A Novel Gene Therapy Strategy Provides Hope for Hemophilia

Recent technological advances in genome engineering have facilitated targeted stable transgene expression for gene therapy applications. In particular, site-specific endonucleases are efficient agents for targeted gene editing, whereby the endonuclease gene is typically delivered into cells by viral vectors and expressed under control of a vector-borne promoter. Recombinant adeno-associated virus (rAAV) vectors are commonly used in human clinical trials since they are derived from a nonpathogenic virus, are not immunogenic, and allow high gene-targeting rates. The rAAV can exist as several different serotypes, which differ in the composition of their capsid protein coat and are thus able to transduce various cell types.However, despite promising results, adverse effects such as off-target endonuclease cleavage events, as well as activation of neighboring genes (including oncogenes) by the integrated vector-borne promoter, remain a major concern.Dr. Adi Barzel and colleagues at Stanford in California have circumvented these issues with a novel approach using a promoterless, endonuclease-independent rAAV for gene targeting. As proof of concept, they used hemophilia B mice and targeted the human coagulation factor IX (F9) to the albumin gene locus, which is highly expressed in the liver, and demonstrated amelioration of the bleeding diathesis. They used the liver-tropic rAAV8 serotype and an innovative experimental strategy. The vector was engineered by inserting 5' and 3' homology arms spanning the stop codon of the mouse albumin (Alb) gene. A codon-optimized human F9 cDNA without the gene promoter, preceded by a 2A-peptide coding sequence, was then inserted in between the arms. Homologous recombination resulted in integration of the 2A-F9 coding sequence immediately 5' of the Alb stop codon. This created a fused gene under control of the strong albumin promoter, which compensated for the lack of the F9 promoter. Transcription resulted in a fused mRNA coding for a polyprotein composed of albumin, 2A peptide and F9. Since 2A peptides are self-processing peptides derived from RNA viruses, inclusion of the 2A peptide is a critical component of these researchers’ ingenious strategy. During translation, the 2A peptide impairs normal peptide bond formation using a ribosomal skipping mechanism, which allows multiple proteins to be produced from a single vector. This process adds approximately 20 amino acids to the C terminus of albumin, which is the first translated protein, but this does not diminish its function. The F9 protein gains a single proline at the amino terminus, which does not adversely affect its processing in the endoplasmic reticulum and subsequent secretion. This strategy therefore does not disrupt the robust expression of hepatic albumin and results in stoichiometric expression of the F9 protein.The engineered rAAV8-F9 vector was injected into neonatal and adult hemophilia B mice and plasma F9 levels between 7 and 20 percent of normal were achieved, which remained stable for up to 20 weeks. The treated mice also had normal coagulation times. The researchers demonstrated that F9 protein was only produced from on-target integration, since a vector-borne promoter was absent. Integration only occurred in about 0.5 percent of the hepatic albumin alleles; however, since the Alb locus has high transcriptional activity and 2A ribosome skipping was highly efficient, therapeutic levels of F9 clotting factor were achieved. Using an inverse control vector, they found no evidence of off-target events, and no liver toxicity was observed. Apart from this favorable safety profile, another advantage of their strategy is that due to targeted vector integration, there should not be a loss of expression of the transgene over time, even in dividing tissues.