USHER SYNDROME TYPE 3 A : RETINAL PHENOTYPE AND GENE THERAPY APPROACHES By RACHEL MICHELLE STUPAY A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2017
2017 Rachel Michelle Stupay
To my mother Cynthia Kachnik for her continued support through my many endeavors over the years and to my boyfriend Marcus Hooper for surviving the Ph.D process with me W ords cannot express my gratitude to you both
4 ACKNOWLEDGMENTS I thank my mentor Dr. William Hauswirth and all the members that have been a part of the Dr. Hauswirth lab over my time in the lab particularly my co mentor Dr. Astra Dinculescu for working with me on a daily basis and for training me with everything I have learned. I thank my committee members Dr. John Ash, Dr. Alfred Lewin, and Dr. Wesley Clay Smith for their continued support and assistance throughout my Ph.D and their effort in helping to train me as a scientist. I thank the Genetics Cha ir Dr. Margret Wallace for her continued guidance, mentoring, and training during this process and for allowing me opportunities that were not readily available for most graduate students. I thank my boyfriend Marcus Hooper for his continued support over the years, helping out with Loki m y beloved dog, and for helping to edit this dissertation. I finally thank my family, particularly my mother, Cynthia Kachnik, for her continued support over the y ears from almost 2 decades in ballet, to a very productive undergraduate career, and finally during an extensive Ph.D program.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 LIST OF ABBREVIATIONS ................................ ................................ ........................... 10 ABSTRACT ................................ ................................ ................................ ................... 13 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 15 Primary Cilia ................................ ................................ ................................ ........... 15 Cilia Structure and Function ................................ ................................ ............. 15 Sensory Cilia ................................ ................................ ................................ .... 19 Diseases Associated with Primary Cilia (Ciliopathies) ................................ ...... 22 The Biology of Vision ................................ ................................ .............................. 24 Retina l Structure ................................ ................................ ............................... 24 Phototransduction Cascade ................................ ................................ ............. 27 Retinal Function and Morphological Assays ................................ ........................... 28 Electroretinogram ................................ ................................ ............................. 28 Spectral D omain Optical Coherence Tomography ................................ ........... 29 Autosomal Recessive Retinal Dystrophies ................................ ............................. 30 Patient Phenotypes ................................ ................................ .......................... 30 Current Treatment Methods for Retinal Degeneration ................................ ...... 32 Usher Syndrome ................................ ................................ .............................. 40 Usher Interactome ................................ ................................ ............................ 45 Usher Syndrome Animal Models ................................ ................................ ...... 47 2 OVERVIEW, RATIONALE, AND SPECIFIC AIMS ................................ ................. 49 Aim 1: Identify endogenous Clarin 1/CLARIN 1 expression in the retina. ............... 50 Aim 2: Identify a retinal phenotype in N48K Knock in (KI) (N48K) and Clarin 1 Knock out (KO) ( Clrn1 / ) mice. ................................ ................................ ........... 50 Aim 3: Identify the optimal AAV gene construct, capsid serotype, and delivery method for Clarin 1 gene therapy treatment and assay for phenotypic rescue using AAV mediated gene therapy. ................................ ................................ ..... 51 3 IDENTIFYING WHERE ENDOGENOUS CLARIN 1 RNA AND PROTEIN IS EXPRESSED IN THE RETINA ................................ ................................ ............... 52 Background ................................ ................................ ................................ ............. 52
6 Methods ................................ ................................ ................................ .................. 58 RT PCR ................................ ................................ ................................ ............ 58 Immunohistochemistry ................................ ................................ ...................... 59 Western Blot Analysis ................................ ................................ ...................... 63 Results ................................ ................................ ................................ .................... 64 Clarin 1 mRNA Isoform Specific Expression ................................ .................... 64 Endogenous Clarin 1 Protein Localization Using Immunohistochemistry ......... 66 Endogenous Clarin 1 Protein Localization Using Western Blot ........................ 69 4 IDENTIFYING A RETINAL PHENOTYPE IN CLARIN 1 KNOCK OUT (KO) (Clrn1 / ) AND N48K KNOCK IN (KI) MICE* ................................ .......................... 80 Background ................................ ................................ ................................ ............. 80 Methods ................................ ................................ ................................ .................. 82 Arrestin 1 and Transducin Translocation Assay ................................ ............... 82 GFAP Expression and Olfaction Assays to Test for Alternative Phenotypes .... 84 Electroretinography ................................ ................................ .......................... 85 Spectral Domain Optical Coherence Tomography ................................ ........... 87 Light Damage ................................ ................................ ................................ ... 88 Results ................................ ................................ ................................ .................... 89 Light Driven Protein Translocation to Characterize a Retinal Phenotype ......... 89 GFAP Expression in WT vs KO Mice ................................ ............................... 90 Olfactory Structure and Clarin1 Expression in Olfact ory Cells .......................... 90 Validation of the Previously Published Novel ERG Phenotype ......................... 91 Abnormal Retained Electroretinogram in N48K Mice ................................ ....... 93 Dual Flash ERG Recovery Response ................................ .............................. 94 Light Damage in WT, KO, and KI Mice ................................ ............................. 94 5 ASSAYING FOR PHENOTYPIC RESCUE USING AAV MEDIATED GENE THERAPY* ................................ ................................ ................................ ........... 111 Background ................................ ................................ ................................ ........... 111 Adeno Associated virus ................................ ................................ .................. 111 Sero types and Modifications ................................ ................................ .......... 112 Methods ................................ ................................ ................................ ................ 114 Results ................................ ................................ ................................ .................. 116 AAV Delivered Clarin 1 Retinal Localization ................................ .................. 116 Optimized AAV Vector Capsid, Promoter, and Titer for Safe and Effective Gene Therapy ................................ ................................ ............................. 118 Gene Therapy Rescue of C larin1 Retinal Phenotypes ................................ ... 121 6 CONCLUSIONS ................................ ................................ ................................ ... 134 LIST OF REFERENCES ................................ ................................ ............................. 141 BIOGRAPH ICAL SKETCH ................................ ................................ .......................... 164
7 LIST OF TABLES Table page 3 1 Previously published Clarin 1 RT PCR primers ................................ .................. 71 4 1 Dual flash ERG recovery response ................................ ................................ .... 97 4 2 Two way ANOVA for WT light damage range of intensities ............................... 98 4 3 Two way ANOVA in Clarin 1 KO and N48K KI mice, 2000 lux light damage ..... 99 5 1 AAV constructs ................................ ................................ ................................ 123
8 LIST OF FIGURES Figure page 3 1 RT PCR and Clarin 1 isoform expression. ................................ ......................... 72 3 2 Novus Biologicals CLARIN 1 antibody st aining for immunohistochemistry ........ 73 3 3 Cryo preserved C57BL/6J Novus Biologicals CLARIN 1 staining ...................... 74 3 4 Novus Biologicals CLARIN 1 localization in WT vs Clarin 1 KO, and N48K KI A/J mice ................................ ................................ ................................ .............. 75 3 5 Novus Biologicals CLARIN 1 colocalization with gamma tubulin ........................ 76 3 6 Subretinal and intravitreal AAV injected CLARIN 1 colocalization with Novus Biologicals CLARIN 1 antibody ................................ ................................ .......... 77 3 7 Novus Biologicals CLARIN 1 western blot analysis ................................ ............ 78 3 8 Arrestin 1, rhodopsin, and secondary only control western blots ....................... 79 4 1 Normal localization of arrestin 1 and transducin ................................ ............... 100 4 2 Immunofluorescence and quantification of arrestin 1 in light adapted Clarin 1 KO and WT retinal sections ................................ ................................ .............. 101 4 3 GFAP expression in WT vs Clarin 1 KO C57BL/6J mixed albino mice ............. 102 4 4 Normal olfactory structure in both WT and N48K KI A/J mice .......................... 103 4 5 RT PCR of WT A/J olfactory epithelial tissue lysates ................................ ....... 104 4 6 ERG difference in C57BL/6J vs WT and Clarin 1 KO A/J mice ........................ 105 4 7 Validation of the ERG phenotype in A/J Clarin 1 KO mice ............................... 106 4 8 ERG phenotype post 1 hour ligh t exposure in the N48K KI mice ..................... 107 4 9 Dual flash ERG recovery response in the N48K KI mice ................................ .. 108 4 10 Light damage optimization in WT A/J mice ................................ ....................... 109 4 11 SD OCT and ERG in Clarin 1 KO vs N48K KI A/J mice pre vs post light damage ................................ ................................ ................................ ............ 110 5 1 Localization of vector expressed CLARIN 1 Venus following subretinal and intravitreal delivery ................................ ................................ ............................ 124
9 5 2 Toxicity and photoreceptor cell death post subretinal injection of CLARIN 1 ... 125 5 3 Photorecep tor cell death and retinal morphology in full titer vs 1:1000 dilution of subretinally delivered CLARIN 1 ................................ ................................ ... 126 5 4 Evaluation of re tinal function and retinal morphology in CLARIN 1 HA injected mice following subretinal delivery ................................ ........................ 127 5 5 ERG and Immunohistochemistry analysis of subretinal vs intravitreal AAV2 Quad Y F smCBA h CLARIN 1 HA in C57BL/6J mice 6 weeks post injection .. 128 5 6 SD OCT of subretinally injected AAV2 Quad Y F pTR GRK 1 h CLARIN 1 HA ................................ ................................ ................................ .................... 129 5 7 ERG analysis of subretinal vs intravitre al AAV2 Quand Y F pTR GRK1 h CLARIN 1 HA 3 months post injection ................................ ........................... 130 5 8 Evaluation of retinal function and GRK1 h CLARIN 1 HA expression following subretinal delivery ................................ ................................ ............................. 131 5 9 Arrestin 1 translocation quantification after intravitreal injection ....................... 132 5 10 Evaluation of retinal function in WT vs Clarin 1 / KO untreated and trea ted retinas following intravitreal delivery ................................ ................................ 133
10 LIST OF ABBREVIATIONS AAV Adeno Associated Virus ABR Auditory Brainstem Response Aipl1 Aryl Hydrocarbon R ecepto r Interacting Protein Like 1 ANOVA Analysis of Variance BBS Bardet Biedl Syndrome BBSome Bardet Biedl Syndrome Interactome bp Base Pairs BPC (s) Bipolar Cell CC Connecting Cilium of Photoreceptors Cdh23 Cadherin Protein number 23 cDNA Complimentary DNA Clrn1 / Clrn1 Clarin 1 Gene/Protein mouse CLRN1 /CLRN1 Clarin 1 Gene/Protein human CLS Ciliary Localization Signal DAPI 4',6 diamidino 2 phenylindole DNA Deoxyribonucleic Acid ER Endoplasmic Reticulum ERG Electroretinogram GC1 G uanylate C yclase 1 GCL Ganglion Cell Layer GFAP Glial Filamental Acidic Protein GFP Green Fluorescent Protein GPCR G Protein Coupled Receptor GRK1 G Protein Coupled Receptor Kinase 1 Promoter
11 HA Hemagglutinin Sequence Tag IFT Intraflagellar Transport ILM Inner Limiting Membrane INL Inner Nuclear Layer IPL Inner Plexiform Layer IRBPE Interphotoreceptor Retinoid Binding Protein Enhancer IS Inner Segment of Photoreceptors (PR) N48K KI Clarin 1 N48K Knock in mouse Clrn1 / KO Clarin 1 Knock out mouse ml Milliliters MKS Meckel Grouber Syndrome, including protein complex NB Novus Biologicals NB CLRN1 Novus Biologicals CLARIN1 antibody NPHP Nephronopthesis, including protein complex OCT Optical Coherence Tomography OLM Outer Limiting Membrane ONL Outer Nuclear Layer OPs Oscillatory Potentials OPL Outer Plexiform Layer OS Outer Segment of Photoreceptors (PR) PBS Phosphate Buffered Saline PC Primary Cilia PFA Paraformaldehyde Pch15 Protocadherin number 15 PCR Polymerase Chain Reaction
12 PDZ Post Synaptic Density Protein (PSD95), Drosophila Disc Large Tumor Suppressor (Dlg1), and Zonula Occludens 1 Protein (zo 1) Protein Sequence PR (s) Photoreceptor ( s ) RGC (s) Retinal Ganglion Cells RNA Ribonucleic Acid RPE Retinal Pigmented Epithelium RP E65 Retinal Pigmented Epithelium Specific Protein,65kDa Molecular Weight RT PCR Reverse Transcriptase Polymerase Chain Reaction sc Self Complimentary scAAV2Q Y F scAAV2 Quadruple Capsid Protein Mutant ( Y272F+Y444F+Y500F+Y730F) scAAV2Q uad T491V scAAV2 Quadruple Capsid Protein Mutant ( Y272F+Y444F+Y500F+Y730F) plus T491V Mutation SDS PAGE Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis SEM Standard Error of the Mean smCBA Small Chicken Beta Actin Promoter SNAREs soluble N ethylmaleimide sensitive factor receptors Spata7 S permatogenesis A ssociated P rotein 7 TZ Transition Zone ul Microliters USH Usher Syndrome USH3A Usher Syndrome Type 3A UTR Untranslated Region vg Vector Genomes
13 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy USHER SYNDROME TYPE 3 A : RETINAL PHENOTYPE AND GENE THERAPY APPROACHES By Rachel Michelle Stupay December 2017 Chair: William W. Hauswirth Co chair : Astra Dinculescu Major: Medical Science Genetics Primary cilia (PC) are extracellular microtubule based sensory organelles which can be found on almost every cell type in the mammalian sys tem. Defects in PC result in a broad range of phenotypes including polycystic kidneys, retinal degenerati on, hearing loss obesity, polydactyly, an d severe cerebral malformations As a class these diseases are termed ciliopathies While some mutations are more pleiotropic other mutations only affect certain cell types such as the human sensory organ systems including rod and cone photoreceptors in the retina and cochlea hair cells in the inner ear which have highly modified PC For my project I have chosen to work on a s elect class of ciliopath ies termed Usher syndrome (USH) USH is an autosomal recessive disorder that affects both hearing and vision in patients and accounts for more than 50% of combined deafness and blindness cases. Usher Syndrome Type 3A ( USH3A ) is caused by mutations in the Clarin 1 ( CLRN 1 /CLRN1 ) gene and t here is no cure for this d evastating disorder. Therefore, my major goal wa s to develop a gene replacement based therapeutic approach for USH3A. In order to accomplish this, I first needed to i dentify the endogenous localization of Clrn 1 in the retina by utilizing commercial ly
14 available antibodies against CLRN1 allowing us to target the correct retinal cell type for treatment and to avoid off target toxicity In addition, I sought to identify a retinal phenotype in Clrn1 deficient mouse models in order to design and evaluate an effective gene therapy approach and be able to efficiently assay for phenotype rescue M y results show that both Clrn1 / knockout ( KO ) and N48K knock in (KI) mice display a novel retinal phenotype with a delay in arrestin 1 movement upon exposure to light in photoreceptor cells M y final goal was t o determine the effects of AAV mediated delivery of Clrn1 on retinal function and morphology in USH3A mouse models. I sh ow that all neuronal cell types in the retina sustain AAV mediated CLRN1 expression after intravitreal and subretinal vector delivery as detected by anti GFP and anti HA tag antibodies In addition, the previously utilized CLRN1 antibody can recognize AAV expressed CLRN1 through co localization experiments
15 CHAPTER 1 INTRODUCTION Primary Cilia Cilia Structure and Function P rimary cilia (PC) was first identified in 1675 by Anton Van Leeuwenhoek and were believed to be vestigial organelle s not essential for cell survival and/or function (41) Today it is know n that PC ar e found on almost every cell type in the body and are conserved from the green algae Chlamydomonas reinhardtii to the nematodes Caenorhabditis elegans all the way up t o humans (85) PC have a diverse range of physiological roles including embryonic development, control of cell growth, and signal transduction (45, 56, 228) T issue specific modified PC also act as chemical, physical, or ligand based sensors which will be discus sed in Sensory Cilia Overall, PC play a vital role in the development and function of a wide range cell types and organ systems. PC are microtubule based organelles that extend from the apical surface of the plasma membrane and are present on almost all eukaryotic cell types (18, 41, 64, 85, 113, 195) During the cell cycle, in mitosis the centrioles act to separate a single cell into two daughter cells. O nce the nuclear envelope re forms the mother centriole duplicates a neighboring daughter centriole. Once established, the mother and daughter centrioles proceed to dock along the plasm a membrane in G0/G1 phase and become the basal bodies that generate the ciliary axoneme (41, 8 5, 113) The mother centriole contact s a ciliary vesicle, analogous to the plasma membrane, which tether s the mother centrio le to a bilayer membrane by centriole distal appendages called transition fibers. I n the cil iary vesicle, a ciliary bud begins to form and will elongate from the base to the tip, formi ng the cilia membrane The mother centriole will elongate the alpha tubulin a nd
