Induction of Endogenous Repair Mechanisms by Single-Stranded DNA Oligonucleotide Therapy for Correction of the DNA-PK Mutation in Murine Severe Combined Immune Deficiency.
Abstract Gene correction is an attractive strategy for gene therapy since it allows the corrected gene to remain regulated within its native genome location. We have explored gene correction of murine severe combined immunodeficiency (SCID) with single-stranded DNA oligonucleotides (SSO). Murine SCID is characterized by severe T- and B-cell lymphopenia and is caused by a point mutation in the DNA protein kinase subunit (DNA-PK). To correct the mutant missense sequence (T to A substitution), a silent mutation was introduced by synthesizing the SSO non-transcribed sequence (45 bp) surrounding the site of the SCID mutation and replacing the T nucleotide with a C nucleotide to permit production of wild-type (wt) DNA-PK protein. Since the fetus is potentially an ideal permissive environment for gene correction due to the high proliferative rate of its tissues, SSO were injected in utero either directly into the liver of the fetus or transplacentally (via hydrodynamic infusion to the pregnant dam). E15/16 BALB/c-SCID recipients (N = 78) were injected with SSO (20 mcg/fetus). Twenty nine mice survived to term and, when evaluated by peripheral blood (PB) FACS at 15–30 weeks of life, 11 had significant phenotypic evidence of immune restoration defined as ≥ 2% CD4+ or CD8+ T cells: 6 had both CD4+ and CD8+ T cells, 2 had CD4+ cells only and 3 had CD8+ T cells only. The highest level of CD4+ cells seen was 9%, the highest level of CD8+ cells was 2% and both had TCR rearrangement and 27% and 15% genotypic correction of the mutated bp by quantitative pyrosequencing (PSQ) of DNA isolated from whole blood. Since placental membranes are permeable to some molecules, SSO were hydrodynamically delivered to pregnant BALB/c-SCID dams (100 mcg). Two of 8 evaluable mice injected on day E5/6 had significant numbers of T cells, one of which had 20% CD8+ with 3% CD4+ cells at 13 weeks of life, and PSQ showed a 13% correction rate. Sixteen offspring injected at E13/14 were analyzed: 2 had 4% and 5% CD4+ cells and the latter also had 6% of CD8+ cells with PSQ correction rates of 22% and 11%, respectively. Of 40 mice evaluated after transplacental injections at age E15/16, 9 had >2% CD4+ or CD8+cells. The four with the highest T cell count had a genotypic correction of 12–25% of wt levels. Notably, littermates with no phenotypic correction had no evidence of gene correction at the DNA-PK mutation site. However, in all immune-restored animals that were analyzed for gene correction, (2/78 after in utero; 7/64 after transplacental delivery) an A to T rather than the anticipated A to C correction occurred. This is consistent with the hypothesis that SSO stimulated homologous recombination with a preferred utilization of the endogenous T rather than the exogenous C due to preferential pairing of two pyrimidines (A with T) than pyrimidine with purine (A with C). In summary, we show that SSO therapy for correction of DNA-PK mutation is possible when SSO are injected in utero at late gestation or are hydrodynamically delivered to the pregnant dam. These findings also suggest that while DNA homology around the mutation site is necessary for correction, the wt nucleotide is favored by the endogenous DNA repair pathway.