Abstract Details

C1 - Base Editing and Prime Editing

98: The Impact of Base Editing on Rhesus Macaque Hematopoietic Stem Cell Engraftment and In Vivo Clonal Behavior

Type: Oral Abstract Session

Presentation Details

Session Title: Base Editing and Prime Editing I






CRISPR/Cas9 editing has transformed hematopoietic stem and progenitor cell (HSPC) gene therapies; however, Cas9 generates double-stranded breaks (DSBs), enhancing off-target risks, p53 activation, and HSPC genotoxicity. Base editors (BEs) mitigate these concerns by facilitating single-nucleotide conversions without DSBs. Investigation into BE impact on long-term HSPC engraftment and clonality is necessary to evaluate BEs for clinical HSPC gene therapies.
Depleting CD33, a myeloid differentiation marker, protects HSPCs from CD33-targeted immunotherapies without effecting functionality. We evaluated the clonal impact of HSPC CD33-targeted adenine BE (ABE, ABE8e variant) in a competitive rhesus macaque (RM) autologous HSPC transplantation (HSCT) model (n=3). Following total body irradiation, RMs were infused with two genetically barcoded CD34⁺ HSPC populations: Lib#1 was unelectroporated (Barcode-only arm) and Lib#2 was electroporated for CD33-targeted ABE (Barcode+ABE arm) (Fig. 1A). ABE efficiency in the Barcode+ABE arm varied from 11.7-88.2% in the 3 RMs (Fig. 1B). In vitro, no significant differences were observed in cell growth, viability, CFU plating efficiency, or lineage distribution between both arms.
Following HSCT, blood cells were collected weekly to track engraftment via GFP expression and ABE alleles and to follow clonal diversity via quantitative barcode retrieval. In vivo, GFP marking peaked earlier than the editing level in granulocytes (Gr), suggesting slower engraftment of Barcode+ABE cells (Fig. 1C). CD33 ABE levels peaked at expected levels 2-3 weeks post-HSCT, but markedly decreased over time, indicating deficient long-term ABE cell engraftment. Unexpectedly, in the 2 RMs receiving ABE cells with the highest editing frequencies (13U036 & 13U052), the overall GFP⁺ population, consisting of Barcode+ABE and Barcode-only cells, declined in tandem with the decrease of edited cells. In contrast, GFP⁺ cells in the less efficiently edited RM (14U026) maintained consistently high GFP⁺% for up to 14 months (m) despite a significant decrease in edited cells. These findings suggest high ABE efficiency may induce toxicities that impact infused HSPC engraftment via niche effects or paracrine signals, requiring further investigation.
In RM 13U036, 1m CD33^+/- Gr and 3 and 5m GFP⁺CD33^+/- Gr were sorted for barcode clonal analysis. The majority of CD33^- Gr showed expected CD33 ABE (Fig. 1D), and the high-contributing clones revealed distinct clonal patterns between CD33⁺ and CD33^- Gr (Fig. 1E). Initially, short-term engraftment was polyclonal with high diversity in both CD33⁺ lib#1 (Barcode-only) and CD33^- lib#2 (Barcode+ABE) Gr. However, clonal diversity decreased over time, with CD33^- lib#2 Gr's diversity declining the most, exhibiting oligoclonality (Fig. 1E-F).
Despite Barcode+ABE cells maintaining normal characteristics in vitro, our preliminary findings indicate reduced long-term engraftment potential of Barcode+ABE cells in vivo. These findings suggest high ABE levels may enhance HSPC toxicity, increasing risk for HSPC loss and oligoclonal hematopoiesis. While the combination of lentiviral transduction and ABE8e may have added to HSPC stress, these findings provide relevant information for the optimization and safe clinical development of ABE-based gene and cell therapies.

Ashley Gin¹, Chuanfeng Wu¹, Diana Abraham¹, Yifan Zhou¹, Taha B. Hayal¹, Richard Lozano¹, Gregory A. Newby², David R. Liu², Cynthia Dunbar<sup>1

1</sup>National Heart, Lung, and Blood Institute, Bethesda, MD,²Broad Institute of MIT and Harvard, Cambridge, MA"