Selective haematological cancer eradication with preserved haematopoiesis

Haematopoietic stem cell (HSC) transplantation (HSCT) is the only curative treatment for a broad range of haematological malignancies, but the standard of care relies on untargeted chemotherapies and limited possibilities to treat malignant cells after HSCT without affecting the transplanted healthy cells1. Antigen-specific cell-depleting therapies hold the promise of much more targeted elimination of diseased cells, as witnessed in the past decade by the revolution of clinical practice for B cell malignancies2. However, target selection is complex and limited to antigens expressed on subsets of haematopoietic cells, resulting in a fragmented therapy landscape with high development costs2–5. Here we demonstrate that an antibody–drug conjugate (ADC) targeting the pan-haematopoietic marker CD45 enables the antigen-specific depletion of the entire haematopoietic system, including HSCs. Pairing this ADC with the transplantation of human HSCs engineered to be shielded from the CD45-targeting ADC enables the selective eradication of leukaemic cells with preserved haematopoiesis. The combination of CD45-targeting ADCs and engineered HSCs creates an almost universal strategy to replace a diseased haematopoietic system, irrespective of disease aetiology or originating cell type. We propose that this approach could have broad implications beyond haematological malignancies.


Identification of base editable residues in the CD45 extracellular domain to provide shielding from antibody binding
We paired the sgRNAs with BEs covering different architectures, e.g.cytosine base editors (CBE) and adenine base editors (ABE) with different PAM restrictions available at the time of the screening (SPACE-NG 1 , ABE8e-NG 2 , ABEmax-SpRY, ABEmax-SpG, CBE4max-NG, CBE4max-SpG and xCas9(3.7)-BE4 3 ) (Extended Data Table 1).Combinations of plasmids encoding BEs and sgRNAs were co-electroporated in K562 cells and BE frequencies were quantified from Sanger sequencing using EditR 4 after 3 days (Extended Data Fig. 1e).We established a base editing score (described in the Materials and Methods section) that reflected a cumulative editing activity at all edited positions combined to compare editing performance (Extended Data Fig. 1f).ABE8e-NG displayed the highest editing activity followed by ABEmax-SpG and ABEmax-SpRY.ABEs tended to edit more actively and more often than CBEs.Some sgRNAs did not lead to any editing independent of the BE they were paired with (sgRNAs 19, 22, 28, 29, 30, 31 and 48).Conversely, some sgRNAs induced editing with most BEs (sgRNAs 1, 5, 6, 8, 10, 11 and 38).Next, we analyzed the codon changes resulting from the editing and observed a broad range of editing outcomes: single digit editing at unique bases, silent edits or very high editing (>80%) at one or more bases with the potential to change multiple amino acids.To validate initial screening hits, we sorted successfully electroporated GFP + K562 cells (Extended Data Fig. 1g).For several conditions bystander edits undetected from the initial screen were found.In many cases the most active BE (ABE8e-NG) also yielded the broadest editing window (Extended Data Fig. 1h).To select candidates for further assessment, we considered the editing efficiencies, editing purity, as well as theoretical physicochemical and structural properties of the inferred amino acid substitutions.
Although this variant is highly attractive from an editing point of view, it is located slightly outside the identified MIRG451 epitope and the conservative Ile-to-Val mutation very likely would not fully shield CD45 from antibody binding.Despite high editing efficiencies, some substitutions were excluded from further analysis due to their localization, low surface accessibility and/or their structural role (CD45 Y365H , CD45 I371M or CD45 Y365H+C367R , CD45 C367Y+D368H ) (Extended Data Fig. 1a,b).Finally, sgRNAs 7, 44 and 49 coupled with different base editors generated amino acid substitution profiles altering CD45 E259 (sgRNA-7, D1), CD45 N286 (sgRNA-44, D1) and CD45 K352 (sgRNA-49, D2), found to be key residues of BC8 and MIRG451 epitopes, respectively (Fig. 1c-e).Taking all these parameters into account we selected candidates for subsequent experiments.
Human T cells base edited with sgRNA-7 (CD45 E259G ) or sgRNA-44 (CD45 I283M+H285R+N286D ) were incubated with increasing concentrations of BC8-Saporin (BC8-SAP).3 days later, the cell composition was analyzed by flow cytometry and live cells were sorted and sequenced.1nM free SAP did not affect the ratio of edited to non-edited cells (Extended Data Fig. 3h,i,k,l).In contrast, BC8 + cells were preferentially killed by increasing BC8-SAP concentrations resulting in gradual disappearance of BC8-stained cells (Extended Data Fig. 3h,i,k,l) and an enrichment of the base editing frequency (Extended Data Fig. 3j,m).CD45 E259G base edited cells were shielded from ADC killing up to 0.25nM BC8-SAP.Higher BC8-SAP concentrations resulted in partial killing of the base edited cells, likely due to residual BC8 binding to CD45 E259G (Extended Data Fig. 3h,i; Fig. 1c,d).Similarly, BC8 stained cells were dose-dependently depleted while CD45 I283M+H285R+N286D base edited cells were shielded from all tested BC8-SAP concentrations (Extended Data Fig. 3k,l).Due to the complete nonbinding of BC8 to CD45 H285R and CD45 N286D (Fig. 1c,d), the selective resistance of base edited cells was more apparent and equally resulted in an enrichment of correctly edited cells (Extended Data Fig. 3m).Thus, we identified base editable CD45 variants that shielded primary human T cells from cytotoxicity by a surrogate ADC.CD45 K352E displayed the most desirable characteristics for a genetically encoded shield, i.e. complete loss of binding to a very high affinity mAb (MIRG451) with preserved protein stability (Fig. 1c-e, Extended Data Fig. 2b,c).However, the editing efficiency of ABE8e-NG + sgRNA-49 (CD45 N351D+K352E ) was very low.

Supplementary Methods 2 :
Flow cytometry gating strategy of bone marrow (a), spleen (b) and blood (c) from mice which received sg49.3HSPCs and were treated with saline.Related to Figure3and Extended Data Figure6.

Supplementary Methods 3 :Supplementary Methods 4 :
Flow cytometry gating strategy of bone marrow (a), spleen (b) and blood (c) from mice which received sg49.3HSPCs, MOLM-14 and were treated with saline.Related to Figure4, Extended Data Figure7 and 8.Flow cytometry gating strategy of bone marrow (a), spleen (b) and blood (c) from mice which received sg49.3HSPCs, PDX and were treated with saline.Related to Figure5and Extended Data Figure10.