16 beta tubulin to generate the ciliary axoneme, while the gamma tubulin will remain exclusively at the mother centriole. Once the cilium is fully formed, the ciliary vesicle will dock and merge with the plasma membrane and the ciliary axoneme will be proj ected outside the cell. (85, 113, 195, 228) Once a cilium is formed, the cell is no longer able to divide until the cell resorbs the ciliary axoneme and disassociates the basal bodies from the plasma me mbrane to then act as centrioles for cell division once again. PC exist as motile and non motile PC All have a ring of 9 parallel microtubule doublets that extend from the basal bodies at the cell surface. The basal bodies consist of alpha beta and gamma tubulin while the PC axoneme consists of alpha and beta tubulin (18, 41, 64, 85, 113) Th e outer ring microtubule doublets are also susceptible to post translational modifications including acetylation, glyc y lation, and glutamylation (113) These additions allo w for greater stability and provid e additional interaction foci for other proteins. Motile cil ia contain an additional microtubule doublet in the middle of the axoneme that is tethered to the outer ring by radial spokes and dynein arms that allow for regul ated movement a nd active bending of the axoneme (56, 143, 195) Motile cilia are preferentially found on the surface of epithelia l cells in the trac hea and the ependymal cells lining the ventricles in the brain. Unique 9 + 2 microtubules exist on olfactory sensory neurons that coincidentally are non motile cilia There is an additional ro le for non motile cilia on the embryonic node pericardium on the developing heart which helps to r egulate fluid flow in the abdomen during embryonic development and is respons ible for the specification of th e left right body axis (18, 85, 113) The base of the ciliary membrane is continuous with the plasma membrane and a distinct region, just abov e the basal bodies will become differentiated into a transition
17 zone (TZ) with structurally dis tinct protein complexes that act as a gate to regulate protein entry and exit within the cilium. The TZ consists of Y links connectin g the ciliary membrane to the microtubule axoneme They are tethered to the membrane through transmembrane proteins that associate with beaded extracellular proteins that form an outer ring terme d a ciliary necklace (45, 195, 216, 231) At the TZ, there are two pri mary protein complexes that functionally interact with each other, the Meckel Grouber Syndrome ( MKS ) complex and the Nephronopthesis ( NPHP ) complex. Interactions between these two complexes are critical for the early stages of ciliogenesis including membra ne docking and fusion of the ciliary vesicle (56, 195, 216, 231) The MKS complex consists of MKS 1/ MKSR 1/ MKSR 2/ MKS 3/ / MKS 5/ MKS 6 and the NPHP complex consists of NP HP 1/NPHP 4. Fu rthermore, there is a hierarchal interaction between the proteins within each individual complex, as well as between the two complexes. MKS 5 is essential for TZ localization of all MKS and NPHP complex components. Loss of MKS 5 in combination with MKSR 1, MKSR 2, MKS 6, and NPHP 4 all result in abnormal cilia morphology; and MKS 5 is required for proper TZ localization of MKS 1, MKSR 1, MKSR 2, MKS 3, and MKS 6 (the entire MKS complex). Additionally, MKS 5 is also responsible for complete localization of N PHP 1 and NPHP 4 at the transition zone (231) Together, this indicates that MKS 5 is a critical component of the TZ and may act as a key regulator of the ciliary gating mechanism. In addition to the TZ as a ciliary gate, protein trafficking into and out of the cilia is also regulated by intracellular trafficking. Due to the fact that protein synthesis takes place within the en doplasmic reticulum (ER) and cytoplasm of the cell, ciliary proteins need to be trafficked to the PC through intraflagellar transport (IFT) as well as the
18 BBSome (BBS: Bardet Biedl Syndrome) complex. IFT is an intracellular trafficking system that moves uni directionally along microtubules using Dynein and Kinesin microtubule motors. Kinesin2 regulates anterograde transport and Dynein regulates retrograde transport. These motors a re associated with two primary protein complexes comprised of IFTA and IFTB (85, 113, 202, 203, 234) The BBSome is a hetero octameric complex of 8 core proteins with up to 8 associated proteins that traffics cargo bi directionally through the cytoplasm and cilia The B BSome is not essential f or PC formation as IFT is; however, mutations in BBS proteins result in alter ed membrane protein composition (122, 159) One mechanism tha t helps to regulate the selective trafficking of certain receptors to cilia is that they contain a ciliary localization sequence (CLS) within the protein that allows it to be recognized by the IFT and/or BBSome machinery (85, 216) Membrane trafficking from the ER and/or Golgi apparatus is also regulated by the Rab family of small GTPases as well as SNAREs (soluble N ethylmaleimide sensitive factor receptors) (113, 216, 228) SNAREs are a component of the exocyst complex this will allow trafficking across the ciliary membrane indicating that there is a potential active transport mechanism in place utilizing importins and exportins It has been proposed that this process functions similarly to that of the nuclear pore complex and nucl e oporins that regulates protein trafficking across the nuclear envelope. Indeed, recently, it has also been shown that nucleoporins l ocal ize to the basal bodies at the PC in cell culture as well as in tracheal epithelial cells (85, 130, 161) Not all nucleoporins localize to the basal bodie s only those aligning with the outer nuclear membrane ring, transm e mbrane ring, and linker nucleoporins indicating that there are cilia specific gating proteins as well in addition to the TZ proteins (130)
19 Sensory Cilia In addition to the standard PC found on almost every cell type, certain tissues possess modified PC unique to it s particular cell type, tissue type and function. The primary examples are olefacto ry cilia in nasal tissue, stereocilia and kinocilia in the cochlea, and photoreceptor (PR) connecting cilia (CC) in the retina One common feature of all three cell types is that their PC have adjacent actin based microvilli, either on the same c ell or adj acent cells, which act to assist in cilia function (80) Olfactory neurons are first order neurons and are the initial cell types in the olfactory system that responds to external stimuli primarily odorants and are responsible for our sense of smell Olfactory neurons also have adjacent sustentacular cells that act as supporting cells and they possess a dense layer of mic rovilli on their apical surface (20) The olf a ctory syste m is unique in that the olfacto ry epithelium has a basal cell layer con sisting of ho rizontal and globular basal cells that acts as olfactory stem cells and can regenerate the olfactory neurons if they die and need to be replaced (20) Olfac tory neurons contain between 10 30 primary cilia per cell, and as mentioned previously, o l factory neuronal cilia are non motile cilia even thoug h they possess a central doublet of microtubules (80, 120, 121, 125, 151) Odorants will bind to their respective G protein coupled receptors (GPCRs) w ithin the cilia of olfactory neurons and will then consecutively relay the activation signal to the olfactory glomeruli in th e olfactory bulb which will stimulate mitral cells in the olfactory tract t o relay the signal to the brain (156) In general, each olfactory neuron only expresses a single type of GPCR, and humans possess approximately 4 00 odorant GPCR genes. Although each GPCR is specific for a particular odorant, they all signal through the activation of
20 adenylyl cyclase III which will act to increase the intracellular levels of cAMP and the consecutive opening of cyclic nucleotide gat ed channels (80, 121, 125, 151) In contrast t o the multi ciliated ol factory neurons, ciliated cells in the coch lea have a single PC with adjacent actin based microvilli structures. PC in the cochlea act to receive and process sound waves and the adjacent vestibular system coincides with sensing motion and balance (21, 111) At the surface of each cochlear hair cell th ere is a modified PC called a kinoci lium which contains a central microtubule doublet in addition to the 9 doublets o f the axoneme Each kinocilia has adjacent rows of stereocilia (similar to microvilli) that are composed of actin arranged in semi circle rows and are stacked in a staircase like fashion moving away from the kinocilium (80, 84) The result is a V shaped bundle that is connected together by both ankle and tip links that, upon sound stimuli, will bend and initiate an intracellular signaling cascade to relay the stimulus through the cochlear nerve and into the brain to process the signa l (80, 111) T he direction of kinocilium formation will predict the direction and orientation o f the stereocilia bundle and is believed to be responsible for any positional information required for proper hair cell development (3, 10, 21, 54, 80, 111, 157, 242, 246) In contrast to both ol factory and cochlear cilia, the cilia in PR cells in the retina exist as a single high ly modified PC called a CC ( (15, 67, 78) 133, 170). Rod PR are designed for response to very dim levels of light and cone PR are designed for response to bright levels of light. They each have a similar cellular structure in that their ribbon synapses are located in the outer plexiform layer (OPL) in the retina, their nuclei are localized in the outer nuclear layer (ONL), their inner segments (IS) contain the majority of intracellular content, and they have a small CC that connects the IS to the
21 outer segment (OS). Just below the CC, ar ound the basal bodies, is a periciliary ridge complex that helps to facilitate disk membrane biogenesis as well as general membrane protein trafficking into the cilium (112, 145, 161, 214) The re are additional actin based microvilli like structures that are extensions of IS membrane, and these filamentous structures are termed calyceal processes which are only present in amphibians, birds, and primates (133, 168) Along the OS of PR cells are mem brane invaginations termed incisures that are associated with these calyceal processes (72, 73). This periciliary membrane as well as the calyceal processes are highly concentrated with scaffolding and adhesion molecu les that are very large transmembrane proteins. These proteins act to connect the OS incisures to the calyceal processes and allow for structural integrity of the OS as well as potentially help to facilitate protein trafficking into the OS of PR cells (72, 73, 200). These proteins when mutated or missing result in Usher Syndrome (USH) wh ich will be discussed below in Usher Syndrome (147, 233) The rod OS is composed of very densely stacked disc membranes that are separate from the plasma membra ne whereas the cone OS is composed of folds of plasma membrane (80, 106) PR cilia are in a 9 + 0 arrangement and the TZ occupies the entire length of the CC, which is unique to PR cells (112, 145 ) The OS contains the axoneme of the cilia and it also contains the light absorbing components of the PR. OS discs are shed distally on a daily basis and are phagocytized by the retinal pigmented epithelium (RPE) while new discs are replenished at the base of the OS at the CC. All components of the OS are synthesized in the IS and they are trafficked to the OS through molecular motors within the cell (147, 187, 190, 202, 203, 234) Through some more recent data, i t was identified that primates have well developed additional actin
22 based structures on their PR cells that extend from the IS and a re localized all around the OS termed calyceal processes T hey are present on both rod and cone PR cells and are absent from lower level mammals including mice These processes are composed of actin filaments and contain all of the USH proteins indicating that these processes are involved in an elongated pe ricil iary ridge complex and assist with protein trafficking to the OS (198) The OS of rod PR s have membrane invagin ations termed incisures which are associated with microtubule tracks that allow for protein trafficking along the OS. C alyceal processes are tethered to the OS incis ure microtubule tracks by the USH transmembrane proteins. They are believed to help facilit ate OS protein trafficking as well as possibly provide additional structural integrity to the rod OS (75, 76) This is believed to be why USH syndrome mouse models do not display retinal degeneration, whereas patients that carry similar mutations develop retinal degeneration (198) and will be discussed further in Autosomal Recessive Retinal Dystrophies. PR function and the rhodopsin signaling cascade will be discussed in Phototransduction Cascade Diseases Associated with Primary Cilia (Ciliopathies) Ciliopathies are a class of disorders that result from defects and/o r abnormalities in PC structure and/or function. Ciliopathies are primarily autosomal recessive disorders that present with a very broad range of phenotypes even with mutations in the same gene. Autosomal recessive polycystic kidney disease is defined by bilateral renal cysts and is caused by mutations in the PKHD1 gene. Nephronopthesis (NPHP) is caused by renal cysts and 11 causative genes have been identified, NPHP1 11 (18, 64, 78, 109) an early onset blindness due to retinal degeneration that prese nts as retinitis pigmentosa caused by PR cell death T here are 14 known causative genes. Not all of the se are cilia related genes, as there are several that are
23 crucial for the visual signaling cascade, but the disease is characterized based on patient presentation (1, 187, 188) Senior Lken Syndrome is combined polycystic kidney disease wi th retinal degeneration and is believed to be a result in mutations in select NPHP genes (1, 109 187, 188) Usher S yndrome (USH) is combined hearing and vision loss due to mutations in any of 11 genes. The three forms of USH are classified by the severity and age of o nset of symptoms and USH will be discussed further in Usher Syndrome (1, 187, 188, 230) in addition to renal cysts also presents with retinal coloboma as well as mental retardation and ataxia as a result of hypoplasia of the cerebellum. Mutations in several NPHP and MKS genes can imary diagnostic ch s as a result of cerebral hypoplasia (41, 109) Bardet Biedl Syndrome (BBS) is caused by mutations in the BBSome proteins as well as some MKS and NPHP genes. BBS is characterized by juvenile onset obesity caused by altered leptin signaling in the brain. BBS also presents with retinitis pigmentosa polycystic kidneys, and mental retardation (1, 41, 109, 188, 230) Meckel Grouber Syndrome (MKS) is the most severe form of ciliopathy and is embryonic lethal. In addition to the earlier mentioned symptoms it is primarily characterized by lung hypoplasia, post axial polydactyly, and occipital meningoencephalocele (41, 109) R ecently, different recessive mutations h ave been seen in many cilia genes and can present with a wide range of organ involvement based on the causative mutation or combination of mutations (109, 127) There is also the possibility that certain mutations will present differen tly based on the genetic background of the patient and the effect of modif ier alleles on the causative mu tation(s) (127) This is becoming more widely
24 accepted in the field of ciliopathy research over the last deca de as researchers begin to identify the cellular mechanisms underlying the respective disease phenotypes (127) The Biology of Vision Retinal Structure The eye is a highly specialized organ that allows for the perception of light and v ision. The visual system is a highly complex signaling system with multiple integrated components that are required for visual processing. Light pass es through the cornea and lens and is received i n the retina in the back of the eye to sense a visual stimu l us This stimulus is further processed by the inner ret ina and then by the brain into cohesive images. The retina possesses five primary classes of neurons that perform the initial image processing includ ing rod and cone PR s bipolar cells (BPC s ) amacrine cells, horizontal cells, and retinal ganglion cells (RGC s ) In addition to retinal neurons there are two supp orting cells in the retina the RPE and Mller glial cells (106, 214) Rod and cone PR s are first order neurons in the ONL that initially respond to light Rod PR s mediate dim light vision while cone PR s mediate bright light and color vision. The r od scotopic response range is 10 6 10 2 cd/m 2 while the cone photopic respons e range is 10 2 10 6 cd/m 2 M esopic vision refers to a mixed rod cone response and is at the high response end of rods and the low end of cones between 10 2 10 cd/m 2 Humans have one type of rod cell that responds maximally to a waveleng th of 498 nm and 3 cone pigment cells blue cones (short wav elength cones: s cones) respond ing maximally to a wavelength of 437nm, green cones (medium wavelength cones: m cones) responding maximally to a wavelength of 533 nm, and red cones (long wavelength cones: l cones) respon d ing maximally to a wavelength of 564 nm (106)
25 At the base of PR cells are their synap ses that connect to BPCs Rod synapses are termed the rod spherule and cone synapses are termed the cone pedicle. Both rod and cone synapses conta in a ribbon syna pse that is adjacent to post synaptic structures (106) In addition to synaptic communication between PR s and BPC s there are also lateral contacts between rod s and cones as well a s cones to cones in the retina. These exist between the synaptic membranes as well as small gap junctions along the outer limiting membrane ( OLM ) that also connect both rod and cone PR s to the Mller cell end feet (106) BPC s are second order neurons in the INL that synapse to PR s and act to amplify the visual signal There is one type of BPC for rods, the rod on BPC which connects to multiple rod cells. Cones have two types of BPC s on and off BPCs. These BPCs relay the amplified signal to the third or der RGC s which communicate the visual signal through the optic nerve to the brain. Within the brain, communication and i nterpretation between the le ft and right eyes occurs across the optic chiasm. Finally, t he lateral geniculate nucleus in the thal amus is responsible for visual perceptio n in the brain and this signal is forwarded to the visual cortex for final visual processing (214) In addition to bipol ar cells, the INL has h orizontal cells and amacrine cells that provide late ral interactions between rod and cone cells a s well as between PRs across the retina. Horizontal cells are responsible for close and global communication across the retina between rods and cones A macrine cells have s ynaptic connections with the axons of bipolar cells and RGCs (106) O n the apical side of the OS of PR cells lies the RPE cell layer. The RPE absorbs much of the excess light tha t is not received by the PR cells and provides nutrients and oxygen to PR s The RPE is also responsible for phagocytosing the OS of PR s on a daily basis. Approximately 10% of the distal OS is
26 digested by the RPE every morning and this is an integrin media ted process (106) Mller cells are another cell type found in the retina that span the entire retina from their basal lamina at the inner limiting membrane (ILM) adjacent to the RGCs with the nu cleus in the INL, and the Mller cell end feet creating the OLM at the apical side of the ONL. Mller cells act as supporting glial cells for all cell types in the retina and are responsible for neuroprotective pathways as well as facilitat ing cone opsin r ecycling through a Mller cell specific pathway. Additionally, Mller cells are believed to act as photon guides to direct incoming light direct ly onto the PR OS (106) H uman retinas have a highly dense cone region near the center of the retina along the visual axis called a fovea centralis which is approximately 2.5 mm in diameter and appears as a shallow depression in retinal optical coherence tomography images Surrounding the fovea is a narrow ring called the macula lutea. T hese areas are adjacent to the optic nerve, which is localized nasally to the fovea. The fovea consists of only cone PR s with a lower density of cones spread across the remainder of the retina. Adjacent to the fovea, approxi mately 4.5 mm just outside of the foveal pit, is the densest region of rod PR s that gradually decrease s towards the retinal periphery There are between 60 125 million rod PR s and approximately 3.2 6.5 million cone s in the human retina Cone s account for o nly 10% of the retina but they are responsible for our daytime vision and normal daylight visual acuity (106) The foveal co ne system allows for greater resolution than the rest of the retina in that each cone PR synapses to only one bipolar cell termed a midget cone BPC, and one ganglion cell termed a midget ganglion cell. I n the periphery of the retina, each rod BPC connect s with multiple PR s and each cone BPC and ganglion cell interact with multiple PR s (106, 214) Th e fovea l
