ABSTRACT Antigen-specific therapies have long been pursued to improve outcomes in acute myeloid leukemia (AML). So far most exploited are monoclonal antibodies (mAbs) targeting CD33, a glycoprotein displayed on the cell surface of leukemic blasts in almost all cases and possibly leukemia stem cells in some. Longer survival of some patients treated with the CD33 antibody-drug conjugate gemtuzumab ozogamicin (GO) validates this approach, but GO is often ineffective, prompting efforts to develop improved, more potent CD33-directed therapeutics. Because AML cells are exquisitely sensitive to radiation in a dose-dependent fashion, radionuclides are ideal to arm anti- CD33 mAbs. Indeed, early phase clinical trials demonstrated substantial anti-AML efficacy of the anti-CD33 mAb lintuzumab (HuM195, SGN-33) when coupled with the a-emitter actinium-225 (225Ac). a-emitters deliver a very high amount of radiation over just a few cell diameters, thereby enabling precise and efficient target cell kill, rendering them particularly interesting for specific targeting of AML with radioimmunoconjugates (“RIT”). However, even with 225Ac-lintuzumab, an important shortcoming is CD33 expression on normal myeloid cells, which leads to “on-target, off-tumor cell” toxicities that manifest as severe and prolonged myelosuppression with life-threatening sequelae (e.g. infection). Thus, clinical use of CD33-directed RIT without immediate stem cell rescue is currently limited to suboptimal drug doses. We have recently demonstrated in mice and nonhuman primates that CRISPR/Cas9 nuclease-based editing of CD33 results in functionally normal hematopoiesis that expresses reduced levels of CD33 and is protected from GO and CD33-directed T cell-engaging therapeutics. We hypothesize CD33-edited normal hematopoietic stem and progenitor cells (HSPCs) will resist CD33-directed RIT with a-particle-emitting radionuclides and enable their safe use at maximally effective drug doses. However, the CRISPR/Cas9-based CD33 gene editing strategy suffers from significant off-target activity, and DNA double strand breaks (DSBs) can generate larger deletions and complex chromosomal rearrangements and cause TP53-dependent DNA damage response and cell cycle arrest. To address this limitation, we will optimize and characterize a novel gene-editing strategy to protect normal hematopoiesis from highly potent CD33-directed RIT by utilizing the recently described base editor (BE) technology. BEs induce precise nucleotide modifications without intentional introduction of DSBs, making them an attractive strategy to generate CD33null “normal” hematopoietic cells. We have assembled a multidisciplinary team of investigators with complementary expertise in CD33-directed therapies, preclinical optimization of RIT, and radiopharmaceutics to conduct well-controlled preclinical IND-enabling studies to develop BE-based CD33 engineering of normal human HSPCs for clinical use with a-emitter CD33-directed RIT for patient...