Stephen H. Tsang, MD, PhD - US grants

This proposal is designed to provide a diverse and extensive research experience in molecular biology, biochemistry, and electrophysiology that will allow the applicant to develop into a mature scientist-clinician with an outstanding potential to contribute to both academic ophthalmology and basic science research. All components of training are incorporated in the curriculum of the EyeSTAR program based at UCLA. The research project will focus on the generation of tissue- and time- specific modifications in the gene encoding the gamma subunit of cGMP- phosphodiesterase (PDEgamma) by using an inducible knockout system. The entire coding region of the PDEgamma will be flanked by 34-base pair Cre-recombinase recognition sites (loxP) introduced into the gene locus by homologous recombination. After germline transmission, mice Pdeglox1/Pdegtm1 bearing the modified allele will be crossed with mice expressing the Cre-recombinase/steroid hormone receptor fusion protein under the control of the genomic locus of the beta subunit of PDE (Pdeb). The use of Cre-recombinase/steroid hormone receptor fusion protein will allow the regulation of Cre-recombinase activity with exogenously applied steroids. Gaining temporal and spatial control of gene expression is essential for the elucidation of gene function in the whole organism. The proposed studies will develop tools that will ease the deletion of genes from the genome in specific cells and at specific times. Understanding of the mechanisms controlling photoreceptor signaling and degeneration in Pdeglox1/Pdegtm1 mice may help to develop strategies for the prevention or slowing of human retinal dystrophies. Dr. Debora Farber is nationally recognized for her biochemical and molecular biological studies of retinal degenerations, in particular retinitis pigmentosa. Dr. Farber will serve as mentor for the PI'S postdoctoral studies. In addition, UCLA has an outstanding panel of faculty members many of whom have professional relationships with Dr. Farber and can serve as scientific consultants to the PI. The proposed project will be at the cutting edge of current research is an ideal area with which to begin a scientific career. The EyeSTAR program also incorporates three years of clinical training in ophthalmology leading to certification and specialty board eligibility. This intense three-year residency program providers a vast clinical experience in diagnosis and treatment of the complete scope of ophthalmologic problems as well as a number of didactic activities including lectures, special seminars, and attendance at clinical research meetings.

DESCRIPTION (provided by applicant): About 36,000 cases of simplex and familial retinitis pigmentosa (RP) worldwide are due to defects in rod-specific PDE6, which consists of catalytic (PDE6a and PDE6b) and regulatory (PDE6g) subunits. In mouse models, low PDE6 activity leads to RP-like features. The Pde6brd1 null mouse exhibits dramatic elevation of retinal cGMP and rapid rod degeneration, which is complete within 3 weeks after birth. There are likely multiple changes in intercellular signaling induced by excessive cGMP, but the exact mechanisms underlying these pathways are unknown. A recently discovered weak Pde6b allele, H620Q, can further our understanding of early degeneration mechanisms and enable us to test novel therapeutic hypotheses. Like Pde6brd1 mice, Pde6bH620Q mutants show RP-like features and dramatic elevation of retinal cGMP. However, the time course of degeneration of Pde6bH620Q is significantly slower, with complete rod loss occurring after 6 weeks. Moreover, for the first three weeks after birth Pde6bH620Q display relatively normal rod histology and quantifiable rod physiology, which is not the case in Pde6brd1 mice. Our long-term goal is to use Pde6bH620Q mice to find therapies that inhibit the intracellular effects of excessive cGMP and/or enhance PDE6 specific activity to prevent further rod and cone degeneration. Aim 1. Establish if lower than normal PDE6 activity, combined with normal guanylate cyclase activity, results in elevated cGMP levels in Pde6bH620Qmutant mouse. Aim 2. Identify kinomic (kinases in the genome) survival and apoptotic effectors of cGMP rise in Pde6bH620Q mutants before morphological signs of degeneration. Such effectors will provide novel pharmacological targets to retard degeneration. Aim 3. Determine if progressive rod-cone degeneration can be genetically arrested by increasing PDE6 activity in Pde6b mutant rods. We intent to halt degeneration using Opsin::Pde6b rescue transgene and subretinal injections of Opsin::Pde6b rescue lentivirus. To test if the rod as well as gradual secondary cone loss can be arrested after the onset of rod death, we will employ a tamoxifen-inducible reverse Cre/loxP system to restore wild-type Pde6b in expression in mid-phase of the disease. PUBLIC HEALTH RELEVANCE: Inherited forms of retinal degeneration are incurable and affect about one in 2000 people; 1.5 million people worldwide are affected by retinitis pigmentosa (RP). Can the remaining rod and cone photoreceptor death be halted once degeneration has begun? Our proposal addresses a clinically relevant question, as most retinal degeneration patients have significant night blindness (rod death) when they are first seen by an optometrist/ophthalmologist. If one can impede further rod and cone loss by correcting the primary cause of the pathogenesis at the mid-stage of disease, then there is hope for a drug- or gene-based therapy to restore activities of daily living for newly diagnosed patients.

