Non–Integrating Lentiviral Vectors for Ocular Gene Transfer

Abstract

Purpose: : Vector integration carries with it the finite chance of inducing insertional mutagenesis, as has been reproducibly demonstrated in animal models and patients using integrating retroviral vectors. Attempts to demonstrate gene expression in vivo using non–integrating vectors have previously been unsuccessful. We developed non–integrating HIV–1 vectors and successfully demonstrate efficient and sustained in vivo transduction of mouse and rat ocular tissues, and show rescue of two models of RP.

Methods: : Paired integration–proficient or –deficient vectors encoding GFP were delivered subretinally or intracamerally in mice or rats. In vivo fundus imaging and cryohistology were used to evaluate transduction patterns. An ELISA was used to compare in vivo GFP levels between the various vector forms. Paired vectors encoding RPE65 or Mertk were delivered to RPE65^rd12/rd12 mice and RCS rats, respectively, and ERG analysis performed 6–8 weeks later. qPCR was used to examine for long–term persistence of episomal circles and LAM–PCR assays used to compare integration events between eyes injected with the 2 vector forms.

Results: : Integration–deficient vectors mediated efficient and sustained transduction of mouse RPE and corneal endothelial cells, and rat RPE cells. We found no dependence on species, promoter or presence of post transcriptional regulatory elements to achieve efficient transgene expression. There were no statistically significant differences in GFP levels mediated by the two vector types. Therapeutic integration–deficient vectors substantially rescued both models of RP and with comparable efficacy to integrating vectors. Average 2dLTR/total vector DNA ratios in mouse and rat eyes were 8–fold higher with integration–deficient vectors and this was stable over time. A single integration event was detected in eyes injected with integration–deficient vectors, in contrast to those injected with integration–proficient integration vectors in which integration events were readily detected.

Conclusions: : Integration–deficient HIV vectors result in efficient and sustained in vivo transduction and with comparable efficacy to their integrating counterparts. This points to their potential application in gene therapy of post–mitotic or essentially post–mitotic human tissues, where an effective vector with a highly reduced rate of vector integration has clear advantages for biosafety.

Keywords: gene transfer/gene therapy • retina • cornea: basic science