27 region is specific to primates, however some species have modified regions of the retina that act in a similar fashion for example the area centralis in dogs (16) Phototransduction Cascade In rod PR cells rhodopsin is the primary compo nent of the OS. It is a 40 kDa seven transmembrane GPCR with an associated chromophore molecule 11 cis retinal. The C terminal sequence of rhodopsin is critical because it n ot only contains a sequence of five amino acids (aa) (QVXPX ) at t he C terminus that target s rhodopsin to the OS, but the C terminus and intracellular loop regions are also responsible for binding of r ho dopsin to i ts associated G protein complex (161, 215) In the dark the cyclic nucleotide gated channel is open with an intracellular charge of approximately 40mV and there is a continuous glutamate neurotransmitter release at the PR ribbon synapse (106) Upon exposure to light, the 11 cis retinal undergoes photoisomerization to the all trans confirmation Activated rhodopsin is now able to bind and activate the G protein transducin by catalyzing a GDP exchange for transduci ng subunit of the G protein membrane associated complex which is then release d from the co mplex (206) Once activated, t ra nsducin stimulate s the activation of cGMP phosphodiesterase that then hydrolyzes cGMP to GMP. This results in the closin g of the cyclic nucleotide gated channel s in the membrane and causes a hyperpolarization in membrane potential to approximately 70 mV thus inhibiting the inflow of Na + and Ca 2+ ions into rods (106, 214) For the recovery phase, activated r ho dopsin is phosphorylated by the G protein r hodopsin kinase (GRK1) which then allows for arrestin 1 to bind the C terminus and cytoplasmic loops of r ho dopsin In the dark with high levels of intracellular Ca 2+ a recoverin Ca 2+ c omplex binds to GRK1 to reduce GRK1 activity and inhibit r ho dopsin phosphorylation. Upon exposure to light when the cyclic nucleotide
28 gated channels clos e, the levels of intracellular C a 2+ dr op and therefore results in the disassociation of the recoverin Ca 2+ GRK1 complex, allowing GRK1 t o then phosphorylate the C terminus of r ho dopsin and allow arrestin 1 bind ing Guanylyl cyclase activating proteins (GCAP) will replenish cGMP to lev els suffi cient to open cyclic nucleotide gated channels once more. Additionally the isomerized all trans retinal is re duced to all trans retinol, and is actively transported to the RPE converted back to 11 cis retinal and returned to the PR OS through the interstitial retinoid binding protein 11 cis retinal is then inserted back into r ho dopsin Upon return to the dark adapted state arrestin 1 dissociates from r ho dopsin and return s to the IS ONL, and OPL thus completing the phototransduction cycle (106, 112, 214) Retinal Function and Morphological Assays Electroretinogram The Electroretinogram (ERG) is a standard ophthalmologic al assay used to measure retinal function through the sti mulation of an electrical activity across the eye upon exposure to flashes of light. Upon illumination there is a change in electrical potential on the surface of th e cornea relative to the retina (106) The ERG was first two electrical changes that occurred first the cornea was negatively charged compare d to the baseline, and th en it became positively charged (106) Today the ERG is defined as a subset of 3 functions The a wave is an initial depolarization due to the electrical potential change in the OS as a result of activated rhodopsin The b wave is a positive change in electrical potential as a result of the INL cells receiving signal input from PR cells. Additionally, there is a c wave that appears later as a result RPE cel l responses (106, 175, 184) Accompanying the rising phase of the b wave are a series of oscillatory
29 potentials (OPs) that are believed to originate from via comm unication between cells in the INL in response to the ERG. OP s are visible but they cannot accurately be measured until a digital filter to amplify the signal is applied call ed a Fast Fourier Transform that reduces out the signal from the a wave and b wave in order to obtain a quantitative measure for their frequency and amplitude (106, 227) Additionally, prior to the a wave there is an early receptor potential signal that is a small dip prior to the full a wave and is due to the initial ch romophore activation (106) There are two types of ERG, scotopic and the photopic. The scotopic ERG is measured in response to dim light flashes and measures the response from rod PR s The photopic ERG is measured at bright light flashes and therefore measures cone function. As mentioned previously, there is some overlap of rod and cone response under intermediate flash intensities, termed mesopic conditions. T herefore the brightest light flash measured in the scotopic ERG will contain some cone r esponses as well Similarly, the dimmest light flash measured in the photopic ERG will contain some rod function Spectral Domain Optical Coherence Tomography Spectral Domain O ptical Coherence Tomography (SD OCT) p rovides a method for imaging cross sections of the retina in a live animal and is non invasive. It is the only imaging method that can provide a visual ization of the multiple layers of the retina without sacrificing the animal. SD OCT generates high quality images that are generated through two scans, an A scan and a B scan. The A scan registers the interface between the retinal tissue and the B scan is the ref lected output of the light scattering through the retina that is then registered by the camera (110) Our current OCT system is a Bioptigen instrument This system provides images of horizontal sections of the retina as well as a fundus image of the surface of the retina. SD OCT
30 scans can be used to measure retinal layer thicknes s in an animal over time. D elineations of the ganglion cell layer (GCL) INL and ONL are fairly easy to deduce however the demarcation of the IS/OS/RPE is much more difficult. This can be overcome by averaging multiple scans taken at the same slice in th e retina and merging the se images (19, 81, 238) SD OCT is currently the best tool for assessing subretinal vector injection damage because any subretinal de tachment can be observed through this methodology i n vivo OCT can be used in combination with histological sections to assay for retinal damage and subretinal detachment over time and was utilized to assay for post injection damage and retinal degeneration (19) Autosomal Recessive Retin al Dystrophies Patient Phenotypes A pproximately 2 million patients worldwide are affected b y retinal dystrophies (1, 92) Retinal dystrophies are inherited one of four ways: autosomal dominant, autosomal recessive, X linked or mitochondrial ly (1, 92) Autosomal r ecessive retinal diseases are Mendelian recessive inherited disorders caused by mutations leading to the lack of protein function that will result in cellular dysfunction and eventually cell death which in this case will present phenotypically as vision loss. The different forms of retinal dystrophies are due to mutations in over 100 different genes (92) The patient phenotypes observed in retinopathies can be classified based on the primary cell type(s) involved. Pure rod and cone dystrophies present as congenital stationary night blindness and achromatopsia, respectively. There can also be presenta tion of pathogenesi s first in rods and then cones a hallm ark of r etinitis p igmentosa R etinitis pigmentosa typically presents as night blindness with restricted peripheral vision followed by cone loss and a decline in visual acuity. C one/rod diseases occu r when the
31 cones are more affected initially Patients will exhibit photophobia and poor visual acuity with the possibility of color blindness For this project I will specifically focus on autosomal recessi ve retinitis pigmentosa. Autosomal recessive retinitis pigmentosa onset can vary depending on the gene mutated and the age at initial vision loss rang ing from infants to mid later T he initial presentation is typically night vision defects due to a loss of rod PR function with increasing loss of peripheral vision over time. Subsequent ex t ensive loss of rod s will result in tunnel vision Subsequent cone leads to declining visual acuity, color vision defects and eventually in some cases almost complete vision loss (1, 158, 199, 213, 225, 236) Utilizing imaging of the retinal tissue through a fundus exam, there is typically visible dark pigmented deposits due to the loss of PR cells, optic nerve pallor and thinning of centra l blo od vessels (199, 213) Additional phenotypic assays that can help identify the retin al pathology include the ERG which would present as a gradual decrease in the rod response and SD OCT which would present as a thinning of the ONL over time beginning in the periphery and gradually moving towards the fovea and might include retinal fluid a ccumulation (158, 213) Once the overt pathology of the patient is defined, the next step to understanding potential treatment therapies is identifying the genetic mutation(s ) causing the disease. R ecent advances in whole genome sequencing along with more thorough databases of characterized retinal mutations has allowed for more accurate gene specific treatments (11, 24, 25, 43, 44, 55, 70, 92, 118, 123, 124, 132, 150, 171, 172, 185, 186, 188, 200, 201, 212, 213, 221, 239) This also provides the patients with a somewhat more accurate disease prognosis and p attern for disease progression and allow s for more appropriate patient selection
32 for early phase gene therapy trials (20, 29, 66, 96, 104, 108, 119, 181) Unfortunately, even if the causative mutations are known, potential available treatments are often limited due to the lack of animal models that accurately recapitulate the disease phenotype, the lack of therapeutic outcomes in animal models in ability to qualify for a given clinical trial. There is also a concern in patients t hat pres ent with very early autosomal recessive retinitis pigmentosa because, although some of the more successful treatment trials have employe d gene replacement therapies, if PR cells are lost, there is no way currently to generate more in the retina (49, 50, 58, 8 6, 116, 118) Current Treatment Methods for Retinal Degeneration C urrent treatment modalities for retinitis pigmentosa are dependent on the age of onset and the severity of the retinal degeneration. In the most severe forms of retinitis pigmentosa patients have lost the majority of their vision and only ma intain a small fraction if any, of their PR cells in their central retina In these patients, INL cells may also b e damaged or lost but the primary degeneration is in PR cells. These are the most difficult to treat because there are few or no PR cells le ft to transduce as PR are terminally differentiated cells that do not divide, and there are no natural retinal progenitor cells tha t can give rise to new PR cells (136) The Argus II, the first commercially available prosthesis for retinitis pigmentosa is a 6x10 grid electrode array that is surgically implanted on the surface of the retina adjacen t to the GCL on the ILM with a receiving antenna that is sutured to the temporal region of the face. The external component is a transmitting antenna attached to a pair of glasses with a small video camera that will record the external environment. There is an additional vid eo unit that will process the images received through the video camera
33 and trans form them into an electrical signal that is sent through the antenna to initiate a focal electrical stimulation of small subsets of RGCs This results in a vis ion like signal being transmitted to the brain. The Argus II clinical trial was approved by the FDA in 2013 and patients have been monitored over the course of the trial for improvement in visual task performance, SD OCT, and a Functional Low Vision O bserver Rated Assessm ent Afte r 5 years of performance and safety validation, the patients presented with an overall impr ovement in quality of life. They were able to perform everyday tasks such as sidewalk tracking and direction d iscrimination statistically better when the sy stem is turned on vs off. The Argus II is now the sole retinal prosthesis that is approve d by the US, Canada, and Europe (58, 59, 74, 91, 174) An alternative approach is the use of optogenetics to stimulate normally light insensitive retinal cells to respond to a light stimulus, for example turning the bipolar cells and/or RGC s into new light sensors to generate a vis ual response to the environment (50) Currently, optogenetics therapies have made advances in pre clinical studies and in August of 2016 RetroSense therapeutics enrolled 3 patients in an FDA approv ed Phase I/II clinical trial funded by the Foundation for Fighting Blindness. The t reatment wa s applied as an AAV mediated gene therapy using channelrhodopsins ta rget ing RGCs in order to stimulate retinal activity (ClinicalTrials.gov, NCT02556736). The mos t novel approach that has recently been studied is the us e of stem cells to regenerate retinal cells In vivo through the injection of embryonic stem cells (ESC) and induced pluripotent stem cells (iPSC) Previously, there has been some progress in generati ng successful gene therapy trials for diseases of the RPE including macular degeneration and Stargardt macular dystrophy through transplanted ESCs and iPSCs
34 to the subretinal space between PR and the RPE (20, 86) There has also been progress made on generating large quantities of mammalian cochlea cell s in culture in order to generate sufficient cells to administer for cochlear iPSC gene therapy (155) There still needs to be extensive work done to assess whether these cells can integrate accurately and efficiently into degenerated retina s and cochleas. If these organ oids are able to recapitulate both cell types i n vivo this would be a p romising treatment not only for general forms of retinitis pigmentosa, but also for the more severe forms of disease that occur at varying ages from infan cy to adulthood, such as USH, since the age of onset will no longer be an issue, and because this would help to circumvent problems with gene the rapy treatments for large genes. In patients that present with later onset retinitis pigmentosa or slowly progr essive retinitis pigmentosa there are more available treatment options especially if regions of the retina are relatively intact. G ene therapy is simply defined as the administration of a therapeutic DNA to cure a defective cell. By restoring the wild type (WT) gene or cDNA there is the potential to achieve a therapeutic amount of WT protein that was previously absent from the cell. Alternatively the delivered gen e of interest may act to silence a det rimental gain of function gene. S ome of the most hopeful progress in gene therapy has been achieved for retinal diseases due to the fact that the eye is a relatively immune privileged organ that is isolated from other tissue systems and therapeutic treatment can be administered fairly easily through an intraocular injection There are two primary gene delivery methods for therapeutic treatments, nanoparticles that are bound to the DNA of interest and viral delivery vect ors that will encapsulate the DNA of interest. T hese two methods can further be modified with
35 additional mechanism s of gene therapy once the cell type(s) of interest is/are identified (108) The initial gene therapy treatments began with n on viral gene therapy in animal models, not in humans, and employed plasmid DNAs transfecting cells at the site of injection. These can be delivered as naked plasm id DNA s through electroporation, sonication, or with the DNA bound to polymers or cationic lipids to incre ase cellular uptake and gene expression. Non viral gene therapy is a n interesting delivery option due to its ability to carry genes of any size and it has an extremely low immunogenicity H owever, with time the DNA will be naturally degraded or if the cells divide the gene of interest will be diluted with each cell division If the delivered DNA integrates into chromosomal loci it may stably persist but there is a significant risk of endogenous gene disruption or gene activation that may have dele terious consequences, inclu d ing tumor initiation. Some clinical applications of non viral delivery therapy trials include intramuscular injection of naked DNA, a melanoma cancer vaccine using cationic lipids, chronic limb ischemia, and a canine melanoma va ccine (108, 222) T here are several different types of viral vectors for gene delivery each with its own advantages and disadvantages. The pri mary methods for retinitis pigmentosa treat ment therapy are Adenovirus lentivirus, a nd Adeno associated virus (AAV) I will first discuss is Ad enovirus vectors which are fairly widely used and have a broad range of therapeutic applications. Ad enovirus is an icosahedral capsid DNA virus with a nascent DNA of 36 kb. Ad enovirus is a potentially useful g e ne therapy delivery tool because it can package up to 38kb of DNA. The Adenovirus capsid will prefere ntially infect hepatocytes in the liver, but different co factors as well as different capsid mutations allow for the virus to evade initial liver infection and allow the virus to travel
36 through the circulatory system to reach the tissue of interest. Adenovirus has been utilized in a broad range of treatments including multiple types of cancers and cardiovascular treatments. One of the primary complications with Adenoviral therapy is that it has a strong inherent immunogenicity and this will in turn st imulate a hos t immune response against the therapeutic vector construct. This has been an ongoing concern with the use of Adenovirus for clinical therapies (108) Because Adenovirus can package very large DNA constructs it may be the best possible deli very system for some genes of interest. There have been several Adenovir al gene therapies te s ted in the retina but most had low transduction efficiencies or infected the wrong cell type (108) Recently, there have been some significant advances in the optimization of the Adenoviral vector for optimal retinal delivery to the cell type of interest. For example, in 2014 Dr Hu developed a helper dependent adenoviral vector retinal gene therapy to deliver large DNA constructs. This new delivery system has the inverted terminal repeats necessary for DNA replication without the presence of the viral coding genes. They tested a ubiquitous CAG promoter, along with a separate construct containing the interphotoreceptor retinoid binding protein enhancer (IRBPE) i n order to specifically target PR cells. This system increased the transduction efficiency which resulted in a greater number of viral particles per cell. As a re sult, they observed sustained expression th roughout the RPE with a subretinal injection using low viral doses for at least 4 months (140) This suggests that it may be possible use Adenovir al vectors to treat the retina with larger genes such as in USH to develop gene therapy treatments for retinitis pigmentosa
37 Lentiviral vectors are a nother mode of viral gene therapy which is an HIV RNA based virus that can package up to 10 kb of a therapeutic construct. Lentivirus can infect dividing and non dividing cells and stable transfection has been demonstrated for multi ple organ systems. Similar to the other viral therapies, lentivirus can be pseudo typed to optimize for cell type specific targeting of the gene of interest. Additionally, the promoter and regulatory regions flanking the gene of interest can also modify th e level of cell type specific expression. Integration deficient vectors have also been designed for infection in non dividing cells, however, this is not ideal for dividing cells since, like in all cases, the gene of interest will be diluted out over time as the cell divides. As with the other two viral methods, lentivirus also poses a risk of an innate or adaptive immune response to the capsid (108, 222) Most recently in 2014 D r Cosgrove developed an EIAV lentiviral vector to develop a gene therapy for Myosin7A, th e causative gene for USH1B. He utilized a CMV driven Myosin7A vector with a subretinal injection Affected mice were analyzed 1 month post injection and they reported expression of Myosin7A in both PR and RPE cells (240) Previously, they show ed that Myosin7A KO mice have a delay in transducin translocation upon exposure to light He also showed that after short term 6 day continuo us light exposure (2500 lux) and lon g term cyclic light/dark e xposure of medium light (1500 lux) induced significant PR degener ation in Myosin7A KO vs WT mice (176, 219) Post EIAV lentiviral treatment with CMV Myosin7A revealed a significant rescue of both the transducin translocation and light damage phenotypes. The y also validated the safety and tolerability in n on human primates and prove d that the gene therapy was safe and well tolerated (51, 240) This is extremely promising for the progress of large gene therapy in the retina, particularly for USH diseases.
