Development of a UM171-Supported, CD33 Epitope-Engineered Blood Stem Cell Graft for Advanced AML Therapies

Targeted immunotherapies for myeloid malignancies suffer from the lack of selectively expressed epitopes on the surface of leukemic versus normal hematopoietic cells, especially hematopoietic stem and progenitor cells (HSPC). Building on our clinically proven UM171 cord blood expansion cell therapy (Cohen et al., Blood Advances, 2023; PMID 37467030) for high risk myeloid malignancies and our Leucegene program for surface antigen identification (Bordeleau et al., Cell Reports, 2024, PMID 38838225), we are now developing a strategy that exploits advances in genetic engineering and stem cell expansion/rejuvenation to shield healthy transplanted cells from immuno-therapeutics and their associated on-target toxicities while preserving the maximal anti-leukemia benefit of CB transplants.

To this end, we first compared Cas9 or Cytosine base editing (CBE) mediated knockout of the therapeutic antibody target CD33 (Gemtuzumab Ozogamicin, GO, Mylotarg, Pfizer), a proof-of-principle surface protein commonly expressed on both malignant and normal myeloid cells, as well as on CD34+ stem and progenitor cells.

Using previously optimized electroporation conditions based on a human HSC-reporter (Lehnertz et al., Blood, 2021, PMID 34499717), transient delivery of either Cas9 or CBE (TadCBEd) encoding, fully N1-methyl-pseudo-Uridine substituted mRNA along with a synthetic sgRNA targeting CD33 exon 2 (Synthego) achieved highly efficient and reproducible CD33 gene ablation (typically >90%) in cord blood derived CD34+ HSPCs through introduction of frame-shift or nonsense mutations, respectively. As suggested before (Schiroli et al., Cell Stem Cell, 2019, PMID 30905619; Lehnertz et al., Blood, 2021, PMID 34499717), Cas9-mediated CD33 editing significantly impacted CD34+ cell expansion ex vivo culture presumably due to introduced DNA double strand breaks. In contrast and in the presence of UM171, the CBE-mediated CD33 ablation allowed similar expansion kinetics as observed in unedited control cells.

Accordingly, no significant differences in engraftment between CBE-edited UM171-treated HSPC and unedited controls were detectable in vivo at all timepoints (weeks 5, 10, 15 and 20 post-transplant), while Cas9 treated HSPCs yielded a 11.5-fold reduction in engraftment at week 20 post-transplant (p=3.13e-3)

As a next step, we identified and optimized an Adenine base editing (ABE) approach to selectively introduce key missense mutations in the Gemtuzumab Ozogamicin target epitope in exon 2 of CD33. Importantly, introduction of these missense mutations completely abolishes recognition of the therapeutic CD33 antibody clone but maintains full expression and function of CD33 in edited cells. We demonstrate that this novel approach is highly efficient in an HSC engineering context (75-95% de-epitoped CD33+ cells in transplanted mice), renders edited cells completely insensitive to anti-CD33 therapeutic antibodies and CAR-T cells and synergizes with the benefits of UM171-mediated cord blood CD34+ cell expansion.

In summary, these results demonstrate a broader feasibility of base editing mediated epitope engineering supported by UM171-mediated HSPC ex vivo expansion for next-generation AML therapies and gene correction therapies. The maintenance of all previously documented UM171 benefits with regards to HSPC expansion and rejuvenation in a gene engineering context is of high clinical relevance, particularly in cases where compatible graft sizes are limited. Other epitopes targeted to specific AML subgroups are being developed in this next generation UM171 amplified transplants.