No abstract

Comparative analysis of the transcriptional profiles of approximately 6000 genes in the retinas of wild-type mice with those carrying a targeted disruption of the rhodopsin gene was undertaken by microarray analysis. This revealed a series of transcripts, of which some were derived from genes known to map at retinopathy loci, levels of which were reduced or elevated in the retinas of Rho(-/-) mice lacking functional photoreceptors. The human homologue of one of these genes, encoding inosine monophosphate dehydrogenase type 1 (IMPDH1), maps to the region of 7q to which an adRP gene (RP10) had previously been localized. Mutational screening of DNA from the Spanish adRP family, originally used to localize the RP10 gene, revealed an Arg224Pro substitution co-segregating with the disease phenotype. The amino acid at position 224 of the IMPDH1 protein is conserved among species and the substitution is not present in healthy, unrelated individuals of European origin. These data provide strong evidence that mutations within the IMPDH1 gene cause adRP, and validate approaches to mutation detection involving comparative analysis of global transcription profiles in normal and degenerating retinal tissues. Other genes showing significant alterations in expression include some with anti-apoptotic functions and many encoding components of the extracellular matrix or cytoskeleton, a possible reflection of a response by Muller cells to preserve the remaining outer nuclear layer of the retina. We suggest that those genes identified are prime candidates for etiological involvement in degenerative retinal disease.

Retinitis pigmentosa (RP), the hereditary degenerative disease of the photoreceptor neurons of the retina, probably represents the most prevalent cause of registered blindness amongst those of working age in developed countries. Mutations within the gene encoding inosine monophosphate dehydrogenase 1 (IMPDH1), the widely expressed rate-limiting enzyme of the de novo pathway of guanine nucleotide biosynthesis, have recently been shown to cause the RP10 form of autosomal dominant RP. We examined the expression of IMPDH1, IMPDH2 and HPRT transcripts, encoding enzymes of the de novo and salvage pathways of guanine nucleotide biosynthesis, respectively, in retinal sections of mice, the data indicating that the bulk of GTP within photoreceptors is generated by IMPDH1. Impdh1(-/-) null mice are shown here to display a slowly progressive form of retinal degeneration in which visual transduction, analysed by electroretinographic wave functions, becomes gradually compromised, although at 12 months of age most photoreceptors remain structurally intact. In contrast, the human form of RP caused by mutations within the IMPDH1 gene is a severe autosomal dominant degenerative retinopathy in those families that have been examined to date. Expression of mutant IMPDH1 proteins in bacterial and mammalian cells, together with computational simulations, indicate that protein misfolding and aggregation, rather than reduced IMPDH1 enzyme activity, is the likely cause of the severe phenotype experienced by human subjects. Taken together, these findings suggest that RP10 may represent an attractive target for therapeutic intervention, based upon a strategy combining simultaneous suppression of transcripts from normal and mutant IMPDH1 alleles with supplementation of GTP within retinal tissues.

The intragenic heterogeneity encountered in many dominant disease-causing genes represents a significant challenge with respect to development of economically viable therapeutics. For example, 25% of autosomal dominant retinitis pigmentosa is caused by over 100 different mutations within the gene encoding rhodopsin, each of which could require a unique gene therapy. We describe here an RNA interference (RNAi)-based mutation-independent approach, targeting as an example murine rhodopsin. Native transcripts are suppressed by a single RNAi molecular species, whereas transcripts from replacement genes engineered at degenerate third-codon wobble positions are resistant to suppression. We demonstrate suppression of murine rhodopsin transcript by up to 90% with full concomitant expression of replacement transcript and establish the validity of this approach in cell culture, retinal explants, and mouse liver in vivo.

Mutations within the inosine 5'-monophosphate dehydrogenase 1 (IMPDH1) gene cause the RP10 form of autosomal dominant retinitis pigmentosa (adRP), an early-onset retinopathy resulting in extensive visual handicap owing to progressive death of photoreceptors. Apart from the prevalence of RP10, estimated to account for 5-10% of cases of adRP in United States and Europe, two observations render this form of RP an attractive target for gene therapy. First, we show that while recombinant adeno-associated viral (AAV)-mediated expression of mutant human IMPDH1 protein in the mouse retina results in an aggressive retinopathy modelling the human counterpart, expression of a normal human IMPDH1 gene under similar conditions has no observable pathological effect on retinal function, indicating that over-expression of a therapeutic replacement gene may be relatively well tolerated. Secondly, complete absence of IMPDH1 protein in mice with a targeted disruption of the gene results in relatively mild retinal dysfunction, suggesting that significant therapeutic benefit may be derived even from the suppression-only component of an RNAi-based gene therapy. We show that AAV-mediated co-expression in the murine retina of a mutant human IMPDH1 gene together with short hairpin RNAs (shRNA) validated in vitro and in vivo, targeting both human and mouse IMPDH1, substantially suppresses the negative pathological effects of mutant IMPDH1, at a point where, in the absence of shRNA, expression of mutant protein in the RP10 model essentially ablates all photoreceptors in transfected areas of the retina. These data strongly suggest that an RNAi-mediated approach to therapy for RP10 holds considerable promise for human subjects.