RPE65 is specifically expressed in the retinal pigment epithelium and is essential for the recycling of 11-cis-retinal, the chromophore of rod and cone opsins. In humans, mutations in RPE65 lead to Leber congenital amaurosis or early-onset retinal dystrophy, a severe form of retinitis pigmentosa. The proof of feasibility of gene therapy for RPE65 deficiency has already been established in a dog model of Leber congenital amaurosis, but rescue of the cone function, although crucial for human high-acuity vision, has never been strictly proven. In Rpe65 knockout mice, photoreceptors show a drastically reduced light sensitivity and are subject to degeneration, the cone photoreceptors being lost at early stages of the disease. In the present study, we address the question of whether application of a lentiviral vector expressing the Rpe65 mouse cDNA prevents cone degeneration and restores cone function in Rpe65 knockout mice. Subretinal injection of the vector in Rpe65-deficient mice led to sustained expression of Rpe65 in the retinal pigment epithelium. Electroretinogram recordings showed that Rpe65 gene transfer restored retinal function to a near-normal pattern. We performed histological analyses using cone-specific markers and demonstrated that Rpe65 gene transfer completely prevented cone degeneration until at least four months, an age at which almost all cones have degenerated in the untreated Rpe65-deficient mouse. We established an algorithm that allows prediction of the cone-rescue area as a function of transgene expression, which should be a useful tool for future clinical trials. Finally, in mice deficient for both RPE65 and rod transducin, Rpe65 gene transfer restored cone function when applied at an early stage of the disease. By demonstrating that lentivirus-mediated Rpe65 gene transfer protects and restores the function of cones in the Rpe65^−/− mouse, this study reinforces the therapeutic value of gene therapy for RPE65 deficiencies, suggests a cone-preserving treatment for the retina, and evaluates a potentially effective viral vector for this purpose.

The polycomb transcriptional repressor Bmi1 promotes cell cycle progression, controls cell senescence, and is implicated in brain development. Loss of Bmi1 leads to a decreased brain size and causes progressive ataxia and epilepsy. Recently, Bmi1 was shown to control neural stem cell (NSC) renewal. However, the effect of Bmi1 loss on neural cell fate in vivo and the question whether the action of Bmi1 was intrinsic to the NSCs remained to be investigated. Here, we show that Bmi1 is expressed in the germinal zone in vivo and in NSCs as well as in progenitors proliferating in vitro, but not in differentiated cells. Loss of Bmi1 led to a decrease in proliferation in zones known to contain progenitors: the newborn cortex and the newborn and adult subventricular zone. This decrease was accentuated in vitro, where we observed a drastic reduction in NSC proliferation and renewal because of NSC-intrinsic effects of Bmi1 as shown by the means of RNA interference. Bmi1^-/- mice also presented more astrocytes at birth, and a generalized gliosis at postnatal day 30. At both stages, colocalization of bromodeoxyuridine and GFAP demonstrated that Bmi1 loss did not prevent astrocyte precursor proliferation. Supporting these observations, Bmi1^-/- neurospheres generate preferentially astrocytes probably attributable to a different responsiveness to environmental factors. Bmi1 is therefore necessary for NSC renewal in a cell-intrinsic mode, whereas the altered cell pattern of the Bmi1^-/- brain shows that in vivo astrocyte precursors can proliferate in the absence of Bmi1.

Ciliary neurotrophic factor (CNTF) has recently been demonstrated to be one of the most promising neurotrophic factors to improve both the survival and regeneration of injured retinal ganglion cells (RGCs). In the present study, we used optic nerve transection as an in vivo model to evaluate the effectiveness of a self-inactivating, replication-deficient lentiviral-mediated transfer of human ciliary neurotrophic factor (SIN-PGK-CNTF) on the survival of axotomized adult rat RGCs. Counts of dextran-fluorescein isothiocyanate conjugated (D-FITC)-retrogradely labeled RGCs revealed that the percentage of RGCs was drastically reduced (<90% cell death) 21 days after optic nerve transection. Retinal sections stained with X-gal revealed that intravitreal injection of the control LacZ-expressing lentiviral vector (LV-LacZ) resulted in the transduction of RGCs and retinal pigment epithelium (RPE) cells. A single intravitreal injection of LV-CNTF at the time of axotomy significantly enhanced RGC survival at 14 and 21 days postaxotomy compared to controls. These results demonstrate for the first time that rapid and prolonged delivery of CNTF using lentiviral-mediated gene transfer to the retina is an effective treatment for rescuing axotomized RGCs for an extended period of time. These results suggest that early and continuous administration of CNTF could serve as a potential treatment for retinal disorders involving optic neuropathy and RGC injury such as in glaucoma.

Retinitis pigmentosa (RP) is a retinal degenerative disease characterized by the progressive loss of photoreceptors. We have previously demonstrated that RP can be caused by recessive mutations in the human FAM161A gene, encoding a protein with unknown function that contains a conserved region shared only with a distant paralog, FAM161B. In this study, we show that FAM161A localizes at the base of the photoreceptor connecting cilium in human, mouse and rat. Furthermore, it is also present at the ciliary basal body in ciliated mammalian cells, both in native conditions and upon the expression of recombinant tagged proteins. Yeast two-hybrid analysis of binary interactions between FAM161A and an array of ciliary and ciliopathy-associated proteins reveals direct interaction with lebercilin, CEP290, OFD1 and SDCCAG8, all involved in hereditary retinal degeneration. These interactions are mediated by the C-terminal moiety of FAM161A, as demonstrated by pull-down experiments in cultured cell lines and in bovine retinal extracts. As other ciliary proteins, FAM161A can also interact with the microtubules and organize itself into microtubule-dependent intracellular networks. Moreover, small interfering RNA-mediated depletion of FAM161A transcripts in cultured cells causes the reduction in assembled primary cilia. Taken together, these data indicate that FAM161A-associated RP can be considered as a novel retinal ciliopathy and that its molecular pathogenesis may be related to other ciliopathies.