0917458
Tsang

The eye captures light but the brain is where vision is experienced. The capacity of retinal prostheses to restore some vision in blind individuals has been demonstrated at the retinal level, but it is unknown how the brain will be receptive to the new neural message after a prolonged period of visual deprivation resulting from retinal degenerative diseases. Neural remodeling that occurs in retinal degenerative diseases can have an impact on the usefulness of retinal prostheses in visual restoration. Mammalian retinal degenerations, initiated by gene defects in rods, cones or the retinal pigmented epithelium, often trigger loss of the photoreceptors; thus, effectively leaving the neural retina without sensory input. The neural retina responds to this challenge by remodeling, first by subtle changes in neuronal structure and later by large-scale reorganization. This type of retinal remodeling can alter normal light transmission pathway in the retina and result in activation of different ganglion cell groups between light and electrical stimulation.
In this proposal, the PI proposes to use transcorneal electrical stimulation (TcES) to probe the neural connection between the retina and the visual cortex via direct stimulation of the retinal ganglion cells from a current passing through the cornea. By using positron emission tomography (PET) and 18F-fluorodeoxyglucose (FDG) to compare brain activation between light stimulation and electrical stimulation of the retina in patients with RP, we will be increase our understanding how retinal remodeling is reflected in the visual cortex. The study can then help us gain insight into cortical reorganization resulting from the retinal degenerative process.

PROJECT SUMMARY Of the retinal degenerative diseases that affect 9 million Americans, cone photoreceptor dystrophies are arguably the most devastating. Gene therapy is a potential means to strengthen photoreceptor viability. However, the first human gene therapy trial for retinal degeneration found improved visual function but did not slow degeneration of photoreceptors. The goal of this gene therapy-oriented proposal is to determine whether therapy is achievable in the context of an already diseased retina and if metabolic reprogramming could be an efficacious treatment option. During the previous funding period, we succeeded in restoring retinal function for more than 11 months in a mouse model of rod degeneration even after the onset of degeneration and at late-stage disease. We now intend to determine whether the same outcomes are achievable in cone-based dystrophies. To do this, we will generate a novel, inducible genetic rescue system in the cone-specific G-protein, guanine nucleotide binding ?- transducin 2 (Gnat2), which will allow us to conditionally reverse GNAT2-deficiency while controlling the temporal and spatial aspects of phenotypic reversal. Using Gnat2floxSTOP/Gnat2CreERT2, we will establish that the model faithfully recapitulates cone-mediated dystrophies (Aim 1). We will restore the model to wild type via tamoxifen injection at early, middle, and late disease stages and assess effects on the rate of degeneration (Aim 2) to determine the temporal limitations of gene therapy. Finally, we will induce metabolic reprogramming and assess its utility as a possible non-gene- specific strategy for treating cone degenerations -based dystrophies (Aim 3). The Gnat2floxSTOP/Gnat2CreERT2 programmable model will provide a platform for contributing to ongoing efforts aimed at increasing restoration of visual function following gene therapy for cone-mediated dystrophies. It will also allow us to address several compelling, clinically relevant questions: Is the brain?s circuitry sufficiently plastic to recover from the pathological changes caused by the Gnat2 mutation? Is there a point of no return after which, despite reversion of the genotype to wild type, cones cannot be salvaged? Can temporal barriers to gene therapy be relieved by metabolic reprogramming? Taken together, this proposal is certain to 1) define the factors limiting interventional therapy; 2) validate a new, inducible model of cone-mediated retinal degeneration; and 3) determine whether metabolic reprogramming can serve as an efficacious, non-gene-specific strategy for treating retinal degeneration.