38 AAV is the most prevalent method of viral g ene therapy to date and has had the most clinical success particularly in the retina. In patients that present with later onset retinitis pigmentosa gene therapy is an available treatment option, particularly if there are relatively intact remain ing regions of the retina The eye is a relatively immune privileged organ that is isolated from other tissue systems, and therapeutic treatment can be admini stered fairly easily through an intraocular injecti on. Additionally, retinal and RPE cells are terminally differentiated, meaning that any AAV DNA that is delivered and taken up by the retinal cells will remain within those cells for the remainder of their life since it will not become diluted ou t over time as the cells divide (108) AAV is a non pathogenic parvovirus that can infect both dividing and non dividing cells. AAV has the smallest packaging capacity of all the viral vectors with a size of onl y 5kb. The AAV genome is a linear strand of DNA that consists of two reading frames containing the replication, capsid and assembly coding genes. There are 9 or more gene products that are controlled by 3 promoters with the addition of differential splici ng and alternative translation start site Flanking these sequences are the inverted terminal repeat sequences that are necessary for packaging (66, 108, 222) The replication gene codes for 4 proteins that are involved with viral replication and packaging of the virus (Rep40, Rep52, Rep68, and Rep78). The capsid proteins form the outer surface of the viral capsid (165) There are 3 capsid p roteins that are composed of two smaller proteins (VP1 and VP2) and one larger protein (VP3). These capsid proteins are in a ratio o f 1:1:10, with VP3 being the mo st abundant. There are multiple different capsid serotypes that have been identified to date as well as multiple different capsid serotype specific mutations that have been identified to optimize for bet ter cell type specific
39 transduction T his will be discussed later in Chapter 5: Serotypes and modifications The AAV constructs that are utilized for gene therapy have all of the replication and capsid genes removed so that all that remains are the inverted terminal repeat sequences for packaging and then the promoter, enhancers, and gene of interest are inserted in the middle along with a poly aden ylation sequence at the end (108) There are many studies looking at the success of AAV mediated gene therapy in the retina and the optimization of targeting specific cell types with different capsid mutants. The early gene therapy studies in the late type specific transduction after subretinal and intravitreal injection as well as long term expression, stability, and toxicity (108 222) These preliminary studies initiated the development of successful gene therapy replacement treatments for several animal models of retinitis pigmentosa The first most ext ensively studied model was for mutations in the RPE65 gene. Part of the early success for this clin ical trial was that there were b oth naturally occurring mouse and dog models (47 48, 102, 115) They showed that subretinal treatment of an AAV2 CBA RPE65 lead to increased impro vement in ERG post treatment T his was also sho wn with several other serotypes and was additionally assessed in both mice and dogs (46 48, 102, 115, 117) After initial treatment in humans, it was show n that treatment mediated improved vision, however it could not altogethe r halt the PR degeneration over several years in patients It was also reported that the only way to halt PR cell death in dogs was to treat prior to any onset of degeneration (27, 28, 47 49, 102) Patients hav e been monitored for sev eral years with s ome treated patients develop ing injection in order to gain visual acuity. This indicated that the patients were able to
40 undergo cortical learning in order for the brain to receive a useful visual signal where it did not previously exist (20, 46, 47, 102, 117) Sim ilar preliminary studies have been done for other autosomal recessive retinal conditions, but so far only in animal models, particularly in LCA mouse models including aryl hydrocarbon receptor interacting protein like 1 (Aipl1), guanylate cyclase 1 (GC1) and spermatogenesis associated protein 7 (Spata7) (28, 30 32, 137, 244) Retinal gene therapy has also been assessed in th e BBS retinitis pigmentosa mouse models which show that PR cell death and rhodopsin miss localization can be rescued post treatment in both BBS4 and BBS1 KO mice (42, 205, 209) Similarly, BBS AAV gene therapy has also been utilized for other organ systems incl uding olfactory epithelia and Adenovirus gene therapy has also been used for other ciliopathy gene s such as Ift88 (153, 154, 232) Some cilia genes are too large to fit in a standard AAV vector for gene therapy, and therefore other methods of deliver y n eed to be utilized such as a dual vector AAV system to deliv er large gene s in two parts, allowing them to recombine in the cell. This was used for USH1B rath er than a lentiviral or Adenoviral vector carrier (144) Most recently in February of 2017, it was shown that AAV mediated delivery of USH1C can prolong auditory hair cell survival, rescue sensory transduction, and rescue the auditory brain stem response post treatment (173) Together, these results confirm the therapeutic potential for the use of AAV as gene therapy tool, not only in the retina but several other cell types as well. Usher Syndrome Usher Syndrome (USH) is a recessively inherited cla ss of disease s defined as combined hearing and vi sion loss due to mutations in nine known genes. Recently, it has also been shown that USH mouse models possess a defect in their odorant detection abilities as well that may also provide an addition al diagno stic tool for patients
41 (120) USH disorder s affect approx imately 1/23 ,000 p eople in the US and are the cause of a bout 50% of combined blind deaf patients. USH was first reported by Charles Usher in 1914 in a patient cohort (233) The re are three forms of USH that are classified by the severity and age of onset of symptoms as well as additional vestibular dysfunctions USH 1 is the earliest onset and is the most severe. It presents with severe bilateral hearing loss at birth and children are either born completely deaf or will present with profound hearing loss by one year of age. USH1 patients also present with extensive vestibular dysfunction in addition to hearing los s and often have delays in motor devel o pment (236) Retinitis pigmentosa in USH1 patients presents very early in childhood and first displays as tunnel vision along with decreased visual acuity followed by rapid vis ion loss (10, 158, 233, 236) USH1 is known to be caused by five known genes. USH1B is caused by mutations in the Myosin7A ( MYO7A ) gene. MYO7A mutations are the most prevalent of all USH mutations resulting in ~50% of all USH cases based on the genetic population. MYO7A is a n actin based motor domain protein that is a non conventional myosin. It is co mposed of a motor head domain, two FERM (F for 4.1 protein, E for ezrin, R for radixin, and M for moesin) domains, two MyTh4 domains, five IQ calmodulin binding motifs, and a Src homology 3 (SH3) domain (192) USH1C is caus ed by mutations in the h armonin gene USH1C is the most prevalent mutation among the Acadian population in Louisiana (236) Harmonin is a scaffolding protein that has three isoforms. All isoforms contain two PDZ (Post synaptic density 95 (PSD95), disc large, ZO1 proteins) domains, class A and B isoforms contain an additional PDZ domain and class B isoforms also have a 2 nd coil coil domain (192) USH1D and UDH1E are cause d by mutations either in the c adheri n 23 (Cdh 23) or
42 protocadherin 15 (P ch 15) g enes both cadherin related protein s Cadherin proteins are very large calcium dependent single pass transmembrane cell adhesion proteins that are indirectly connected to the actin cytoskeleton through their C terminal PDZ binding motif (192) Recently, M. Zallocchi has shown that Pch15 functionally interac ts with the USH3A protein CLARIN1 (CLRN 1) in zebrafish cochlear hair cells via immuno precipitation and the fact tha t Pch15 is mislocalized when the C terminus of Clrn1 is deleted (169) Finally, USH1F is caused by mutations in the SANS gene which encodes a scaffolding protein (10) Sans contains three Ankyrin domains with a CT PDZ binding motif (192, 236) In addition to USH1 proteins being present in PR s and cochlear hair cells, it h as recently been reported that h armonin, SANS Cdh23 and Pch 15 are all present in olfactory sensory neurons at both the mRNA and protein levels (120) USH2 patients are characterized by the age of onset of hearing loss in their early teens although some can present earlier on in life or in their adolescence and hearing loss tends to be stable throughout the pat with no known vestibular dysfunction Retinitis pigmentosa also arises later in USH2 patients and usually is not diagnosed until after puberty (158) Interestingly, it has been shown that USH2 patients also present with a significant reduction in choroidal thickness that is correlated with age (52) It has been shown that in retinitis pigmentosa patients there can be altered choroidal blood flow which could provide addit ional confounding variables with disease pr ogression (141) This is significant because analy zing choroidal thickness may also help to provide an additional monitor for disease progression and help researchers to understand more about the disease mechanism USH2 is known to be caused by three genes. USH2A patients carry mutations in the g ene u sher in which cause s between 50
43 90% of USH2 cases depending on the genetic population. There appears to be a founder effect in the French Canadian population, accounting for approximately 50% of the USH2A cases with a secondary founder population of European or igin accounting for approximatel y 15 45% of all mutated alleles (158) There are two isoforms of u sherin, the short isoform is a laminin type EGF (epidermal growth factor) like protein with 4 fibronectin domains. The long isoform is a transmembrane protein with a C terminal protein binding motif (192, 236) USH2C is c aused by mutations in the VLGR1 (very large G protein coupled receptor) gene which is the largest of all USH genes VLGR1 is a seven transmembrane Ca2+ exchanger that only appears to account for a small percentage of USH2 cases at 3 6% of patients (158, 236) VLGR1 is believed to play a role of adhesion in the synaptic membranes and assist in G protein signaling during synaptogenesis (192) USH2D i s caused by mutations in the w hirlin gene and is a proline rich protein with 3 PDZ domains. Whirlin mutations appear to be extremely rare compared to other USH mutations (10, 158, 236) Similar to the US H1 proteins, the USH2 proteins u sherin and Vlgr1 have also been reported to be present in olfactory sensory neurons at b oth the mRNA and protein levels (120) USH3 A is the lates t onset of all the USH disorders and is cause d by mutations in only one gene to date, Clarin1 ( C LRN 1 /CLRN1 ). USH3 A is the least prevalent of all USH forms but mutations in CLRN1 are mostly seen among the Finnish and Ash kenazi Jewish populations which account s for more th an 40% of patients due to a strong founder effect in both cohorts USH3 A patients present with progr essive retinitis pigmentosa (131, 239) Affected patients will experience progressive tunnel vision and reduced visual acuity over time. Different from the other types of USH,
44 USH3 A hearing loss will present between the 1 st and 2 nd decade of life, but will become profound over time (171, 172, 239) This is unique compared to USH1 and 2 because hearing loss in USH3A presents post lingual ly and allo ws for patients to develop fairly normal speech abilities. This also suggests that USH3 A may be the most detrimental in terms of patient quality of life because patients are born with normal hearing and vision, but progressively lose both senses later in l ife (10, 158, 185, 186, 225) Patients show a significant level of visual field loss and analysis of patient retinas by ERG sh ow definitive PR dysfunction with a greater de ficit in rod function than in cone s and t heir fundus exam s show significant retinitis pigmentosa phenotypes in the peripheral retina Upon SD OCT analysis, the ON L was significantly thinner than the standard thickness and the IS/OS band was disrupted within 4 degrees of the fovea (189) As stated previously, there is currently no treatment for retinitis pigmentosa in U SH3A patients, however many patients have received cochlear implants. Those patients with cochlear implants show a significant improvement in hearing and also showed a significant improvement in word recogniti on post treatment (183, 185) CLRN 1 is a tetraspanin TM protein with 3 isoforms The main isoform contains an N48 glycosylation residue with a proposed PDZ bindi ng motif on the intracellular C terminus (4, 132, 192, 218, 241) The true function of CLRN 1 is unknown but it is believed to localize at the CC, kinocilia and stereocilia as other USH proteins (53, 146, 241) Although the other USH proteins were recently shown to be expressed in the olfa ctory epith elium, there is no evidence as of yet whether there is any CLRN 1 expression as well. Cl rn1/CLRN1 will be discussed in Chapter 3 background.
45 Usher Interactom e The USH proteins are believed to functionally interact with each other along the stereocilia and k inocilia in the inner ear and along the CC, calyceal processes, an d incisures in PR cells (5, 198) Each of the proteins in this network have different functions that provide membrane membrane connections, scaffolding proteins for structural maintenance, and motor proteins that allow for the movement of cargos. MYO7A is the only known motor protein asso ciated with this functional network and has been shown to interact with h armonin and w hirlin to connect the USH interactome to the intracellular actin cytoskeleton (135, 147, 192) MYO7A has also been shown to directly interact with Pch15 in the inner hair cells (204) with additional protein protein specific interactions between the C terminal PDZ binding motif in SANS to the PDZ domains in whirlin (3, 5, 35, 135, 147, 170, 191, 192, 229) Of a ll the USH proteins, CLRN 1 is the only protein that has not yet been conclusively linked to the interactome. As stated previously, r ecent studies showed that the C terminal domain of C lrn 1 functionally interacts with Pch 15 in zebrafish neuromast hair cells and this inter action is essential for Pch15 incorporation into cochlear hair cell bundles and its proper localization at the tip links of stereocilia. Interestingly, a P ch15 deletion did not alter C lrn 1 localization suggesting that C lrn 1 is required for p roper Pch15 localization and function but that C lrn 1 can localize independently of Pch15 (146, 169) C lrn 1 was also found to be essential for the hair cell synaptic ribbon arrangement Clrn1 zebrafish morpholinos displayed a definitive increase of synaptic ribeye puncta and an increase in puncta mislocalization. This phenotype could be rescued w ith the full length Clrn1 however, the C terminal deletion of C lrn 1 did not. This indicates the C terminus is required for proper ribeye localization at the ribbon synapses as well as for synaptic formation (169)
46 Furthermore, C lrn 1 was shown to be required for normal vesicle re cycling a t the ribbon synapses and the C terminus was not required for proper vesicle recycling (16 9) The USH interactome proteins are also linked to other ciliopathy proteins including Cep290, RPGR, RPGRIP1, lebercilin, and BBS6. In particular, SANS has been shown to directly interact with Cep290 (211) The defects in C lrn 1 vesicle recycling was also assessed using a rab11a/b cilia marker for trafficking and it w as shown that C lrn 1 was required for accumulation of rab11a/b at the apical pole in inner hair cells (169) Given that this interactome is present in both cochlear hair cells a nd PR cells, its functions appear to be cell type specific, although little information relates to the function of USH proteins in olfactory neurons (120) In the cochlea, t his interactome is believ ed to pl ay a role in sensing sound waves through cilia bending and interactions at the tip and lateral links on the inner ear stereocilia (5, 54) When the tympanic membrane senses a sound it relay s a vibration that is sent through the fluid matrix in the cochlea and the stereocilia will bend in response to fluid movement. The USH interactome proteins are believed to activate when the stereocilia move relaying the signal into the cell and results in neurotransmitter release at the cellular synapses (23, 148) Additionally, harmonin functionally acts as a scaffold for ion channel complexes at the ribbon synapses in order to regulate Ca2+ signaling and electrical signaling at inner hair cell synapses (84, 147, 211) In PR cells the USH interactome functions as an adhesion complex along the CC, periciliary membrane and the calyceal processes (54, 158, 199) As mentioned previously, they are present in frogs birds, and primates but are lacking in mice. They are believed to act as stabilizers of the OS and USH proteins help facilitate O S disc formation and biogenesis (148, 211)
47 Usher Syndrome Animal Models Many of the naturally occurring mouse models display a distinctive behavior of head tossing and circling around themselves. This is believed to be a consequence of inner ear vestibular dysfunction and abnormal balance issues. This behavior seems to be restricted to the USH1 mutations, however there is one model for USH2 that does indicating that some U SH2 patients could perhaps have mild vestibular problems. A commonality of all USH mouse models is that they present with definitive inner ear defects, having a loss of stereocilia and kinocilia that result in hearing loss However, they have very mild, if any, retinal phenotypes that ha ve been extremely difficult to characterize (233) Although there does not appear to be any structural abnorm alities in t he retina, there is a reduced a and b wave ERG in some of the USH1 and 2 mouse models. A mouse model for USH1C shows mild peripheral degeneration and USH2D mice have shortened PR OS (237) MYO7A KO mice have an accumulation of rhodopsin in their OS and their RPE have an abnor mal accumulation of melanosomes (51, 144) Currently, the only USH mouse model that has been reported to have measurable PR loss is an USH2A mo use which loses approximately 50% of PR by 2 years of age (233) Recently, in a new USH3A mouse model Y. Im a nishi has reported a slight decrease in b wave ERG at 7 months of age, but as the WT animals age, their ERG will drop to KO levels at 9 months of age (217) This late onset loss of function is believed to be a result of abnormal RPE function in this mouse strain a s well as an increase in inflammation (160) The re are also some additional retinal phenotypes of delayed phototransduction protein localization upon exposure to light. This has been seen in mouse mo dels for all three forms of USH (72, 176, 219, 220) The USH3A mouse models and phenotypes will be discussed further in Chapter 3 and 4.
48 Recently, several zebrafish morpholino models have also been generated in order to attempt to identify additional retinal phenotypes as well as study USH protein localization and function in a mo del organism. Zebrafish have a much short er developmental time frame and their retinas are more cone dominant, making them relatively more similar to a primate retina than a mouse They are also diurnal and more dependent on visual processing (22) Zebrafish are a good model to test a potential for a gene therapy treatment because the zebrafish retina can also regenerate PR c ells i n vivo (22) There are currently zebrafish models for all three USH syndrom es and successful th erapeutic treatments have been of recent growing interest in these models in order to validate the effectiveness of treatments.
