Purpose: With the increasing number of retinal gene-based therapies and therapeutic constructs, in vitro bioassays characterizing vector transduction efficiency and quality are becoming increasingly important. Currently, in vitro assays quantifying vector transduction efficiency are performed predominantly for non-ocular tissues. A human retinal pigment epithelial cell line (ARPE19) and a mouse cone photoreceptor cell line, 661W, have been well characterized and are used for many retinal metabolism and biologic pathway studies. The purpose of this study is to quantify transduction efficiencies of a variety of self-complementary (sc) adeno-associated virus (AAV) vectors in these biologically relevant ocular cell lines using high-throughput fluorescence-activated cell sorting (FACS) analysis.
Methods: ARPE19 and 661W cells were infected with sc-smCBA-mCherry packaged in unmodified AAV capsids or capsids containing single/multiple tyrosine-phenylalanine (Y-F) mutations at multiplicity of infections (MOIs) ranging from 100 to 10,000. Three days post infection fluorescent images verified mCherry expression. Following microscopy, FACS analysis was performed to quantify the number of positive cells and the mean intensity of mCherry fluorescence, the product of which is reported as transduction efficiency for each vector. The scAAV vectors containing cone-specific (sc-mCARpro-green fluorescent protein [GFP]), rod-specific (sc-MOPS500-eGFP), retinal pigment epithelium (RPE)-specific (sc-VMD2-GFP), or ubiquitous (sc-smCBA-GFP) promoters were used to infect both cell lines at an MOI of 10,000. Three days post infection, cells were immunostained with an antibody raised against GFP and imaged. Finally, based on our in vitro results, we tested a prediction of transduction efficiency in vivo.
Results: Expression from unmodified scAAV1, scAAV2, scAAV5, and scAAV8 vectors was detectable by FACS in both ARPE19 and 661W cells, with scAAV1 and scAAV2 being the most efficient in both cell lines. scAAV5 showed moderate efficiency in both ARPE19 and 661W cells. scAAV8 was moderately efficient in 661W cells and was by comparison less so in ARPE19 cells; however, transduction was still apparent. scAAV9 performed poorly in both cell types. With some exceptions, the Y-F capsid mutations generally increased the efficiency of scAAV vector transduction, with the increasing number of mutated residues improving efficiency. Results for single scAAV1 and scAAV8 capsid mutants were mixed. In some cases, efficiency improved; in others, it was unchanged or marginally reduced. Retinal-specific promoters were also active in both cell lines, with the 661W cells showing a pattern consistent with the in vivo activity of the respective promoters tested. The prediction based on in vitro data that AAV2 sextuple Y-F mutants would show higher transduction efficiency in RPE relative to AAV2 triple Y-F capsid mutants was validated by evaluating the transduction characteristics of the two mutant vectors in mouse retina.
Conclusions: Our results suggest that this rapid and quantifiable cell-based assay using two biologically relevant ocular cell lines will prove useful in screening and optimizing AAV vectors for application in retina-targeted gene therapies.