? DESCRIPTION (provided by applicant): Age-related macular degeneration (AMD) is a major cause of retinal damage, the leading cause of blindness in Western countries and, as the name implies, a disease of aging. Like other age-related diseases, AMD is particularly challenging to study because it takes decades to develop and so any research model must recapitulate the conditions of an older organism. Genome-wide association studies (GWASs) and linkage analyses have provided the first clues to what might cause AMD. These studies identified three single nucleotide polymorphisms (SNPs) that are strong risk factors for AMD.1 One SNP lies in the 402H allele in the CFH gene and the two others are tightly linked and lie in the neighboring HTRA1 and ARMS2 genes. These SNPs confer the most significant genetic risk factors in the history of GWAS studies in human genetics. People homozygous for these SNPs have a 50-fold increased risk of AMD. How these mutations might cause sight to deteriorate is unclear, however, because the underlying molecular mechanisms of AMD are unknown. Recently, however, our unbiased proteome analysis suggested super oxide dismutase (SOD) mechanisms are perturbed in affected cells and that, over time, this introduces reactive oxidative species (ROS) mediated cellular insults that eventually manifest as AMD. If ROS metabolism is indeed disrupted in AMD, then we might finally begin to address the causes of the disease. We believe the hurdles faced in finding causes and treatments for AMD could be circumvented by stem cell technologies. To this end we have found a way to differentiate stem cells from patients into retinal cells. Moreover, we developed a protocol that recapitulates aging in these patient-stem-cell-derived retinal cells. Finally, through gene-targeting technology, we can manipulate the stem cell genome, targeting disease- associated SNPs, to determine the individual contributions of each. By applying these powerful methodologies, we believe we can finally identify the root causes of AMD and so begin to develop new therapies. Our goals will be accomplished in two specific aims: Aim 1A. Use the CRISPR/Cas9 system to convert HTRA1 and ARMS2 alleles from low-risk to high risk in patient-derived stem cells. Aim 1B. Determine the individual contribution of human HTRA1 and ARMS2 alleles to AMD pathogenesis. Test whether CRISPR conversion from low to high-risk AMD alleles in Aim 1A affects ROS levels in cells. Aim 2. Test the function of patient-stem-cell-derived RPE in a human-mouse chimera, in vivo assay.

? DESCRIPTION (provided by applicant): Retinal degenerative diseases affect 9 million Americans. Among these conditions, retinitis pigmentosa (RP) is among the most devastating. Mutations in genes encoding subunits of the rod-specific enzyme, cyclic guanosine monophosphate (cGMP) phosphodiesterase 6 (PDE6A and PDE6B), are responsible for approximately 72,000 cases of RP worldwide each year, making therapeutic modeling highly relevant to developing mechanisms based therapies. In both RP and age-related macular degeneration (AMD), progressive atrophy of rod photoreceptors leads to secondary death of cone photoreceptors. Gene therapy could enhance rod viability and prevent secondary cone loss. These FDA trials used wild-type (wt) gene supplementation (i.e., overexpression of a normal version of the gene) in diseased cells to overcome the abnormality of the patients' mutated genes. The first human gene therapy trial for early onset retinal degeneration found visual function recovered initially, but did not retard the rate of photoreceptor degeneration, and the gene therapy treated RP patients now continue their march toward blindness. The failure of FDA trials to attenuate the progression of rod death suggests that there is a point of no return for rod viability in retinal disease and presents a major obstacle to the treatment of RP. We hypothesize that this retinal point of no return derives from 1) changes in Ca2+ homeostasis and 2) inadequate activation of the mammalian target of rapamycin (mTOR) self-survival pathway. Preliminary data suggest that the point of no return could be halted by administering two therapeutic constructs via bipartite vectors in an autosomal recessive PDE6-RP model (arRP) (Aims 1 and 2). In photoreceptors, incoming light activates PDE6, which hydrolyzes free cGMP. Lower free cGMP levels close cGMP-gated Na+/Ca2+ cation (CNG) channels in the plasma membrane, reducing cation influx and propagating nerve impulses. Aims 1 and 2 test therapeutic strategies aimed at remedying PDE6 deficiency. Specifically, shRNA knockdown will be used to identify therapeutic targets using two novel bipartite AAV8 vectors described in detail herein. Finally, wt gene supplementation leaves the patient's mutant genes intact, which could continuously trigger ongoing damage despite the presence of a wt gene in a diseased cell. A gene editing approach could overcome this defect. Thus, Aim 3 explores in vivo AAV mediated CRISPR-gene editing in PDE6-RP models available on campus: 3a) the Pde6a mutant mouse; and 3b) human stem cells from an RP patient bearing the PDE6A mutations (OMIM# 180071), which offers an in vitro model for comparing CRISPR efficacy in human cells vs. Pde6a mouse retina.