49 CHAPTER 2 OVERVIEW, RATIONALE, AND SPECIFIC AIMS USH retinal diseases are caused by a group of autosomal recessive genetic disorders that present with combined hearing and vision loss in patients. The three classifications of USH account for 50% of combined blindness/deafness in humans and affects 1/25,000 children worldwide (10, 23, 135, 199, 233, 236) The three USH types are clinically differentiated by the age of onset and the mutated gene involved (23, 135, 236) USH3A is the slowest onset of the USH sub types and is therefore the most promising target for treatment. Current treatments for USH3A are costly and limited only to hearing defect s utilizing hearing and vision aids, cochlear implants being the most effective In USH3A, hearing loss results from the loss of function of vestibular cochlear cells in the inner ear, whereas vision loss and progressive night blindness is caused by a loss of PR cells in the retina and is characterized clinically a s retinitis pigmentosa (20, 107, 171, 172, 183, 185, 186, 189, 192) USH3A is caused by mutations in the Clrn1 gene, which encodes a 232 amino acid four transmembrane domain protein of unknown function. Its pattern of endogenous expression in the retina remains unclear It has been shown that Clrn 1 contains an N48 glycosylation and PDZ domains which are likely responsible for proper protein localization and facilitation of its interaction with other USH proteins The USH proteins are localized to photoreceptor IS, CC, ONL and PR/ bipolar cell synapses in the OPL, as well as to the kinocilium and stereocilia bundle of cochlear c ells in the inner ear (3 6, 8 10, 14, 17, 23, 35, 53, 54, 6 8, 69, 88, 89, 93, 94, 105, 107, 114, 117, 131, 135, 138, 139, 142, 147 149, 152, 158, 169, 172, 178, 191 194, 198, 199, 204, 211, 217, 233, 237, 241, 242) T he retinal location of Clrn1 /C lrn 1 remains controver sial because previous localization data is conflicting depending on the
50 assay performed (4, 87 89, 114, 182, 218, 241) A primary difficult y with regard to understanding USH3A function is that no current animal models fully recapitulate patient phenotype s The g enetically engineered models, N48K knock in ( KI ) and Clrn1 knock out ( KO ) (Clrn1 / ) mice, described to date present with a hearing loss phenotype while their retinas appear morphologically and functionally normal based on histology, OCT and ERG analysis (87 89, 218) In order to develop an effective treatment for USH3A, a preclinical an imal model is required that more accurately mimics the patient phenotype because there is currently no restorative retinal therapy for Usher patients. Therefore, uncovering a ret inal phenotype in available USH3A mouse models and determining the endogenous localization of Clrn1 /C lrn 1 in the retina are crucial for the development of clinical t herapies T he aims of my study are to define an USH3A mouse retinal phenotype, identify the endogenous retinal location of C lrn 1, and successfully rescue an identified retinal phenotype using AAV mediated gene therapy Aim 1: Identify endogenous Cl a r i n 1/ CL A R I N 1 expression in the retina. Hypothesis: Clrn1 / C lrn 1 is expressed in photoreceptor cells and its N48 glycosylation and PDZ binding domain are responsible for proper localization. Objective 1.1: Document Clrn1 /C lrn 1 retinal expression and localization in the presence or absence of photoreceptors. Objective 1.2: Identify endogenous C lrn 1 l ocalization by immunohistochemical (IHC) analysis employing new CLRN1 antibodies in mouse retinal tissue. Objective 1.3: Document C lrn 1 subcellular localization when N48 glycosylation is absent. Aim 2: Identify a retinal phenotype in N48K Knock in (KI) (N4 8K) and Cl a r i n 1 Knock out (KO) ( Clrn1 / ) mice. Hypothesis: There is a subtle but quantifiable visual phenotype in the USH3A mouse models.
51 Objective 2.1: Perform light driven protein translocation experiments in mice containing the Clrn1 / KO or N48K KI mutation to characterize a retinal phenotype. Objective 2.2: Replicate the identified decreased b wave phenotype in the new A/J mouse strain. Objective 2.3: Perform prolonged light exposure experiments and light damage experiments in Clrn1 / KO or N48 K KI mice to characterize a retinal phenotype and potentially induce degeneration compared to wild type mice. Objective 2. 4 : Perform novel ERG methods using light exposure prior to ERG in order to characterize a retinal phenotype in Clrn1 / KO or N48K KI mice. Aim 3: Identify the optimal AAV gene construct, capsid serotype, and delivery method for Cl a r i n 1 gene therapy treatment and assay for phenotypic rescue using AAV mediated gene therapy. Hypothesis: Clrn1 needs to be expressed at optimal levels by the correct cell type in order to develop a successful gene therapy treatment that can restore the previously defined phenotype through AAV mediated gene therapy. Objective 3.1: Assess CLRN1 localization following AAV injection containing human C LRN 1 Obje ctive 3.2: Optimize vector capsid, promoter, and titer for C LRN 1 /CLRN1 safe and effective AAV gene therapy treatment. Objective 3.3: Determine if exogenously expressed C LRN 1 /CLRN1 can rescue the phenotypes previously described in Aim 2 Clrn1 / KO or N48K KI mice.
52 CHAPTER 3 IDENTIFYING WHERE ENDOGENOUS CLARIN 1 RNA AND PROTEIN IS EXPRESSED IN THE RETINA Background It has been established that the majority of USH proteins localize to the CC and calyceal processes in PR cells and at the kinocilia and stereocili a bundle in cochlear hair cells (3, 147, 192, 198) As mentioned previously, many USH proteins contain PDZ domains, PDZ binding motifs, as well as transmembrane components all of whi ch have been proposed to link the USH interactome to the intracellular cytos keletal network through macromolecular complexes (3, 135, 211) As of yet USH3A is the only USH protein that has not been definitively linked to the USH interactome. USH3A is caused by mutati ons in the Clrn1 gene which consists of 4 exons. There are 3 proposed protein coding transcripts in the retina: isoform 1 containing exons 1, 3 and 4 which is the 232 aa isoform that is 26 kilo Daltons (kDa) and is believed to be the primary variant, isofo rm 2 containing all exons 1 4 which is the 250 aa isoform, and isoform 3 containing exons 1 and 4 the smallest isoform at 180 aa (Ensemble: Isoform 1: Clrn1 201 ENSMUST00000051408.7 Isoform 2: Clrn1 202 ENSMUST00000055636.12 isoform 3 Clrn1 203 ENSMUST0 0000072551.6 ) It is unclear if all three isoforms are expressed in the retina or not, and if so, which re t inal layers are expressing which isoform. Isoforms 1 and 2 are thought to contain four transmembrane domains and isoform 3 contains only two with in tracellular N and C terminal ends (4) The main variant codes for a tetraspanin transmembrane protein with a n extracellular glycosylation resid ue on Asparagine (N) 48 (218) This transmembran e tetraspanin structure has been supported through recent 3D protein modeling utilizing multiple different software for protein domain analysis as well as I TASSER for final
53 structure prediction (132) T he most common patient mutation in the US is a loss of this N48 glycosylation which is believed to render the protein non functiona l and target it for degradation (114) Cell culture experiments looking at multiple mutations including those in transmembrane 1, transmembrane 3, and the N48 loss of glycosylation (N48K), show that WT Clrn1 localizes to the plasm a membrane and the mutant C lrn 1 lo calizes almost completely with the ER (107, 114, 218) In addition to its glycosylation, C lrn 1 also contains a PDZ binding domain on its intracellular C terminal tail that may facilitate its interaction s with other transmembrane proteins (100, 101, 146) In vivo C lrn 1 has been proposed to be associated with the ribbon synapses CC, kinocilia, stereocilia bun dle a nd the cellular cytoskeleton in both PR and cochlear hair cells (87 89, 93, 182, 218, 241, 242) Functionally, C lrn 1 was reported to be essential for the development, neuronal activation, and synaptic maturation of cochlear hair cells (87, 88, 241, 24 2) Because C lrn localization pattern and functional significance in cochlear hair cells is similar to other USH proteins, this suggests that C lrn 1 also belongs to the USH protein network in both the hair cells and photoreceptors (3, 135, 198, 211, 241) D ue to conflicting results depending on the assay performed t he precise retinal localization of Clrn1 /C lrn 1 has no t been conclusively established (53, 169, 182, 241) One group utilizing retinal transcriptome analysis reported Clrn1 expression only in starbu r st amacrine cells (207) Dr analyzed Clrn1 mRNA expression in the retina by i n situ hybridization, followed by laser capture microdissection and RT PCR. They reported that Clrn1 mRNA was expressed through out development and was absent at adulthood. cDNA analysis indicated that the transcript detec ted correspond ed to isoform 1 (232 aa variant) and 3 (150 aa variant) and do not mention isoform 2 (250
54 aa variant) even though their RT PCR gel shows a faint product for that isoform (Geller figure 1) (Figure 3 1) (87) Clrn1 appears throughout the embryo, head, and brain during development as well as in the retina T hey further observed Clrn1 in multiple brain tissues, however for all of these assays they used prime rs in exon 3 and 4, with no way to distinguish whether the isoform 2 (250 aa variant) is expressed in addition to isoform 1 (232 aa variant) because these primers would amplify both transcript s (Geller et al figure 1) (87) I n situ hybridization showe d mRNA expression in the INL and not in PR cells in WT fetal retinas, but no mRNA expression was observed in KO mice and Clrn1 expression declined to no transcript by adulthood ( Geller figure 2) (87) Laser capture dissection showed Clrn1 in the INL using isoform 1 primers (232 aa variant) and t hey suggest Clrn1 is only expressed by retinal Mller glial cells (Geller figure 5) (87) This group also generated the first Clrn1 KO mouse thr ough loxP Cre excision of exon 1 containing the ATG start codon. A difficult y with this mouse model is the genotype for a Clrn1 / is the lack of a PCR product which creates the possibility that a h et erozygote mouse could be mistakenly genotyped as a KO due to a failed PCR reaction (87) They furth er tested a novel antibody generated against human CLRN 1 and saw identical staining in WT and KO retinas ; h owever, they do not test this antibody in cochlear tissue Another group looked at laser capture microdissection in the mouse cochlea and report ed the highest level of Clrn1 expression in the spiral ganglion cell region similar to all the other USH proteins they tested (166) In another study, Dr. Zallocchi and colleagues develop ed another novel CLRN 1 human polyclonal an t ibody that recognized all CLRN1 isoforms and they report labeling at the stereocilia and synapses of inner hair cells and the CC and ribbon synapses in
55 PR cells (241) One difference between their methods and the methods of others was that they did not utilize any fixative prior to sectioning and staining which may help to preserve the antibody epitope better than if the samples are fixated prior to analysis. Dr. Zallocchi and colleagues provided a supplemental appendix demonstrating the specificity of their antibody by reporting no s taining in the ir Clrn1 / KO mouse model (53) When Clrn1 expression was assessed via RT PCR, they observed products only for what they reference as isoforms 2 (232aa variant) and 3 (180aa variant) which correspond to isoforms 1 (23 2 aa variant) and 3 (180aa variant) in Ensembl, respectively. They further showed that only one i soform (232 aa variant) was observed via western blotting. Their subcellular localization in both inner hair cells and photoreceptors coincides with the localization of all other USH proteins at the stereocilia, kinocilia, and ribbon synapses in the inner ear as well as the CC and calyceal processes in PR cells (198) Given these similari ties, I hypothesize d that Clrn1 /C lrn 1 is indeed expressed in PR cells and that the opt imal therapeutic vectors should be targeted for PR cells in the retina This will be discussed in Chapter 5 Due to the phenotypic and biochemical inconsistencies with the current mouse models, a few groups have utilized zebrafish for a model organism to b etter recapitulate the retinal phenotypes of USH in patients. In zebrafish Clrn1 has been observed from embryo through adult hood in both the cochlea and retina. Zebrafish i n situ analysis shows Clrn1 in hair and supporting cells in the ear and PR and INL cells in the retina (93, 169, 182) In zebrafish, Dr. Phillips and colleagues r eport ed i n situ hybridization of Clrn1 in sensorineuronal cells with weak expression in the brain, retina, and inner ear. In the retina, both ONL and INL cells contained Clrn1 transcripts Interestingly, Clrn1
56 INL staining was present at embryonic stages and re stricted to amacrine cells in adult fish, similar to the Dr. Geller et al. pap er; however ONL Clrn1 expression was present at all ages (182) The y made a cus tom antibody for zebrafish Clrn1 targeting the 1 st extracellular loop containing the N48 glycosylation. They transfected embryos with a C lrn 1 HA tagged construct and looked for colocalization of an anti HA antibody with their Clrn1 antibody They showed C lrn 1 loca lization at the synapses and apical cell body of inner hair cells with no staining in the stereocilia. In the retina, C lrn 1 was expressed in the INL basal portion where amacrine cells reside as well as in the ONL and G CL. Furthe rmore, they saw significant C lrn 1 expression at the OLM as well as the OPL and partial co localization with Mller cells at the OLM and INL. In the adult retina they observ ed additional C lrn 1 staining at cell cell contact s between rod and cone PR cells (182) A second z ebrafish study utilized a Novus Biologicals (NB) rabbit anti human CLRN1 antibody and t hey observed C lrn 1 localization at t h e cochlear hair cell bundle where it colocalize d with F actin and they observed n o C lrn 1 antibody staining in the KO morpholinos (93 ) This is significant because it show s C lrn 1 localization at the cochlear hair bundles and that their commerci al antibody is specific for C lrn 1. This also provides f urther evidence that C lrn 1 has an overall localization pattern similar to all other USH proteins. The only complication with this study is that morpholino treated zebrafish retinas possess ed a normal ERG response compared to WT fish (93) A primary complication with studying USH3A mouse mo dels is that the mice exhibit extensive hearing loss by 4 months of age but do not develop any previously defined retinal phenotype (87 89, 93) Additionally, Clrn1 / KO mi ce display a characteristi c circling behavior by 6 months of age. As mentioned above, some studies
57 claim Clrn1 /C lrn 1 is expressed in the inner retina, whereas others claim Clrn1 /C lrn 1 is in photoreceptor cells (53, 87, 88, 241) An additional complication is that the Clrn1 / KO mice may potentially still express an isof orm of Clrn1 / C lrn 1 because the previous antibody stained both WT and KO retinas (87, 88) There are 11 proposed Clrn1 mRNA splice variants predicted to produce protein products and there are three known C lrn 1 isoforms thought to be expressed in the retina, but it is unclear which retinal layer or cell type is producing which isoform (4, 226) Due to the disagreements among the previously published data, it is critical to identify the endogenous retinal localization of Clrn1 /C lrn 1 in order to develop a successful gene therapy treatment f or USH3A. As noted above, Clrn 1 / KO and N48K KI mice were also generated on both the CBA/J and A/J backgroun ds. CBA/J mice carry a mutation in phosphodiesterase PDE6B that cause s a complete loss of photoreceptors by 3 months of age (rd1 mutation) (43) By utilizing these mice, I c ould determine if Clrn1 is expressed in PR cells and w hich isoforms of Clrn 1 are removed in Clrn1 / ; rd1 / double KO mice The second strain of A/J mice additionally carry an Usher Syndrome Type 1D mutation in Cadherin 23 (Cdh 23) that causes hearing loss, but no clear retinal phenotype (7, 26, 69, 70, 129, 138, 139, 142, 167, 201, 208, 243) Cdh 23 was shown to interact with the Usher Syndrome Type 1B gene (USH1B) Myosin 7A, and is a glyco sylated protein with a C terminal PDZ binding motif involved in mediating cell cell contacts and calcium dependent interactions (7, 10, 26, 69, 70, 129, 138, 139, 142, 201, 208) These mice w ere used to compare Clrn1 / ; Cdh23 / and Clrn1 +/+ ; Cdh23 / mice in the context of a secondary USH mutation to potentially induce a retinal phenotype and perhaps identify potential interacting partner s of C lrn 1. Additionally, the
