Exploration of an AAV-mediated TULP1 replacement gene therapy and characterization of the Tulp1-/- murine model

Abstract

The vertebrate family of tubby-like proteins (TULPs) include the founding member TUB and related members of the tubby-like protein family such as TULP1, TULP2, TULP3, TULP4. TULP1 encoded by TULP1 gene is believed to play a crucial role in retinal physiology due to its selective expression in the retina. Mutations in TULP1 are causative of early onset recessive retinal degenerations. The majority of TULP1-linked patients manifest similar disease phenotypes such as; rapid photoreceptor degeneration with granular pigmentation, retinal pigment epithelial (RPE) atrophy, optic disc pallor, maculopathy the affected rods and cones resulting nystagmus and night blindness during the first weeks of life. Absence of TULP1 in photoreceptors is thought to be the major contributor to the disease. There is an underlying principle that replacing TULP1 should prevent the devastating events that drive photoreceptors towards the fate of cell death in TULP1-linked disease. For this reason, an AAV-mediated TULP1 replacement therapy was explored during the course of the study presented in this thesis. The study presented in Chapter 2 of this thesis involved the designing, cloning and producing of high titre AAV2/5 vectors encoding a functional human TULP1 gene. In the study the Tulp1-/- mouse, a naturally occurring mouse model of a recessively inherited retinopathy was employed. Similar to the human patients, the retinopathy in Tulp1-/- mice presents as a severe, very early onset disease and hence delivery of gene therapy to young mouse pups was required in order to provide benefit. The generated AAV-TULP1 was administered subretinally at postnatal day (p) 1-2 to provide sufficiency levels of TULP1 in photoreceptor cells at early stage. To evaluate the AAV-mediated TULP1 replacement therapy in the Tulp1-/- murine model, the research described in Chapter 2 of this thesis employed a variety of assays to evaluate the effects of the therapy. Interestingly, initial studies were undertaken with the AAV-TULP1 gene replacement vector and no rescue of the disease phenotype was observed. Subsequently, the AAV2/5 vectors were re-engineered to contain the codon-optimized mouse Tulp1 replacement gene to determine if more effective rescue in this mouse model might be achieved using this mouse gene, given the divergence between the mouse Tulp1 and human TULP1 genes. However, the optimization and evaluation of novel AAV-Tulp1 gene therapy showed no morphological nor functional rescue in Tulp1-/- even though the correct localization of TULP1 protein and efficient Tulp1 mRNA expression were obtained. Chapter 3 focused on characterizing the nature of retinal degeneration in the mouse model in greater depth in addition to mouse models of inherited retinopathies such as Rds-/- and Rho-/- mice. The study revealed that common features of photoreceptor degeneration were detected in Tulp1-/-, Rho-/- and Rds-/- retinas. In contrast, other alterations in bipolar, amacrine and ganglion cells, were specific to Tulp1-/- mice and suggest early involvement of the inner retina in TULP1-linked inherited retinal degenerations (IRDs). Peak expression of Tulp1 was detected in wild type mouse retina RNA at p8 and protein TULP1 in the inner nuclear layer at p5-8. Bioinformatic analysis indicated Tulp1 or TULP1 expression in retinal progenitor and non-photoreceptor cells. Additionally, TULP1 interactors differed significantly in different retinal cell types, suggesting TULP1 may have different functions in these cell types. The early expression of TULP1 in the inner retina indicates that perhaps gene therapies targeting TULP1 may need to be expressed in both the inner and outer retina. Chapter 4 involved evaluating the impact of inner retinal alterations on retinal vasculature in Tulp1-/- murine model including a number of other IRD mouse models such as Rds-/-, P23H and P347S mice in comparison with wild type mouse lines available in the TCD laboratory. Attenuation of retinal blood vessels is hallmark of IRDs, typically reported in conjunction with photoreceptor death. Exploring retinal flatmounts, we tested early changes of the retinal vasculature in IRD mouse models using AngioTool and morphometric analyses. The correlation between retinal degeneration and vascular changes was evaluated by advanced statistical analysis. The study demonstrated that retinal vasculature in deep plexus was compromised significantly in all tested IRD models showing a strong correlation between photoreceptor loss and vascular alterations in deep plexus. Interestingly, the retinal vasculature in the superficial and intermediate plexuses was significantly altered in Tulp1-/- and Rds-/-. GFAP down-regulation was also characteristic for those models suggesting that the inner retinal remodelling (early observed in Chapter 3) may affect the inner retinal vascular growth and development. The work presented in the thesis provides new insights into the complex retinal degeneration in TULP1-linked IRDs and potential challenges in treating patients with this condition.