Promoterless gene targeting without nucleases rescues lethality of a Crigler‐Najjar syndrome mouse model

Abstract

The Crigler-Najjar Syndrome Type I (CNSI) is a rare monogenic disease caused by null mutations in the UGT1A1 gene, resulting in bilirubin conjugation deficiency and unconjugated hyperbilirubinemia, with lifelong risk of neurological damage, kernicterus and death. Current clinical practice consists of phototherapy treatment (PT) for more than 10-12 h/day, profoundly affecting (social) life quality. However, PT becomes less effective with age, leaving liver transplantation as the only therapeutic option, which is associated with significant morbidity and mortality (90% survival, 10 years after transplantation). rAAV-mediated liver gene therapy in neonates is associated with the loss of viral genomes over time, and insertional mutagenesis due to integration of the expression cassette into genomic oncogenic loci. Gene targeting/editing strategies using co-expression of a nuclease increase the probability of off-targeting and mutagenesis, as well as an immune-response directed against the nuclease. To overcome these issues, we applied a novel and safer in vivo gene targeting approach recently developed and successfully applied to a hemophilia B model, based on the targeted insertion of a promoterless version of the Factor IX cDNA into the albumin locus, without the use of nucleases. To implement this approach for Crigler-Najjar, we used a lethal mouse model of the disease previously developed in our laboratory. The strategy is based in the in frame knock-in of the human UGT1a1 cDNA into the albumin locus using a promoterless rAAV8 vector. The hUGT1a1 cDNA is inserted downstream of the albumin ORF, with a ribosomal skipping 2A sequence. This results in a single chimeric mRNA driven from the endogenous albumin locus that produces two proteins: albumin and UGT1a1. We IP injected mutant pups, temporarily treated with PT up to P15 (a procedure that rescues all mutant mice), at different ages with a rAAV8 donor vector to determine the most effective timepoint. We observed a therapeutic reduction in plasma bilirubin of AAV8-treated mice (n=4-6 per timepoint, p£0.001, ANOVA). RT-PCR analysis of total liver RNA showed the presence of the hybrid Albumin-Ugt1a1 mRNA, which was confirmed by DNA cloning and sequencing, proving that the Ugt1a1 transcript was derived from the endogenous albumin locus. More importantly, when we injected P4 mice and treated them with PT up to P8 (a procedure that alone results in 95%-100% mortality), we rescued lethality in all rAAV-treated mutant mice, while all untreated animals died (n=7, p£0.001, Log-rank test). Future steps will involve the improvement of the therapeutic efficiency by transducing a higher number of viral genomes, and mRNA translatability by using codon optimized versions of the hUGT1a1 cDNA. The results presented here support the conclusion that the promoterless gene targeting therapy without nucleases, already effective for hemophilia B, proved to be efficient for CNSI, strongly suggesting that this approach is applicable for providing therapeutic levels of an intracellular enzyme responsible for a lethal liver metabolic disease.