An insulator strategy to prevent insertional oncogenesis in gene therapy

Abstract

Instances of insertional mutagenesis by retroviral vectors and consequent tumorigenesis during human gene therapy have prompted the need for strategies to prevent the phenomenon of oncogene activation by transcriptional enhancers used in integrating gene therapy vectors. We tested a small, well-defined enhancer-blocking insulator element from the chicken beta-globin locus (cHS4-FII) for its ability to prevent genome-wide utilization of oncogenes by a potent, T-lymphomagenic murine leukemia virus. We first showed effective blocking of gammaretroviral enhancer activity by the cHS4-FII insulator in cultured T-cells using a transiently tranfected reporter gene. The insulator activity was CTCF and copy-number dependent and enhancer function was almost eliminated when multiple (>4) insulators were introduced in tandem. To then test the effect of the insulator on gammaretroviral lymphomagenesis, replication-competent murine leukemia viruses containing one or more copies of cHS4-FII adjacent to viral enhancers were generated to infect mice. Lymphomagenesis was significantly reduced by 3 or more insulators but only a modest effect was observed at lower copy numbers. A genomics analysis of proviruses in the ensuing tumors revealed a frequent loss of insulators at higher insulator-copies, presumably favoring tumorigenesis. However several instances where oncogene utilization was not prevented despite presence of multiple insulators in orientations where their activity was expected to be optimum were also observed; such instances may represent breakthrough events. Furthermore, as a strategy to lower the propensity of tandemly repeated insulators to become deleted during viral replication presumably by reverse transcriptase jumping, multiple, known vertebrate insulators of divergent sequences were introduced in tandem and found to significantly improve insulator stability in replicating viruses in culture and in mice. Mice infected with recombinant viruses bearing such heterogenous insulators in multiple copies showed a significantly improved survival compared to control counterparts as demonstrated by a reduced disease incidence and prolonged latency. Thus, our preliminary results strongly suggest that small and multiple, tandem insulators will be necessary to maximize protection against enhancer activation of cellular oncogenes. Furthermore, insulator stability can be improved by reducing sequence identity between insulator repeats. Further development and validation of such strategies may improve the safety of future human gene therapy.