Does recombinant adeno-associated virus-vectored proximal region of mouse rhodopsin promoter support only rod-type specific expression in vivo?

PURPOSE We have previously found that the -385 to +86 portion of the mouse rod opsin promoter (mOP500) can limit recombinant adeno-associated virus (rAAV)-mediated transgene expression to photoreceptor cells when delivered subretinally. However, the photoreceptor (PR) subtype-specificity of expression remains unclear. Here, we evaluated whether the presence of certain cis-elements in this proximal promoter, such as the rod-specific, neural retina leucine zipper protein (NRL) response element (NRE), can render it a driver of rod-specific expression. METHODS Subretinal injections of a serotype 5 rAAV vector carrying the green fluorescent protein (GFP) cDNA, driven by mOP500, were administered to male Sprague-Dawley rats at postnatal day (P) 40-48. Two weeks to eight months later, the distribution of GFP-expressing cells in the retina was characterized by GFP-, cone-specific alpha-transducin-immuno-, and peanut agglutinin-lectin histochemistry and by morphological criteria. The same viral suspension was also injected sub-retinally into rhodopsin-knockout rho (-/-) mice either at P18 or P78, and retinas were analyzed by immunohistochemistry and PNA lectin histochemistry two weeks later. RESULTS GFP reactivity was found exclusively in the outer nuclear layer (ONL) of rat retinas two weeks after treatment, with abundant reporter gene expression observed in both rods and cones. GFP-positive cones, defined by their typical morphology and the co-linearity of PNA-lectin labeling with GFP-immunoreactivity, were found in all regions of the transduced retinas. GFP-positive cones constituted up to 6% of the total GFP-positive photoreceptors. By eight months post-injection, a low level of GFP-reactivity was additionally observed in the inner nuclear layer (INL) and ganglion cell layer. Photoreceptor-specific GFP expression was also seen in the rho (-/-) mice at both ages tested. In pups injected at P18, costaining with PNA-lectin revealed that up to 15% of the GFP-positive photoreceptors were cones. Despite only a single row of photoreceptors remaining in these knockout mice by P90, numerous GFP-positive cones were still present. CONCLUSIONS Subretinal delivery of rAAV5 harboring a reporter gene driven by mOP500 results in passenger gene expression in both rod and cones, indicating that this promoter is photoreceptor-specific but not rod-specific. The lack of photoreceptor subtype-specificity suggests that although cones do not express the NRL and NR2E3 trans-factors considered necessary for activation of mOP500, other general transcription factors in cones may compensate.