Successful cancer treatments eliminate cancer cells without causing great damage to healthy tissues. For blood cancers like acute myeloid leukemia (AML), achieving this ideal has been difficult because cancerous cells express many of the same surface markers as normal cells. Consequently, current AML immunotherapies target tumors, but they also cause dangerous losses of healthy blood cell progenitors. 

Scientists at the Dana Farber Cancer Institute led by gene therapy researcher Pietro Genovese developed an epitope editing-based strategy to create hematopoietic stem cells that resist common immunotherapies, including chimeric antigen receptor (CAR) T cells and monoclonal antibodies (1). Clinicians can use this technique to populate the diseased bone marrow with healthy hematopoietic stem cells that immunotherapies will not harm, which would minimize treatment side effects. It might also allow patients to safely receive repeated treatments when needed and to consider drug combinations that might otherwise be risky. 

Previously, scientists explored CRISPR-Cas knockout or exon skipping to remove targeted proteins from healthy cells to shield them from cancer treatment. However, uncertainties remained about the long-term influence of such modifications on myeloid cell functionality in humans. “[Genovese’s team’s] strategy is creative because they are only modifying the epitope without affecting the function of the receptor or signaling transduction,” said Ghayas Issa, a medical oncologist at the MD Anderson Cancer Center who was not involved with the study.

It came as a surprise that single amino acid changes were sufficient to completely abrogate antibody binding. 
– Gabriele Casirati, Dana Farber Cancer Institute

Gabriele Casirati, hematologist and study coauthor, aimed to identify the minimal number of residues that prevented the therapeutic monoclonal antibodies 4G8, Fab-79D, and 7G3 from binding their respective proteins, FLT3, CD123, and KIT, all of which are common in both healthy and AML cells. Through a combinatorial library, the researchers identified single amino acid substitutions that resulted in the loss of monoclonal antibody binding while maintaining protein expression. “It came as a surprise that single amino acid changes were sufficient to completely abrogate antibody binding,” Casirati said. 

They then used adenine base editing to make the single amino acid change in the target protein epitope, resulting in the loss of monoclonal antibody recognition without significant protein knockout and without compromising the proteins' functionality, ligand binding, or downstream signaling pathways. “This [approach] is likely safer than [CRISPR-Cas9] editing strategies because base editing doesn't require making double strand breaks [that can cause mutations] in the DNA,” said Johanna Olweus, a clinical immunologist at the University of Oslo who did not participate in the study. 

In in vitro killing assays, CAR T cells killed a leukemia cell line expressing unmodified FLT3, KIT, or CD123, while cells expressing epitope-engineered variants were resistant to CAR T cells. For in vivo experiments, Casirati’s team introduced the genetic changes into healthy hematopoietic stem cells before transplanting them into mice. These modified cells persisted and resisted immunotherapy attack, which otherwise eliminated cells that had not undergone this specific modification.

Lastly, the authors explored the potential of multiplex epitope editing by targeting multiple antigens simultaneously. They conducted experiments using dual- and triple-edited leukemia cell lines, incorporating edits for FLT3, CD123, and KIT, and cocultured these cells with bi- or trispecific CAR T cells. They found that only the cells edited for multiple epitopes survived the specific CAR-mediated killing while still expressing the targeted antigens. 

“It is a very elegant approach that looks very promising,” said Olweus, and other research groups have started to explore similar strategies (2). However, challenges remain. For instance, CD123's high expression on long-lived dendritic cells poses risks. “It often takes a long time before dendritic cells are replaced by the transplanted cells, which could cause toxicity,” she said. 

Shielding healthy hematopoietic stem cells from immunotherapies might also not be enough to increase the safety of human AML immunotherapies, and potential off target effects elsewhere in the body are possible. “CD123 and KIT are also expressed in the brain,” Olweus added. “Whether CAR T cells [reactive to these targets] could cause any damage there is difficult to investigate in mouse models.” 

Issa agreed. "It's very different to do a screen in vitro versus a product that would eventually be infused in patients. But this is great progress.”

References

1. Casirati, G. et al. Epitope editing enables targeted immunotherapy of acute myeloid leukaemia. Nature  621, 404–414 (2023).
2. Wellhausen, N. et al. Epitope base editing CD45 in hematopoietic cells enables universal blood cancer immune therapy. Sci Transl Med  15, 714 (2023).