58 A/J strain has a smaller ERG amplitude than control C57BL/6J mice and they lose sign ificant ERG amplitude over time (Figure 4 7, 4 8) (160, 217) A/J WT mice also undergo significant loss of ONL nuclei over time c ompared to C57BL/6J mice This retinal degeneration in A/J mice suggests that their retinas age mu ch faster than other strains (160) This could pote ntially reveal an ERG phenotype due to the accelerated aging process. This will be discussed in Chapter 4. Overall, my experimental plan is aimed to determine whether Clrn1 /C lrn 1 is expressed in inner retina, outer retina, or both and identify potential interacting partners. I therefore documented mRNA and protein expression using RT PCR immunohistochemistry and western blot analysis in all 3 mouse strains and the results are discussed below. Not all data was acquired for all strains because not all str ains were available throughout the course of this work However, t he strain for each result is defined with each data set. Methods RT PCR The Dr Flannery mixed albino C57BL/6J strain Clrn1 / KO mice the Dr Imanishi A/J albino strain Clrn1 / KO mice, and the Dr Imanishi CBA/J strain Clrn1 / KO mice were used for RT PCR experiments (87, 88) Mice were euthanized with dual CO 2 /cervical d islocation retinas were dissected out and RNA was isolated using a Trizol phenol chloroform method (Life Technologies, Trizol Reagent #15596026). Samples were processed according to the published protocol ( https://tools.thermofisher. com/content/sfs/manuals/trizol_reagent.pdf ) and all centrifuge spins were performed at 16,000 x g for 1 hour. RNA for RT PCR wa s converted to cDNA using a Bio Rad i Script cDNA Synthesis Kit ( Bio Rad #170 8891 http://www.bio rad.com/webroot/web/pdf/lsr /literature/ 4106228C.pdf ) at a max imum concentration of 1ug total RNA and RNA for
59 qRT PCR experiments was processed using a Bio Rad iScript Reverse Transcription Supermi x for RT qPCR ( Bio Rad #170884 1, http://www.bio rad.com/en us/product/ iscript reverse transcription supermix for rt qpcr ) at a max concentration of 1ug total RNA Primers for RT PCR s were based on previo usly published data (Table 3 1) (87) Primer positions are defined in Figure 3 1 A A ll isoforms are based on NCBI Ensembl accession numbers. Isoform 1 is exons 1, 3, and 4 ( ENSMUST00000051408.7) the 232 aa isoform Isoform 2 is exons 1 4 (ENSMUST00000055636.12) the 250 aa isoform and Isoform 3 is exons 1 and 4 (ENSMUST00000072551.6 ) the 180 aa isoform and the predicted RT P CR product sizes were obtained through alignment of the published primers to the NCBI reference sequences in Vector NTI (Figure 3 1B) Previously, Dr. group has tested novel AAV2 serotype capsid mutations that allow for better transduction effi ciency in the retina. I utilized the construct that had the best expression in PR cells in the retina, which was an AAV2 quadruple mutant. The AAV2 Quadruple mutant contains 4 tyrosine to phenylalanine mutations at residues Y F 272, 444, 500, and 730 (AAV2 Quad). The AAV vector was produced and optimized as described previously (65, 71, 82, 83, 103, 128, 134, 179, 180, 197, 245) C57B L/6J mice along with A/J WT and Clrn1 / KO mice were then injected subretinally with AAV2 Quad Y F sc smCBA h C LRN 1 HA at 8.43 x 10 12 vg/ml (Table 5.1 construct 1) as described previously (128, 181, 182, 199). These mice were anal yzed 1 month post injection by RT PCR as described above. Immunohistochemistry Mice were utilized from both strains described above and mice were housed as described above (87, 88, 217) C57BL/6J control mice were utiliz ed for initial antigen retrieval and antibody staining optimization. All mice were housed in a 12 hour dim
60 light/dark cycle and were euthanized as described above Eyes were incubated in 4% paraformaldehyde (PFA) fixative for 1 hour prior to moving to a 1x phosphate buffered saline (PBS) solution For paraffin sections, eyes were embedded using a paraffin processor (RMC Ventana Tissue Processor PTP 1530) and sectioned at a thickness of 4 um for all samples. For cryosections eyes were incubated in 30% sucrose until saturated and then embedded in O.C.T. (Optimal Cutting Temperature) compound (Fisher Scientific Healthcare Tissue Plus O.C.T. compound 23 730 571) an d sectioned on a Leica Cryostat (Leica CM 1900 Cryostat) at 12 um per section for all samples For paraffin immunostaining, antigen retrieval was optimized beginning with base (57) The final optimized paraffin protocol is as follows: Prior to a ny treatment the slides were incubated in a dry incubator lying flat at 60 o C for 1 hour. T he tissue sections were then de paraffinized either with Histoclear II (National Diagnostics Inc. HS 202) or Xylenes (C8H10 X3P 3875 Fisher Scientific). Tissue sectio ns were rehydrated with dilutions beginning with 100% 95%, 90%, 80%, and 70% ethanol for 3 minutes each, followed by 1x PBS. Slides were then incubated with a Protinase K solution at a concentration of 10 mg/ml for 10 minutes at 37 o C then washed 3 times, followed by incubation in 0.5% Saponin for 20 min at room temperature and washed again 3 t imes in PBS. Slides were serum blocked with 1% bovine serum a lbumin (BSA) for 1 hour and incubated with primary antibody in 0.5% BSA overnight at RT. Primary antibodi es are as follows: Novu s Biologicals rabbit anti CLRN 1 (NB CLRN1) 1:2000 dilution ( NBP69142 ), mouse anti Gamma Tubulin 1:1000 dilution (Thermo Scientific Pierce MA1 850) mouse anti arrestin 1 1:100 dilution (courtesy of Dr. WC Smith C10C10 ) (73) mouse anti
61 r hodopsin 1:100 dilution (courtesy of Dr. WC Smith, B6 30) (2) and mouse anti GFP 1:100 dilution ( Invitrogen A11121 ). The second day slides were washed 3 times in PBS and stained with the following secondary antibodies for 2 hours at a dilution of 1:400: goat anti rabbit 488 (A11008 ), goat anti rabbit 594 ( A11012 ), goat anti mouse 488 ( A32723 ), goat anti mouse 594 ( A11 005 ), donkey anti rabbit 488 ( A21206 ), donkey anti rabbit 594 ( A21207 ), donkey anti mouse 488 ( A21202 ), and donkey anti mouse 594 ( A21203 ) S econdary antibodies were acquired from Molecular Probes/Invitrogen, Eugene, OR Lastly, slides were rinsed 3 times in 1x PBS and mounted with VectaShield containing DAPI ( Vector Laboratories H1200 ). Sections were analyzed using a spinning disc confocal microscope at the University of Florida Cell and Tissue Analysis Core ( Olympus DSU IX81) Images were acquired with 10 15 confocal z stacks and images were consecutively de convolved and merged into a single projection image and each individual color channel was saved separately and merged. Cryosections were stained wi th the following protoc ol: T issue sections were air dried at room temperature for 30 min prior to staining. S ections were consecutively incubated in PBS for 5 minutes, followed by incubation in 0.5% Saponin or Triton X 100 for 20 min at room temperature and washed again 3 times in PBS. Slides were then serum blocked with 1% BSA for 1 hour and then incubated with primary antibody in 0.5% BSA overnight at RT. Primary antibodies are as follows: NB CLRN1 1:2000 dilution (NBP69142 ) and mouse anti rhodopsin 1:100 dilution (co urtesy of Dr. WC Smith, B6 30) (2) The second day slides were washed 3 times in 1x PBS and stained with the following secondary antibodies for 2 hours at a dilution of 1:400: goat anti rabb it 488 (A11008 ) and goat anti mouse 594 (A11005 ) ( Molecular Probes/Invitrogen,
62 Eugene, OR ) Lastly, slides were rinsed 3 times in 1x PBS and mounted with VectaShield containing DAPI (Vector Laboratories H1200). Sections were analyzed using a spinning disc confocal microscope in the Cell and Tissue Analysis Core at the University of Florida ( Olympus DSU IX81) Images were acquired with 10 15 confocal z stacks and images were consecutively de convolved and merged into a single projection image and each individual color channel was saved separately and merged. For the C57Bl/6J mice that were AAV treated for c o staining with the NB CLRN antibody, the AAV vector was produced and optimized as described previously (65, 71, 82, 83, 103, 128, 134, 179, 180, 197, 245) These mice were then injected subretinally with AAV2 Quad Y F sc smCBA h CLRN1 Venus at 1.54 x 10 12 vg/ml (Table 5.1 construct 3) or intravitreally with AAV2 Quad Y F sc smCBA h C LRN 1 Venus at 1.54 x 10 12 vg/ml (Table 5.1 construct 4) that both contain a Venus GFP protein analog with the previously described methods (179, 180) M ice were euthanized, and sections were obtained and analyzed 1 month post injection by paraffin immunohistochemistry as described above. Sections were stained with primary antibodies: NB CLRN1 1:2000 dilution (NBP69142) and mouse anti GFP 1:100 dilution (Invitrogen A11121). Secondary antibodies were goat anti rabbit 594 (A11012) and goat ant i mouse 488 (A32723) ( Molecular Probes/Invitrogen Eugene, OR ) Lastly, slides were rinsed 3 times in 1x PBS and mounted with VectaShield containing DAPI (Vector Laboratories H1200). Sections were analyzed using a spinning disc confocal microscope through the Cell and Tissue Analysis Core at the Universi ty of Florida ( Olympus DSU IX81) Images were acquired with 10 15 confocal z stacks and images were de convolved and merged into a single proje ction image and each color channel was saved separately and merged.
63 Western Blot Analysis Mice from both strains described above were housed and euthanized as described above (87, 88, 217) C57BL/6J control mice were utilized for initial antibody staining optimization Eyes were dissected as eye cups, retinas, and remaining ocular tissue. Whole eyes were removed from the mouse, eye cups and retina diss ections were performed by cutting off the cornea and removing the lens. Retinal dissections were then separated from the other ocular tissues by placing a slit between t he retina and c horoid tissue, slowly separating the two layers The NB CLRN1 antibody was tested on all tissue s and eye cups were used for the majority of experiments Samples were sonicated in 200 ul of homogenization buffer for 5 seconds (30mmol/l Tris HCl, 10mmol/l EG TA, 5 mmol/l EDTA, 1% Triton X 100, 250 mmol/l sucrose and 1 mmol/l phenylmethylsulfonyl fluoride, pH 7.5) and stored on ice. Samples were centrifuged at 16,000 g for 1 min to prec ipitate out debris and loaded in a 4 20% SDS PAGE gradient gel (Bio Rad Mini PROTEAN TGX precast gel #456 1094) in a running buffer ( 25mM TrisBase, 192 mM Glycine, and 1% SDS ) for 1 hour at 200 volts 0.08 amps and 30 watts of current. G el s were transferred in a wet transfer system to a 0.45 um pore sized PVDF membrane (Immobilion FL IPFL00010) for 1 hour in a transfer buffer (25mM TrisBase, 192 mM Glycine, and 10% methanol) at 200 volts, 0.2 amps, and 40 watts. Membranes were washed in 50% Odyssey blocking buffer in PBS for 1 hour ( Li Cor 927 40000) followed by 3 washes in PBS with 0.1% Tween 20 and incubated at room temperature overnight with the primary antibody The primary antibodies are as follows: NB CLRN1 1:2000 (NBP69142) mouse anti r hodopsin 1:200 (courtesy Dr. WC Smith B6 30 ) (2) mouse anti arrestin 1 (courtesy Dr. WC Smith C10C10 ) (73) mouse anti gamma tubulin Thermo Scientific Pierce MA1 850). The second day membranes were
64 washed 3 times in 1x PBS with 0.1% Tween 20 and incubated with the secondary antibodies in a blocking buffer mix for two hours (5 ml PBS with 0.1% Tween 20, 1 ml Odyssey blocking buffer, and 10 ul 20% SDS). Secondary antibodies were diluted at 1:10,000 and were as follows: goat a nti mouse IRDye 800CW (Li Cor926 32210) and goat anti rabbit IRDye 680LT (Li Cor 926 68021). Membranes were finally rinsed 3 times with 1x PBS with 0.1% Tween 20 and imag ed on an Odyssey imaging system. For the new A/J Clrn1 / ; Cdh23 / double KO and A/J Clrn1 +/+ ; Cdh23 / WT mice that were treated for co staining with the NB CLRN antibody, the AAV vector was produced and optimized as described previously (65, 71, 82, 83, 103, 128, 134, 179, 180, 197, 245) These mice were then injected subretinally with AAV2 Quad Y F sc smCBA h C LRN 1 HA at 8.43 x 10 12 vg/ml (Table 5.1 construct 1) that contained a hemagglutinin ( HA) tag sequence as described previously (179, 180) These mice were euthanized, and eyecup dissections were obtained and analyzed 1 month post injection as described above. Western blots were performed as described above with primary antibodies: NB CLRN1 1:2000 dilution (NBP69142) and mouse anti HA 1:2 00 dilution (3F10, Roche Diagnostics, Indianapolis, IN). Secondary antibodies were diluted at 1:10,000 and were as follows: goat anti mouse IRDye 800CW (Li Cor926 32210) and goat anti rabbit IRDye 680LT (Li Cor 926 68021). Membranes were finally rinsed 3 times with 1x PBS with 0.1% Tween 20 and imaged on an Odyssey imaging system. Results Clarin 1 mRNA Isoform Specific Expression I nitial RT PCR experiments were performed u tilizing the previously published primers from Dr. Geller and colleagues in 2009 with intracellular actin as the positive control (Figure 3 (87) Figure 3 1 A shows the general schematic of primer
65 location. Figure 3 xpressed in each isoform of Clrn1 /C lrn 1 I first tested primers F1 R1 in exons 3 and 4 which pick s up both isoforms 1 and 2 (Figure 3 C) In C57BL/6J control mice there was a clear RT PCR product for Clrn1 at the expected size of 273 bp which would correlate to isoform 1 When CBA/J Clrn1 +/+ WT rd1 / KO mice were assessed, there was an RT PCR product at the expected size similar to the C57Bl/6J mice, indicating that isoform 1 is produced in the inner retina because the rd1 / strain of mice are c ompletely lacking PR cells at 2 months of age (Figure 3.1 C ) (43, 126, 177) When the CBA/J double Clrn1 / KO, rd1 / KO mice were assessed, they did not show any transcript present for the same PCR reaction, indicating that isoform 1 is absent in these mice ( Figure 3 1 C ) This is significant because Clrn1 / KO mice lacking PR cells d o not produce any isoform 1 Clrn1 mRNA from the inner retina, but Clrn1 WT mice without PR cells do. This indicates that isoform 1 is most likely expressed by the inner retinal cells. When I further tested isoform 2 specific primers F2 R4, there is only a product for the un injected C57BL /6J mice indicating that isoform 2 is most likely the isoform that is expressed in PR cells because it is not present in either the CBA/J Clrn1 +/+ WT rd1 / KO or CBA/J double Clrn1 / KO, rd1 / KO mice (Figure 3 1D) The right eyes of the C57BL/6J mi ce were injected subretinally with AA V2 Quad Y F sc smCBA h CLRN 1 HA ( Table 5 1 construct 1 ) and there did seem to be an increased amount of isoform 1 in the injected vs un injected eyes of C57BL/6J mice indicating that I am also detecting the injected construct as well as the endogenous Clrn1 transcript (Figure 3 1C) When I tested primers F2 R4, which should only pick up isoform 2, I only detect the expected fragment 300 bp in the un injected C57Bl/6J eyes,
66 but not in the AAV treated eyes (Figure 3 D) This indicates that isoform 2 is absent in both Clrn1 / ; rd1 / and Clrn +/+ ; rd1 / mice and that isoform 2 is most likely expressed by PR cells because it is absent from both Clrn1 WT and KO mice when on the rd1 ba ckground (Figure 3 D ) The only mice that show a significant presence of isoform 2 from this first group were the C57BL/6J contro ls with a correct band at 300 bp. I next tested the published primer s on the new A/J Clrn1 / ; Cdh23 / double KO and A/J Clrn1 +/+ ; Cdh23 / WT mice that were treated in one eye with subretinally with AAV2 Quad Y F sc smCBA h CLRN1 HA (Table 5 1 construct 1), ( Table 5 1 construct 1 ) With primers F3 R1 I detected the expression of isoform 1 at 326 bp in the WT A/J Clrn1 +/+ ; Cdh23 / mouse un injected retinas. A significant find ing is that when comparing the injected to the un injected KO A/J Clrn1 / ; Cdh23 / mice the untreated eyes do not have any isoform 1 expression at 326 bp, but the treated retinas show expression of isoform 1 with the F3 R1 primers as expected (Figure 3 F). This further supports the conclusion that isoform 2 is most likely expressed by PR cells while isoform 1, and perhaps 3, are expressed in the INL. Endogenous Clarin 1 Protein Localization U sing Immunohistochemistry For immunohistochemistry mice were sacrificed and paraffin samples were processed as described abov e. Antigen retrieval was optimized in C57BL/6J mice first and endogenous Clrn1 was detected using an anti human rabbit polyclonal CLRN1 primary antibody, Novus Biolgicals CLRN1 ( NB CLRN 1 ) I saw Clrn1 staining at the CC of PR cells in the retina (Figure 3 2 ) When I tested the WT A/J Clrn1 +/+ ; Cdh 23 / and KO A/J Clrn1 / ; Cdh 23 / mice I saw NB CLRN 1 staining at the CC of PR cells a nd some staining within the ONL in the WT mice (Figure 3 2 In contrast, a lthough the A/J Clrn1 / ; Cdh 23 / mice show some staining within the ONL, there
67 seems to be an absence of staining at the CC in PR cells. (Figu re 3 2 ) This is p romising given that there is a difference in Clrn1 / KO vs WT mice in the A/J; Cdh 23 / strain Comparison of C57Bl/6, A/J Clrn1 +/+ ; Cdh23 / and Clrn1 / ; Cdh 23 / mouse tissue indicates Clrn 1 is present, but localizes specifically to the CC alone only in C57Bl/6 mice. A/J Clrn1 +/+ ; Cdh 23 / mice that show localization in the CC and outer nuclear layer (ONL) and Clrn1 / ; Cdh 23 / mice show little expression Cryosections were processed as described above. In C57BL/6J mice, cryo preserved sections showed NB CLRN1 localization throughout the entire P R cell, concentrated at the IS and OPL, similar to th at in A/J Clrn1 +/+ ; Cdh 23 / mice in paraffin sections (Figure3 3 ) The difference in paraffin vs cryo staining is most likely due to the different treatment methods prior to staining. The paraffin processed samples undergo much harsher processing methods that the cryo processed samples because the tissues undergo a series of dehydration steps before being immersed in Xylene and embedded. The tissues are then re treated with xylene befo re several rehydration steps as mentioned above. This can act to mask certain antigens or alter some antigen specificity compared to the cryo processed tissue that are not exposed to these treatments. When analyzing NB CLRN1 staining in A/J Clrn1 +/+ ; Cdh 23 / N48K KI mice, NB CLRN1 localize d only to the distal tips of the OS of PR cells (Figure 3 4 ) In previous cell culture experiments the mutant N48K KI C lrn 1 fail ed to traffic to the plasma membr ane and (114) Due to the high levels of protein synthesis in PR cells, it is possible that N48K KI C lrn 1 is trafficked to the RPE cells for degradation rather than being degraded by the PR cell. In PR cells there are several mechanisms that regulate pr otein homeostasis including ubiquitination enzymes, heat shock proteins, and ER