? DESCRIPTION (provided by applicant): Stem cell therapy finds an ideal proving ground in the eye, an organ with relative immune privilege that is accessible yet isolated. Therapies for the eye are generally neither invasive nor systemic; and since the eye is optically transparent, treatment can be monitored easily and non-invasively in living animals-a major advantage over other systems. In preclinical studies of retinal disease, grafts of healthy retinal pigment epithela (RPE) can restore damaged retina. Looking ahead to clinical applications of RPE grafting, we posit that RPE grafts grown from autologous stem cells (a patient's own stem cells) would be the optimal approach to develop. Nevertheless, clinical use of this method awaits resolution of several knowledge gaps: Are an adult's stem cells pluripotent enough to render functional RPE grafts? Even if patient grafts are functional, will in vitro culture make them antigenic? Also, eve if autologous RPE grafts work, will gene therapy be more effective at late stages of RP? These questions are addressed by Aims 1, 2, and 3, respectively. The therapies explored here aim to repair hereditary retinal degeneration. Although, in general our results will pertain to retinitis pigmentosa (RP), these studies focus on RP caused by rare mutations in membrane frizzled- related gene (MFRP). MFRP retinopathy is ideal for our studies for three reasons. First, known MFRP point mutations cause an RP phenotype. In addition, MFRP retinopathy has a decades-long window of opportunity for treatment: Over the life span, electroretinogram (ERG) changes lag behind photoreceptor loss, suggesting that despite retinal damage, MFRP retinopathy patients retain sight and are likely treatable until late stages of the disease. The third reason fr choosing MFRP retinopathy is that we can take advantage of the well characterized, MFRP mutant mouse line, rd6 that also suffers blinding RP. For our study, MFRP-deficient patient stem cells will be isolated, cultured, and transduced with an Adeno-Associated-virus (AAV) to express wild-type MFRP. These MFRP gene repaired patient stem cells will be differentiated into RPE and grafted into right eyes of immunodeficient (Scid), MFRP-deficient (rd6) mice. Left and right eyes will be compared for reversal of the RP phenotype. Our long-term goals are to create cell- and gene-therapy cures for hereditary retinal diseases, and develop strategies that extend to other diseases. Our objective in this proposal is to find ways to use patient stem cells as a source of retinal grafts. We will test our central hypothesis that patient- derived RPE grafts, repaired by AAV in vitro, are already a therapeutic option.

Currently there is no treatment for autosomal recessive retinitis pigmentosa (arRP) due to mutations in rod photoreceptor genes. This multi-investigator, multi-center research project will fill this gap by optimizing recombinant adeno-associated virus (rAAV) vector gene augmentation therapy for cyclic nucleotide-gated channel beta 1 (CNGB1) linked RP. The optimized vector will be taken through the stages needed for an investigational new drug (IND) submission in preparation for a future phase I/II clinical trial. The collaborative team has expertise in rAAV vector development and production, preclinical and clinical trials for retinal dystrophies, clinical assessment and recruitment of arRP patients, CNG channel physiology and small and large animal proof-of-concept gene therapy studies. There are compelling reasons to select CNGB1-RP to fill this unmet need. First, CNGB1 mutations cause a loss of rod function, but only a slow loss of rods, meaning there is a wide window of opportunity for intervention while there are still remaining rods. Second, there are well characterized small (mouse) and large (dog) animal models of CNGB1-RP that recapitulate the human phenotype. Third, in both CNGB1-RP models rAAV gene augmentation therapy can efficiently (1) rescue the function and (2) delay the degeneration of rods. There are four aims to the project: Aim 1 is development of an optimized vector that efficiently and specifically targets rod photoreceptors in nonhuman primates. We will start with an efficient vector with a new short rhodopsin promoter that has already shown efficacy in CNGB1-RP animal models. The final rAAV-CNGB1 vector will be used to start the Good Manufacturing Practice (GMP) process development. This final vector will be used to investigate duration of rescue achievable in mouse and dog CNGB1-RP models and also answer the important question of how late in the process of rod degeneration can rescue and preservation of structure be obtained?? Aim 2 will consist of recruitment of candidate patients for the clinical trial. A barrage of clinical testing methods will be used over a three year period to precisely describe the RP phenotype and identify optimal outcome measures for a future clinical trial. The testing will also ascertain if there is similar disease progression between the two eyes which would allow the second eye to be used as a non treatment control. Aim 3 will consist of the animal toxicology and pharmacokinetic and efficacy studies needed for IND submission. We propose a standard GLP toxicology/biodistribution study in rats coupled with a hybrid efficacy/safety study using the CNGB1-RP dog model. A pre-IND meeting with the FDA will review the study design and ensure that it meets regulatory requirements. GMP vector production will then be completed. Finally, in Aim 4 we will prepare and submit an IND application. This project fulfills the FOA goal of ?development of a therapeutic, which can then be tested in a clinical trial.?

