IntroductionA significant fraction of inherited monogenic disorders is caused by patient-specific mutations dispersed over the entire locus of the affected gene. Correcting these mutations by introducing healthy gene copies into the genome of the diseased cells proved successful in several clinical gene therapy trials. Most of these trials employed retroviral vectors, which by inserting randomly throughout the genome deprived the transduced genes of their endogenous control and caused insertional mutations leading to secondary disease. The development of genome editing tools capable of modifying any prespecified genomic sequence with unprecedented accuracy opened up a wide range of new possibilities in gene manipulation including targeted gene repair. In particular, CRISPR/Cas9, a prokaryotic adaptive immune system, and its swift repurposing for genome editing was widely adopted as the hitherto simplest genome editing tool. CRISPR/Cas9 is an RNA guided endonuclease that uses RNA-DNA base pairing to target genomic DNA. Bound to its target via the guide (g)RNA, Cas9 induces DNA double strand breaks (DBS) at prespecified genomic sites that promptly activate the endogenous DNA repair machinery. DSB repair is accomplished by either non-homologous end joining (NHEJ) or homology directed repair (HDR). Thus far, correction of human mutations in hematopoietic cells relied entirely on HDR requiring gene specific donor templates in addition to the RNA guided endonucleases (RGNs). However, although the HDR offers precision, its efficiency is low and requires positive selection to enrich for gene corrected cell. Because in mammalian cells DSB repair by NHEJ significantly exceeds HDR and even more importantly, is the dominant DSB repair pathway in hematopoietic stem- and progenitor cells (HSPC), we thought to exploit NHEJ for gene therapy because in theory, approximately one third of the indels associated with NHEJ should restore the open reading frame (ORF) disrupted by a disease mutation. This would lead to a significant number of ORF reconstitutions of which some, depending on the position and type of the original mutation, should either completely or partially recover protein function.ResultsTo test gene repair efficiency by NHEJ in human hematopoietic cells, we generated PLB-985 (PLB) reporter cells expressing mutationally inactivated EGFP (mEGFP). Transduction of mEGFP expressing PLBs (mEGFP-PLB) with integrase-deficient lentiviral (IDLV) particles encoding for RGNs targeting the EGFP mutation reconstituted EGFP expression in up to 27% of the mEGFP PLB cells. Indel analysis revealed that 13 out of 28 (46%) restored the EGFP-open reading frame of which 7 (25%) reconstituted EGFP activity. To test whether the donor-template free IDLV strategy would also effectively correct bone fide disease mutations, we performed similar experiments with X-CGD PLB cells expressing transgenes encoding patient-specific frameshift, missense and nonsense CYBB mutations causing X-linked chronic granulomatous disease (X-CGD) which is an inherited, life threatening immunodeficiency disorder. Transduction of the cells with IDLVs carrying RGNs directed against each of these mutations restored CYBB function in up to 10% of cells harboring frameshift mutations which is sufficient to protect X-CGD patients from microbial infections. However, RGNs directed against the nonsense or missense mutations restored CYBB function in only 1-2% of the cells, suggesting that these mutations are less amenable to CRISPR/Cas9 mediated repair. As Cas9 frequently tolerates single nucleotide mismatches, selection against solitary nucleotide substitutions may explain this low efficiency of gene repair.ConclusionsFrameshift mutations can be effectively repaired by NHEJ in hematopoietic cells by CRISPR/Cas9 transducing IDLVs. As about 25% of most inherited blood disorders are caused by frameshift mutations, our results suggest that about a quarter of patients suffering from monogenic blood disorders could benefit from personalized, template free CRISPR/Cas9 gene therapy. DisclosuresNo relevant conflicts of interest to declare.