68 stress response proteins (13) Interestingly, C lrn 1 has been shown to interact with several of these proteins including HSP70, HSPA5, calnexin, and cation independen t mannose 6 phosphate receptor that were identified by Tian et al (218) Since t hen there ha ve been no further studies looking at the functional interactions of C lrn 1 with any of the 54 identified interacting partners. It is possible that in the presence of the N48 K mutation, cell stress pathways stimulate protein degradation in order to maintain cellular homeostasis. Given the OS/RPE localization of N48K KI C lrn 1, this may be what stimulates retinal degeneration in patients over time because PR and RPE cells cannot keep up with the amount of muta nt protein being made in PRs I further validated the localization of NB CLRN1 at the CC of PR cells by co staining with both NB CLRN1 and gamma tubulin, a basal body marker which stain s adjacent to the CC in PR cell s. I saw staining of gamma tubulin with NB CLRN1 indicating that C lrn 1 is indeed l ocalized at the CC in PR cells (Figure 3 5 ) I further validated whether the NB CLRN1 antibody recognized C lrn 1 protein by colocalizing NB CLRN1 with injected AAV constructs expressing the 232 aa isoform of human Clrn1 cDNA C57BL/6J m ice were treated with AAV2 Quad Y F sc smCBA h C LRN 1 Venus at 1.54 x 10 12 vg/ml either subretinally or intravitreally (Table 5 1 construct 3 or 4) and sections were immunostained with NB CLRN1 and CLRN1 Venus (GFP analog). S ubretinally inj ected CLRN1 /CLRN1 localized at the RPE and PR cells and intravitreally injected CLRN 1 / CLRN1 localized throughout the whole retina as I ha d seen previously (Figure 3 6 ; Figure 5 1 A, B ) (72) For both injections, the AAV CLRN1 / CLRN1 protein colocalized with th e NB CLRN1 staining (Figure 3 6 ) NB CLRN1 also stained for endogenous C lrn 1 protein expression given the localization at the CC in the
69 intravitreally treated retinas as seen previously ( Figure 3 2 ; Figure 3 6 Overall, it appears that the NB CLRN1 antibody recogniz es Clrn1/ CLRN1 based on the c olocalization of AAV CLRN1 Venus with the NB CLRN1 antibody Endogenous Clarin 1 Protein Localization U sing Western Blot Tissue samples were processed and w estern blots were performed as de scribed above with C57BL/6J mice being used for initial NB C LRN 1 antibody optimization. For the initial western blot data, the expect ed 26 KDa band was seen for both the retina and eye cup dissections (Figure 3 ) This product is absent from the RPE/choroid dissection samples suggesting that this C lrn 1 protein product is only expressed in the retina These samples appear to be from clean dissections without retinal contamination as evidenced by the lack of arrestin 1 in these samples (Figure 3 7 A ) This confirms that the 26 kDa protein is only expressed in retinal tissue because it is present in the retina and whole eyecup samples but not the RPE/choroid samples. I further validated these dissec ted samples by co staining for arrestin 1 and r ho dopsin along with NB CLRN1 in both C57BL/6J and A/J WT mice (Figure 3 8 A, B) A rrestin 1 only appears in the retina l and eye cup dissections, suggesting they are pure dissections containing only retina tissue (Figure 3 8 A) In the secondary antibody only negative controls, there appears to be no c ross reactivity with the anti rabbit secondary that would recognize NB CLRN1. I do see cross reactivity for the ant i mouse secondary that recognize arrestin 1 and r ho dopsin (Figure 3 8 C) I believe that this is coming from the anti IgG epitopes present in the choroidal vasculature that are only present in the RPE/choroid and whole eye cup dissections since these bands only present in those sample lanes. I further tested the NB CLRN1 antibody in AAV expressed CLRN1 / CLRN1 v ector inject ed eyes vs un injected eyes A/J WT Clrn1 +/+ ; Cdh 23 / and KO Clrn1 / ; Cdh 23 /
70 mice were injected subretinal ly with AAV2 Q uad Y F sc smCBA h C LRN 1 HA ( Table 5.1 construct 1 ) When comparing the inject ed vs un injected eyecups the expected 26 KDa band was seen for all mice which would correspond to a monomer protein product (Figure 3 7 ) Additionally, I have clear co localization of NB CLRN1 with the anti HA antibody that recognize s this vector expressed CLRN1 protein (Figure 3 7 B ) Overall, it seems that the N B CLRN1 antibody does recognize Clrn1/ CLRN1 both as endogenous and an AAV expressed protein Given that there are additional bands that do not run at the expected size for C lrn 1 it is likely that there may be some non specific staining for other proteins in the retina. A potential way to decipher this would be to analyze these samples on a co o massie blue stained gel, excise out the protein f ragments and send both for mass spectrometry analysis in order to identify the exact prot ein(s) running in each band. Possible ways to test this hypothesis would be to generate a Clrn1 Crisper Cas9 KO mouse that w ould delete the entire Clrn1 locus and therefore should have no C lrn 1 expression from any isoform. Alternatively a Clrn1 /C lrn 1 KI m ouse with an N terminal or C terminal HA tag under endogenous promoter control of Clrn1 could be made This should allow for anti HA staining of endogenously expressed C lrn 1 and should therefore be able to clearly demonstrate the endogenous expression of C lrn 1 without any artifact from non specific targets with the NB CLRN1 antibody in immunohistochemistry and western blot analys e s.
71 Table 3 1. Previously p ublished Cl a r i n 1 RT PCR p rimers Primer Oligo Name Clrn1 F1 F o r wa rd GGTCCAAGCCATCCCCGTA Clrn1 F2 F o r wa rd TCATGCCAAGCCAGCAGAAGAAG Clrn1 F3 F o r wa rd AGGCAATGTGGGTTAGGAGCAAG Clrn1 R1 Rev erse TGTTCTGTAGGCATAGGTCCCTTC Clrn1 R4 Rev erse CTCTCCTTTGTCCTCATACAGAGAGTACC Clrn1 R5 Rev erse AGCCCCAGTGGTCCATGAAGAG Actin F1 F o r wa rd ACCAACTGGGACGACATGGAGAA Actin R1 Rev erse CATGGCTGGGGTGTTGAAGGT
72 Figure 3 1 RT PCR and Cl a r i n 1 isoform e xpression. A): Schematic of Clrn1 RT PCR with primer locations, arrows represent published primers (87) Clrn1 three main protein isoforms. C E ) CBA/J mice ( Clrn1 / rd1 / double KO vs Clrn1 +/+ ; rd1 / KO, left eye (L) vs right eye (R), C57 control, un injecte d (UI) vs injected (I) with AAV8 Y733F CBA h C LRN 1 HA subretinally C ) F1 R1 expected size for isoform 1 273 bp, isoform 3 should not be present. D ) F2 R4 should onl y pick up isoform 2 at approximat ely 300 bp. E) Actin control should be 165 bp F A/J mice ( Clrn1 / ; Cdh23 / double KO vs Clrn1 +/+ ; Cdh23 / KO, un injected (UI) vs injected (I) AAV2 smCBA h CLRN1 HA subretinally F ) F3 R1 expected size for isoform 1 326 bp is seen in both the injected retinas. e 165 bp
73 Figure 3 2 Novus Biologicals CLARIN 1 a ntibody s taining for i mmunohistochemistry NB CLRN 1 localizes to CC in A) C57Bl/6J B) Clrn +/+ Cdh23 / WT mice but not in C) Clrn1 / Cdh23 / KO mice 48 8 anti rabbit NB CLRN 1. 594 anti mouse rhodopsin. 488 594 DAPI.
74 Figure 3 3 Cryo p reserved C57BL/6J Novus Biologicals CLARIN 1 s taining magnification of C57B L/6J mice. NB CLRN1 is localized throughout the entire PR cell, p articularly at the IS and OPL. agnification of C57BL/6J mice. A, B) Merged goat 594 anti mouse rhodopsin goat 488 anti rabbit N B CLRN1, merged with DAPI 358. CLRN1 st ainin g alone goat 488 anti rabbit NB CLRN1.
75 Figure 3 4 Novus Biologicals CLARIN 1 l ocalization in WT vs Cl a r i n 1 KO and N48K KI A/J mice A) NB CLRN1 localizes at the distal tips of the OS in N48K KI mice, possibly within the RPE microvilli. Merged donkey 488 anti mouse rhodopsin donkey 594 ant i rabbit NB CLRN1. Compared to B) WT and C) Clrn1 / KO A/J mice.
76 Figure 3 5 Novus Biologicals CLARIN 1 c olocalization with gamma t ubulin A) NB CLRN1 localizes at CC with partial overlap of gamma tubul in (cili a basal body marker) in C57BL/6J mice. ) gamma tubulin localization at the basal bodies in PR cells. NB CLRN1 at the CC of PR cells. B) Reference image for NB CLRN1 localization at the CC in C57Bl/6 J mice A) Goa t 594 anti rabbit NB CLRN1, Goat 488 anti mouse gamma tubulin, DAPI 358 DAPI 358. Goat 488 anti mouse gamma tubulin. Goat 594 anti rabbit NB CLRN1. B) Donkey 488 anti rabbit NB CLRN1, Donkey 594 anti mouse rhodopsin DAPI 358.
77 Figure 3 6 Subretinal and i ntravitreal AAV injected C LARIN 1 c olocal ization with Novus Biologicals CLARIN 1 a ntibody. A) Subretinal injection AAV2 Quad Y F smCBA h C LRN 1 Venus (GFP analog) B) Intravitreal injection AAV2 Quad Y F smCBA h C LRN 1 Venus (GFP analog) 488 anti mouse GFP (Venus analog) Do nkey 594 anti rabbit NB CLRN1. merged with DAPI 358, complete c o localization is observed with GFP and some endogenous NB CLRN1 staining is observed in the connecting cilium of photoreceptors.
78 Figure 3 7 Novus Biologicals CLARIN 1 w estern b lot a nalysis A) Western blot analysis of different dissection methods staining for arrestin 1 and NB CLRN1. NB CLRN1 staining alone. R etina dissections show the expected C lrn 1 band at 26 kDa and arrestin 1 at 50 kDa. RPE/choroid dis sections have minimal to no arrestin 1 present at 50 kDa indication that they do not contain much if any retinal tissue. The RPE/choroid dissections also do not contain any C lrn 1 at the predicted 26 kDa size. Eye cup d issections containing both the r etina and RPE/choroid tissues have both the 26 kDa and 100 kDa bands further indicating that the 26 kDa band for C lrn 1 is only present in the retinal tissue and the larger 100 kDa fragment is comin g from the RPE/choroid tissue. B) West ern blot analysis of inject e d vs un injected retina eyecups. enhanced image of B with a longer exposure time Eyes were treated subretinally with AAV2 Q uad Y F sc smCBA h C LRN 1 HA and blots were stained for anti HA in green and anti CLRN1 in red.
79 Figure 3 8 Arrestin 1 r hodopsin, and secondary o nly c ontrol w estern b lots A) Arrestin 1 is only present in retinal samples. B) Rhodopsin is present in all samples, but the m ajority is in retinal samples. C) Secondary only controls show no staining for anti rabbit (CLRN1) an d IgG staining for anti mouse ( arrestin 1 r ho dopsin ).
80 CHAPTER 4 IDENTIFYING A RETINAL PHENOTYPE IN CLARIN 1 KNOCK OUT (KO) (Clrn1 / ) AND N48K KNOCK IN (KI) MICE 1 Background As stated previously, in dark adapted PR cells, the transducin is localized in the OS disc membranes and is associated with complex and arrestin 1 is lo calized to the rod PR IS and OPL synapse s. Upon exposure to light, transducin move s from the OS to the IS and OPL whereas arrestin 1 will move from the OPL and IS to the OS where it will bind to phosphorylated rhodopsi n (79) The majority of transducin will move from the OS to IS within 2 minutes of ligh t exposure a nd the majority of arrestin 1 will translocate to the OS wit hin 8 minutes of light exposure (37, 79) There is an ongoing debate as to whether these proteins travel through diffusion or through active transport with molecular motors along the actin and tubulin cytoskeletal filaments (36 39, 95, 97, 98, 162 164, 210) Dr Wolfrum as well as others, have assessed arrestin 1 and ability to utilize cytoskeletal filaments in a dark adapted vs light adapted retina. I n the light they believe arrestin 1 and t ra nsducin do not require fi laments ; however in the dark, arrestin 1 and t ra n s ducin travel along cytoskeletal filaments in order to return to their normal dark adapted state (39, 190 ) D r group previously show ed that the shaker1 (MYO7A) USH1B and the whirler (whirlin) USH2D mice have a delay in t ra nsducin translocation upon exposure to light (176, 219) MYO7A mice also display a slight ly attenuated ERG response for both a and b waves, but prior to this study no retinal degeneration had 1 Reprinted with permission from Di nculescu A and Stupay RM et al. AAV Mediated Clarin 1 Expression in the Mouse Retina: Implications for USH3A Gene Therapy. PLoS One 2016.11(2):e0148874.
81 been observed. T he level of light intensity required for transducin activation was much brighter than WT controls They further demonstrated that mice e xposed to contin uo us bright light of 2500 lux and mice housed in mid level light at an intensity of 1500 lux develop retinal degeneration in approximately 6 months; however they do not degenerate when the mice are housed at a low light intensity of 200 l ux (176) D r Cosgrove was also able to demonstrate that lentiviral delivery of WT MYO7A can rescue this transducin translocation phenotype as well as the li ght induced retinal degeneration (240) Similarly, they also showed a t ra nsducin translocation defect in whirlin mutant mice along with a requirement for brighter levels of light to induce complete transducin translocation W hirlin mice d evelop a retinal degeneration phenotype when exposed to 1500 lux for a continuous period of time and also have miss localized r ho dopsin protein in the IS of PR cells (219) Given this translocation phenoty pe in other USH mouse models, I hypothesized that there may be a sim ilar translocation phenotype, retinal degeneration after bright light exposure, attenuated ERG, or a potential additional ERG phenotypes in my USH3A mouse models. As noted previously, a primary complication with studying USH3A is that the current mouse models do not show a retinal phenotype but do present with significant hearing loss and inner hair cell death by 4 months of age (87 90, 93) An additional complicatio n with the original Clrn1 / KO and N48K KI mouse models is that each was generated on different genetic backgrounds which means that any phenotype identified may have confounding variables due to the genetic background of the mouse strain. Additionally, the original Clrn1 / KO mice are albinos and the N48K KI mice are pigmented adding another layer of complexity to comparing any phenotype of
82 translocation and/or light induced degeneration between Clrn1 / KO and N48K KI mice. Particularly with the hypo thesis of a light exposure phenotype, pigmented mice are traditionally less sensitive to light and require much higher levels of light to induce degeneration compared to albinos. This is one benefit to switching to the A/J strain. One c omplication is that the A/J mice develop more pronounced age related retina l degeneration than other strains. This PR cell death is due to an upregulation of inflammatory factors and a decrease in neuroprotective factors and these mice show changes in their RPE prior to PR ce ll death (160) A/J mice have significant thinning of the ONL and a noticeable decline in cone PR number at 8 months old Consequentially there was a decrease in scotopic a nd photopic ERGs at 8 months of age. This PR cell death is indep endent of light exposure since WT A/ J mice raised in the dark also und ergo PR degeneration (160) Upon fundus exa m, A/J m ice have an increase in auto fluorescence at 8 months of age and hav e evidence of inflammation and immune cell activation originating from the RPE. Additionally, 8 month old mice show an increase in RPE cell size and multinucleated cells and they also fai l to regenerate 11 cis retinal over time (160) Additionally, this strain of mice also carr y other background mutations that cause the mice to age at an accelerat ed rate and they have a higher incidence of cancer and mitochondrial de fects These mice also have RPE defects and undergo retinal degeneration over time with an attenuated ERG at 12 months of age (160) Methods Arrestin 1 and Transducin Translocation Assay Mice from both the C57Bl/6J mixed a lbino and the A/J strain described above were housed as described above (87, 88, 217) All mice were housed in a 12 hour dim light/dark cycle prior to translocation experiments. Mice were dilated with Atropine
83 Sulfate Ophthalmic Solution 1% (Akorn Inc, Lake Forest IL USA, NDC 17478 215 05) a long acting dilator prior to be ing dark adapted overnight. The following day mice were dilated again with Phenylephrine Hydrochloride Ophthalmic Solution 2.5% (Paragon BioTeck Inc, Portland OR, USA, NDC 42702 102.15) prior to translocation experiments. One mouse of each genotype was mov ed into a clear plastic box and exposed to 1000 lux of light for 1 hour and the mice were then euthanized as stated previously. Eyes were enucleated and pl aced in a 4% PFA fixative solution for 1 hour prior to moving to a 1x PBS solution All eyes were processed for paraffin sections as described above Immunostaining was performed as described previously with the exception s of not baking the slides at 60 o C prior to staining and having a Protinase K digestion step Primary antibodies we re a s follows: mouse anti arrestin 1 1:100 dilution (courtesy of Dr. WC Smith, C10C10) (73) mouse anti r hodopsin 1:100 dilution (courtesy of Dr. WC Smith, B6 30) (2) and rabbit anti t ransducin 1:1000 dilution (rabbit polyclonal Santa Cruz Biotechnology sc 389). Secondary antibodie s we re as follows at 1:400 dilution: goat anti rabbit 488 (A11008), goat anti rabbit 594 (A11012), goat anti m ouse 488 (A32723), goat anti mouse 594 (A11005), donkey anti rabbit 488 (A21206), donkey anti rabbit 594 (A21207), donkey anti mouse 488 (A21202), and donkey anti mouse 594 (A21203). All secondary antibodie s were acquired from Molecular Probes/ Invitrogen, Eugene, OR Slides were mounted and imaged and analyzed as described above using a spinning disc confocal microscope through the Cell and Tissue Analysis Core at the University of Florida ( Olympus DSU IX81) Images were acquired with 10 15 confocal z stacks and images were consecutively de convolved and merged into a singl e projection image and each color channel was saved separately and merged.
