to on-target killing (mAb-affinity, CAR-T co-culture). We optimized a base-editing protocol to introduce these mutations in CD34+HSPCs and developed advanced in vivo models with co-engraftment of healthy HSPCs, patient-derived AML xenografts (PDX) and CAR-T to assess selective elimination of leukemia and protection of healthy hematopoiesis. Results. To develop our approach, we selected mAbs under development for AML therapy: clone 4G8 (FLT3), Fab-79D (KIT) and 7G3 (CD123). To identify residues involved in mAb binding, we designed Sleeping Beauty epitope mapping libraries. We found that single amino-acid substitutions can disrupt therapeutic mAb binding despite preserved surface receptor. Since epitope engineering can be achieved by point mutations, we reasoned that base editing (BE) could be a suitable and safer option compared to homology directed repair (HDR). By electroporating sgRNA+ABE variants, we achieved successful epitope editing with minimal toxicity. By using fluorescent FLT3L, SCF and IL-3, we confirmed preserved ligand binding to edited receptors. Activation of downstream signaling was confirmed by western blot. By performing in vitro killing assays, we found that, while cells expressing WT FLT3, KIT or CD123 were eliminated, those expressing epitope-edited variants were resistant to CAR-mediated killing and survived up to experiment termination without eliciting T cell activation and degranulation. To introduce our variants into human HSPCs, we optimized a BE protocol on mobilized peripheral blood-derived CD34+ cells, achieving up to 86%, 78% and 78% efficiency for FLT3, KIT and CD123, respectively. Contrary to previous observations with HDR editing, BE efficiencies were similar in bulk and primitive, HSC-enriched subsets (CD90+45RA-) with no skewing of stem phenotype. BE HSPCs were resistant to CAR or mAb-mediated killing in vitro. To confirm the safety of our approach, we compared FLT3, CD123, KIT edited to AAVS1 control HSPCs and found no differences in proliferative response, transcriptional changes (RNAseq), phospho-proteomic profile and colony-forming capacity. Xenotransplantation of BE HSPCs in NBSGW mice showed preserved repopulation and multilineage differentiation capacity, both in primary and secondary recipients. To assess if FLT3 CAR can eliminate AML while sparing FLT3-edited hematopoiesis, we sequentially engrafted NBSGW mice with HSPCs and human PDX cells. We observed a significant increase in the percentage of FLT3 edited cells in CAR-treated mice and relative depletion of CD19+ subsets (pre-B, pro-B), granulocytes, granulo-mono progenitors (GMP) and lymphoid-primed multipotent progenitors (LMPP) only in the AAVS1-BE group, while mice engrafted with FLT3-BE HPSC were protected. Concomitantly, mice treated with 4G8-CAR achieved complete AML eradication. CAR-T exposed to FLT3-BE hematopo -Abstract Truncated-
oncology
What problem does this paper attempt to address?