Efficient prime editing of human hematopoietic stem cells corrects the recurrent SBDS C.258+2T>C mutation and recovers engraftment potential in shwachman-diamond syndrome

Abstract Shwachman-Diamond syndrome (SDS) is an inherited bone marrow failure disorder associated with myelodysplastic syndrome and acute myeloid leukemia predisposition. Nearly all patients carry a hypomorphic SBDS c.258+2T>C mutation, which produces aberrant mRNA splicing, reduces gene expression, and impairs ribosome biogenesis. Prime editing (PE) is a CRISPR-based genome editing method that enables versatile installation of short sequence changes templated by a prime editor guide RNA (pegRNA). PE has minimal off-target potential and designs that evade mismatch repair (MMR) bypass potentially genotoxic double strand break intermediates. Here, we investigate PE to correct the recurrent SBDS mutation in hematopoietic stem cells (HSCs). Using K562 cells engineered to carry SBDS c.258+2T>C, we observed that correcting the mutation alone (using a +3C>T pegRNA) was relatively inefficient, with 5.3% prime edits. We evaluated pegRNAs that introduced additional intronic substitutions predicted to evade MMR and preserve splicing and identified a pegRNA that when employed as PE2-type editing (without a second gRNA nicking the opposite strand) achieved 45.2% precise edits, resulting in efficient rescue of full-length SBDS transcript expression. Compared to PE3-type editing, PE2-type showed a similar editing efficiency but a markedly improved precise edit-to-indel ratio (120:1 vs. 18:1). Mouse embryonic fibroblasts (MEFs) engineered to carry a humanized intron 2 segment containing the c.258+2T>C mutation in the cognate mouse Sbds locus (c.199+2T>C) recapitulate the splicing defect (Peters et al. ASH 2022). PE resulted in efficient editing (68%), rescue of SBDS mRNA and protein expression, and normalization of the polysome profile including the 60S:80S ratio and translation rate as assessed by OP-Puro incorporation in Sbds c.199+2T>C homozygous MEFs and in SDS patient-derived fibroblasts. In 3 healthy donor CD34⁺ HSPCs, PE achieved 29±4% editing efficiency at the SBDSP1 pseudogene, which harbors the c.258+2T>C SBDS sequence, with undetectable editing of the nonmutant SBDS gene, consistent with the exquisite precision of PE. Peripheral blood CD34+ HSPCs were mobilized from a patient with SDS due to compound heterozygous SBDS c.258+2T>C and c.183_184del TAinsCT (p.K62*) mutations. After mobilization with G-CSF and plerixafor, 28 CD34+ HSPCs per mcL were mobilized and 3 blood volumes were processed by apheresis yielding 2.3x106 CD34+ HSPCs per kg. The patient harbored 2 somatic clones carrying TP53 mutations: c.524G>A (3.4% VAF) and c.535C>T (3.8% VAF). Ex vivo PE with PEmax-La mRNA and synthetic pegRNA electroporation produced 30% prime edits of SBDS. Using the same number of starting HSPCs allocated to each condition, untreated control and PE treated SDS patient cells were infused to NBSGW mice. Peripheral blood analysis showed increased human chimerism at 9 and 17 weeks in PE treated compared to untreated controls. In bone marrow, human chimerism was 14% in unedited recipients compared to 52% in PE-treated recipients. The edit frequency in PE-treated recipients increased from 30% in the infused HSPCs to 78% in the engrafted BM cells. Single cell RNA sequencing analysis showed that PE treatment rescues multilineage hematopoiesis and reverses ribosomal biogenesis and TP53 intrinsic apoptotic signaling gene expression signatures. Secondary transplantation showed a 12.4-fold increase in long-term repopulation capacity after PE treatment. SBDS edit frequency was 90% in engrafted secondary recipients. We further evaluated the impact of PE on the somatic TP53 mutant clones. In the controls, TP53 mutation frequency remained stable at 14% after engraftment compared to 15% in input HSPCs. In contrast, after PE treatment, the TP53 mutation frequency decreased from 19% in input HSPCs to 5% in engrafting cells. Colony-forming unit (CFU) analysis from engrafting HSPCs showed that 3 of 15 colonies lacking SBDS editing had a TP53 mutation (20%), while only 1 of 87 colonies with SBDS mutation correction showed a TP53 mutation (1.1%), suggesting that SBDS mutation correction counter-selects for engraftment of TP53 mutant HSPCs. Together, these results demonstrate a near-universal and efficient PE approach with high product purity that corrects the recurrent SBDS c.258+2T>C mutation and restores hematopoietic engraftment function in SDS patient HSPCs.