Retinal pigmented epithelial (RPE) disorders, including autosomal dominant retinitis pigmentosa (RP) and age- related macular degeneration, are an enormous burden and a growing public health concern, given the aging U.S. population. Autosomal dominant (ad) conditions emerge from mutations that allow expression of a defective protein and are generally not amenable to current ?gene addition? therapies in humans, as they require precise repair to remove the gain-of-function mutation. The long-term goal is to identify ways to use patient stem cells as a tissue source for regenerative medicine (RM). The objective of this application is to determine whether autologous induced pluripotent stem (iPS) cell transplantation is a RM approach or treatment of dominant disorders. The eye provides an ideal proving ground for RM therapy, as it is an organ with relative immune privilege that is both accessible and isolated, therapies for the eye are generally neither invasive nor systemic, and its optically transparency allows treatment to be monitored easily and noninvasively in living animals. The central hypothesis is that a gene-editing strategy using a novel state-of-the-art therapeutic editing approach can suppress manifestation of a dominant RPE65D477G/+ mutation in human RP. This hypothesis will be tested by pursuing three specific aims: 1) Obtain functionally recovered RPE after CRISPR repair of corresponding RPE65D477G/+ iPS cells. Gene-editing tools and methods will be optimized for CRISPR repair of the RPE65D477G/+ mutation in iPS cells. The cells will be assessed to determine whether they are able to differentiate into RPE and to confirm their genomic integrity. 2) Determine whether in vitro- repaired patient iRPE functionally integrates into the retinae of live mice by assessing inner retinal electrophysical signals and retinoid signaling. RPE grafts will be derived from either RPE65 gene-repaired or unrepaired patient iPS cells. 3) Determine whether transplanted gene-repaired iRPE is safe and nontumorigenic. Immunocompromised Rpe65rd12/Rpe65rd12; Prkdcscid/Prkdcscid mice will be subretinally injected with iRPE grafts and tested for safety, tumorigenicity and survival. The proposed research is significant, as it will advance treatment guided by the precise pathophysiology of patient-specific mutations. The proposed research is innovative because it uses a new approach to generate repaired RPE cells for transplantation, should not require immunosuppressive therapy, and make use of innovative outcome measurements. The research is expected to have an important positive impact as it represents the initial steps of a precision medicine approach directed toward developing treatments targeting dominant mutations unique to individual patients with RP, AMD, and other dominant disorders, while introducing new methods and model systems. .

Summary FDA-approved gene therapy trials have treated autosomal-recessive (i.e., loss-of-function) disorders by supplementation with the wild-type (WT) version of the mutant gene. For patients with autosomal-dominant (ad) gain-of-function disorders, the best hope for a cure is genome surgery that repairs or removes the malfunctioning genes at the root of the disease. Currently, that hope lies in CRISPR/Cas9-based gene editing {DiCarlo, Mahajan & Tsang, J Clin Invest. 2018;128:2177}. The strength of the first-generation CRISPR-based therapy (CRISPR1.0; Fig. 1)?its mutation-specificity?is also its greatest weakness. This is because the therapeutic components for each mutation (both the guide RNA (gRNA) and the repair template) need to be custom-designed, engineered, tested, and FDA-approved. This presents a considerable and costly challenge for the many ad diseases caused by a slew of different mutations. For example, the blinding Best vitelliform macular dystrophy (VMD) disorder is caused by any 1 of 250 different mutations in the rhodopsin (BEST1) gene. Treatment of all patients would, therefore, require that 250 sets of CRISPR1.0 components be engineered, validated, and FDA- approved. To overcome this major limitation, we developed CRISPR2.0 (Fig.2), a mutation nonspecific strategy. Unfortunately, CRISPR2.0 is not allele-specific and so eliminates both the mutant and WT alleles. As a result, CRISPR2.0 requires gene supplementation, which leads to variable expression of the rescued gene and sustainability concerns. We now propose to develop a third-generation CRISPR-based strategy, CRISPR3.0 (Fig. 2), that, like CRISPR2.0, is mutation nonspecific. However, CRISPR3.0 is allele-specific and therefore ablates the mutant, disease-causing cis allele while leaving the WT allele intact to support normal function. We hypothesize that CRISPR3.0 chromosome-specific genome surgery will produce a more sustained therapeutic response compared to the CRISPR2.0 supplementation strategy.