Ornithine transcarbamylase deficiency (OTCD) is a monogenic disease of ammonia metabolism in hepatocytes. Severe disease is frequently treated by orthotopic liver transplantation. An attractive approach is the correction of a patient's own cells to regenerate the liver with gene-repaired hepatocytes. This study investigates the efficacy and safety of ex vivo correction of primary human hepatocytes. Hepatocytes isolated from an OTCD patient were genetically corrected ex vivo, through the deletion of a mutant intronic splicing site achieving editing efficiencies >60% and the restoration of the urea cycle in vitro. The corrected hepatocytes were transplanted into the liver of FRGN mice and repopulated to high levels (>80%). Animals transplanted and liver repopulated with genetically edited patient hepatocytes displayed normal ammonia, enhanced clearance of an ammonia challenge and OTC enzyme activity, as well as lower urinary orotic acid when compared to mice repopulated with unedited patient hepatocytes. Gene expression was shown to be similar between mice transplanted with unedited or edited patient hepatocytes. Finally, a genome-wide screening by performing CIRCLE-seq and deep sequencing of >70 potential off-targets revealed no unspecific editing. Overall analysis of disease phenotype, gene expression, and possible off-target editing indicated that the gene editing of a severe genetic liver disease was safe and effective.

The urea cycle enzyme carbamoyl phosphate synthetase 1 (CPS1) catalyzes the initial step of the urea cycle; bi-allelic mutations typically present with hyperammonemia, vomiting, ataxia, lethargy progressing into coma, and death due to brain edema if ineffectively treated. The enzyme deficiency is particularly difficult to treat; early recognition is essential to minimize injury to the brain. Even under optimal conditions, therapeutic interventions are of limited scope and efficacy, with most patients developing long-term neurologic sequelae. One significant encumberment to gene therapeutic development is the size of the CPS1 cDNA, which, at 4.5 kb, nears the packaging capacity of adeno-associated virus (AAV). Herein we developed a split AAV (sAAV)-based approach, packaging the large transgene and its regulatory cassette into two separate vectors, thereby delivering therapeutic CPS1 by a dual vector system with testing in a murine model of the disorder. Cps1-deficient mice treated with sAAVs survive long-term with markedly improved ammonia levels, diminished dysregulation of circulating amino acids, and increased hepatic CPS1 expression and activity. In response to acute ammonia challenging, sAAV-treated female mice rapidly incorporated nitrogen into urea. This study demonstrates the first proof-of-principle that sAAV-mediated therapy is a viable, potentially clinically translatable approach to CPS1 deficiency, a devastating urea cycle disorder.