84 Fluorescence intensity plots wer e made with ImageJ software (National Institutes of Health) Densitometric plots were obtained across the PR cell layer and the Plot Profile program was used to calculate arrestin 1 and t ra nsducin staining intensity across the PR cell layer. Nuclear staining wi th DAPI defined the ONL For this analysis three random regions acros s the retina were selected for WT, KO and KI mice. The fluorescent signal in the OPL was defined as a percentage of to tal fluorescence across the entire PR cell layer. Th e statistical difference was calculated via an unpaired t test using GraphPad softwar e and error bar s calculated as the mean standard error of mean (SEM) and was considered to be statistically significant at p<0.05 or greater. GFAP Expression and Olfaction Assays to Test for Alternative Phenotypes Immunohistochemistry was perf ormed on paraffin retinal sections to look for any upregulation in glial fibrillary acidic protein (GFAP) expression in WT vs KO mice. S lides were mounted and imaged as described above using KO mice from the C57BL/6J mixed albino strain. GFAP is an interme diate filament protein that is upregulated in M ller glial cells in response to retinal injury or degeneration (77) Eye cups were processed and stained as described above with an anti GFAP anti body ( Thermo Fisher mouse antibody ASTRO6 #MA5 12023) and Vecta Shield with DAPI. For olfactory assays, mice were anesthetized and sacrificed and o lfactory and r espiratory epithelium and olfactory bulb samples were processed in collaboration with Dr. Jeffrey Marte ns group at UF. Experimental protocols are described in their previous studies (153, 232) Olfactory lysates were obtained from WT A/J and N48K KI mice and RT PCR analysis was done with the previously pub lished primers (87) Samples were dissected and RT PCR was performed as described above.
85 Electroretinography For the LKC system, an electrode wa s placed on the surface of the cornea and a reference electrode wa s placed in the skin at the surface of the skull between the ears. A third grounding electrode wa s placed in the tail to establish a baseline measurement. The mice we re set up on a mobile platform then moved into an upright G anzfeld dome that generate d a unifor m light field for the procedure (EM Win UTAS LKC Tec hnologies Gaithersburg MD ) For the Espion system, a contact electrode wa s placed on the surface of the eye, the reference electrode wa s placed in the cheek of the mouse, and the ground electrode wa s placed in the tail. The mice we re setup on a stat ionary platform and a moveable G anzfeld dome wa s lowered on top of the mice prior to the procedure (40) Both systems contain a platform heat ing pad to keep the mice at a stable body temperature during the experiment. Both systems were utilized through the course of this work, but un less otherwise noted, only the LKC data is presented here. For scotopic ERG, mice we re dark adapted overnight for 12 h ours prior to ERG. The mice we re dilated the day before using Atropine Sulfate Ophthalmic Solution 1% (Akorn Inc, Lake Forest IL USA, NDC 17478 215 05) a long acting dilator. Prior to ERG the mice we re then dilated again using Phenylephrine Hydrochlor ide Ophthalmic Solution 2.5% (Paragon BioTeck Inc, Portland OR, USA, NDC 42702 102.15) Mice we re then anesthetized using a mixture of ketamine (72 mg/kg) (Ketaset: Ketamine HCl injection, USP, Zoetis Inc, Kalamazoo MI, USA, NADA 043 304) and xylazine (4 m g/kg) (LLOYD Shenandoah IA USA, NADA 139 236) by intraperitoneal injection with a volume of 4 ul per gram of weig ht. Then the mice we re setup on the platform and, prior to electrode placement, either G onak Hypermellose Opt halmic Demulcent Solution 2.5% ( Ak orn Inc, Lake Forest IL, USA NDC 17478 064 12 ), Serile eye wash (Altaire
86 Pharmaceuticals Inc, Aquebouge NY USA, NDC 59390 175 35), or GenTeal Lubricant Eye Drops (Alcon Laboratories Inc, Fort Worth TX USA NDC 0078 0518 16) wa s applied to the surface of the cornea to keep the eyes hydrated thro ughout the procedure. The mice we re then placed in the G anzfeld dome and exposed to 3 light intensiti es of 20 dB, 10 d B, and 0 dB (corresponding to 0.02, 0.2 and 2 scot cdsecm 2 stimuli). After ERG the mi ce we re pr ovided anesthesia reversal using a solution of Antisedan Orion Pharma Corporation Fi nland, Zoetis Inc, Kalamazoo MI USA NADA 141 033) (63, 72, 175, 184) Scotopic a wave, b wave, and OPs were compared between control C57BL/6J, A/J WT, and KO mice to validate the previously published ERG phenotype. The a wave was measured from the highest amplit ude immediately after the light stimulus to the lowest amplitude afterward prior to the increase in amplitude for the b wave. The b wave was measured from the lowest amplitude used for the a wave to the highest amplitude after the OPs. The OPs are the osci llations along the increasing amplitude of the b wave and there can be a range of up to 5 OPs. OP1 is measured from the lowest amplitude used for the a wave to the next highest amplitude prior to the next trough OP2 begins at the first trough to the next hig h est amplitude. OP3 is measured similarly from the next trough to the next highest peak. For a photopic ERG, mice we re first light adapted, and m ice we re prepared as described above In order to select the optimal time, mice we re light adapted in the G anzfeld dome for 2, 4, 6, 8, and 10 minutes and then the highest light intensity for a cone response was tested to determine a plateau of maximal response and then all intensities were recorded. Cone response is recorded at 4 light intensities of 3 dB, 3 dB, 6 dB, and 10 dB (25 phot cdsecm 2 maximum intensity). After the procedure m ice
87 we re provided anesthesia reversal as described above (63, 72, 175, 184) Photopic cone ER Gs were compared between C57BL/6J, A/J WT, and KO mice. For the dual flash ERG recovery response s mice were dark adapted overnight, then fully bleached to suppress rod responses using the Espion ERG system, then a scotopic ERG was recorded at 1 cd.s/m 2 which is equivalent to the 0 Db intensity on the LKC system. After the initial flash post light adaptation, the mice were tested at varying time points post light exposure to measure how rapidly the mice were able to recover a scotopic ERG response This protoc ol is defined in Table 4 1 and was only assessed on the initial C57BL/6J mixed albino mice. Due to the amount of time required to set up the mice on the Espion system compared to the LKC s y stem I was unable to reproduce the protocol on the new A/J s train because the A/J mice are extremely sensitive to the anesthesia and cannot be anesthetized for the 1 hour required for the experiment For ERG analysis post 1000 lux light adapt ion mice were dark adapted overnight, exposed to 1000 lux of light for 1 hour, and then a scotopic ERG was measured immediately after light exposure. Maximum scotopic a and b wave ERG recordings were analyzed for WT, Clrn1 / KO and N48K KI A/J strain mice, the amplitudes were averaged between left and right eyes from the sam e mouse and then averaged together for each genotype. The results were analyzed by a Student t test using GraphPad statistical software (GraphPad Prism 6.0, GraphP ad Software, San Diego, CA). ERG responses were considered statistically significant at p< 0.05 *, <0.01 **, and <0.001 *** and a ll ERG data are re presented as the mean SEM Spectral Domain Optical Coherence Tomography SD OCT was used to assay for retinal degeneration post light damage as follows Mice were anesthetized and eyes were dilated and kept hydrated during the
88 procedure as described above. Mice were arranged on the Bioptigen platform and focusing scans were used to align the retina with the optic nerve at the center of the image. Two mor e scans were obtained peripheral to the optic nerve, both nasally and temporally. SD OCT parameters for all scans were the same for all mice. Rectangular volume scans were obtained with an A/B scan ratio of 1000, 5 frames per B scan at 100 B scans and 80 l ines of inactive A scans per B scan ratio taken at 1 volume with a square field of 1.4 mm length and 1.4 mm width and a horizontal offset of 0.1 mm. After the procedure, mice were woken up as described above (19) Images were analyzed with InVivo Vue Bioptigen software and ONL thickness was measured using Bioptigen Diver software with a 9x9 spider plot. ONL thickness was measured both manually and using the auto segmenting program, and values were compared. The ONL thickness was measured and calculated using Diver software calipers through the Bioptigen OCT system placed at the OPL synapses and the OLM at the top of the ONL and plotted in GraphPad software and a two way ANOVA (Analysis of Variance) test was performed to assess whether changes in ONL thickness was statistically significant. Light Damage Mice were housed in the animal facility as described a bove. Prior to light damage mice were analyzed by ERG and SD OCT as described above for a light damage control. During light damage experiments, the previously p ublished protocols were adapted (99, 223, 224) The mice were dark adapted the day before and two mice were placed in each cage, one of each genotype, with minimal bedding and food placed on the bottom of the cage to allow for minimal blockage of light. Mouse cages were housed on a standard ACS (animal care services) rack with LED light s emitting 5500K of light, arranged und er the cage lids. The se lights are adjustable to achieve the desired light
89 intensity Light intensity was calibrated using a traceable dual range light meter (Sper Scientific 840006 Light Meter Lux). Timing of light exposure was also controlled by an elect ric timer. M ice were placed in the rack at 6 pm and light damage was performed for 4 hours until 10 pm at specified light intensities. One week post light damage mice were an alyzed using SD OCT and ERG, and the ONL thickness and scotopic ERG responses were measured, calculated, and analyzed as described above For the initial A/J light intensity optimization, light intensities of 1300, 1500, 2000, or 5000 lux were used to optimize the light damage conditions for the maximum ONL degeneration in WT mice. At all light intensities tested, SD OCT was performed 1 week post light damage, ONL thickness was analyzed using a two way ANOVA (Analysis of Variance) to assess if changes in ONL thickness was statistically significant after light damage. Scotopic ERG respon ses were assessed for all mic e pre and post light damage at all intensities a nd a and b wave amplitudes were measured as described above. A Student T test was performed and the ERG response was considered statistical ly significan t at P<0.05 *, <0.01 ** and <0.001 ***. The 2000 lux intensity was chosen to assess for light damage in the Clrn1 / KO and N48K KI mice and light damage was performed as stated above. SD OCT and ERG were performed pre and post light damage to measure ONL thickness and retina l function as mentioned above. Results Light Driven Protein Translocation to Characterize a Retinal Phenotype Upon dark adaptation both the mixed albino C57BL/6J WT and Clrn1 / KO mice have proper localization of t ra nsducin in the OS and arrestin 1 in the IS, ONL and OPL (Figure 4 1) Upon exposure to 1000 lux of light for 1 hour, both the WT and Clrn1 / KO mice have normal transducin movement (Figure 4 F luorescence quantification
90 is plotted in Figure 4 after 1 hour of light exposure at 1000 lux both the Clrn1 / KO and N48K KI mice show a significant delay in arrestin 1 translocation from the OPL and IS to the OS and this delay is statistically significant for both the Clrn1 / KO and N48K KI mice on the mixed C57BL/6 J albino mice (Figure 4 2 ) This is in contrast to USH1B and USH2D mou se models that have a delay in t ra nsducin translocation rather than arrestin 1 (176, 219, 220) This tran slocation phenotype was also reproducible in the new A/ J Clrn1 / KO and N48K KI mice compared to WT It is unclear what the b iological significance of this arrestin 1 delay is, given that there is no PR degeneration or loss of ERG response o ver time Pre vious studies hav e looked at arrestin 1 KO mice and they see a prolonged photorespo nse in mice completely lacking arrestin 1 These mice also do not have ONL degeneration, but they do have slightly disorganized OS s and a hi ghly altered recovery phase of r h o dopsin (235) I therefore also tested for a similar delay in rod PR cell recovery response but there was no significant difference between the WT and N48K KI mice (Figure 4 10) GFAP Expression in WT vs KO Mice As mentioned above, GFAP is an intermediate filament protein that is upregulated by Muller glial cells in response to retinal injury or degeneration (77) When comparing the mixed albino C57BL/6 J WT and Clrn1 / KO mice there did appear to be some slight upregulation in some Clrn1 / KO mice however this was not consistent across a larger pool of samples (Figure 4 3 ) T he samples were all processed ac cording to the same protocol at the same ti me. This was shown in other previous studies (87) Ol factory Structure and Clarin1 Expression in Ol fac tory Cells A collaboration was arranged with Clrn1 /Clrn1 expression in olfactory tissue. As noted above, olfactory epithelium,
91 respiratory epithelium, and olfactory bulb tissue was acquired, sectioned, and stained in addition to mRNA isolation from olfactory tissue for RT PCR analysis. First I compared the cilia formation in both olfactory a nd respiratory epithelium looking at acetylated tubulin and there did not appear to be any di fferences in either tissue from WT or N48K KI mice. I next looked at whether there was normal olfactory signaling present in the WT vs N48K KI A/J mice by assessing the presence and amount of tyrosine hydroxylase activity. There appeared to be normal local ization in th e juxtaglomerular cells that innervate the olfactory glomeruli (Figure 4 4 Overall there did not appear to be any olfactory phenotype, which was disappointing because previous studies have looked at the other USH proteins in olf action both in animal models and patients and they saw significant decrease in olfactory signaling as well as decreased beat frequency in nasal cilia (12, 120, 196) Next I looked at transcript analysis of Clrn1 in olfactory lysates. Uti lizing the previously published primers that we re used above, I saw isoform specific products for isoform 2 with primers F2 R4 indicating this isoform is expressed in olfactory epithelium. I also saw much fainter products for F2 R5 which detects isoforms 1 and 2, as well as a product for F3 R1 which detects all three isoforms (Figure 4 5 A) Overall this indicate s that Clrn1 is expressed in olfactory tissue and correlate s with other USH proteins being expressed there as well Validation of the Previously Published N ovel ERG Phenotype First I compared the A/J strain ERGs for the WT and Clrn1 / KO mice to my previous control C57BL/6J mice in order to understand any differences that may hinder future studies. Like the previously published data on the A/J strain (160) the A/J mice do have significantly smaller a and b wave ERG amplitudes compared to the C57BL/6J mice. It appears that the cones are most affected in the A/J strain because all light
92 intensities for the pure cone response were all significantly reduced and the only a and b wave intensity that was significantly decreased was at the 0 Db flash intensity which is the brightest rod response condition and does detect some cone response (Figure 4 7 A, B) Furthermore, the A/J cone ERGs were significantly decreased at all three intensities co mpared to the C57BL/6J controls (Figure 4 6 C) I also examined OPs in the A/J mice I n the pure rod response ERG there is a difference betwe en the A/J WT and Clrn1 / KO mice (Figure 4 6 D) This may be significant because the OPs are a result of the inner retinal response after synaptic transmission from PR cells. There is no conclusive agreement in the published data that supports which cel l types in the inner retina are producing the OPs, but they are generally believe d to origin ate from either amacrine cells, horizontal cells, or the on/off rod cone bipolar cell pathways (209) If OPs do in fact originate in amacrine cells, this would be s ignificant because they are reduced in the Clrn1 / KO mice I further confirmed the ERG difference previously published in the WT vs Clrn1 / KO A/J mice (217) A s described abo ve, Dr. Imanishi generated a Clrn1 / KO and N48K KI strain on an A/J background that possess an aging mutation allowing for more rapid retinal degeneration over time. The goal was to identify an additional retinal phenotype for the Clrn1 / KO and N48K KI mice. He show ed that at 3 and 6 months old the Clrn1 / KO mice have a reduced ERG by approximately 30%, however by 9 months of age the WT mice also have a reduced ERG response similar to Clrn1 / KO levels but there did not appear to be any differenc e in the WT vs N48K KI mice (217) I validated this difference in ERG amplitude s for the WT vs Clrn1 / KO mice and saw a similar trend however the KI mice did not have a significant decrease compared to WT at old er ages (Figure 4 7 )
93 Abnormal Retained Electroretinogram in N48K Mice Given the delay in arrestin 1 movement upon exposure to light and the reduced OPs potentially due to p roblems in the inner retina, I test ed for any physiological significance that could be observed i n vivo I perform ed scotopic ERGs on WT, Clrn1 / KO, and N48K KI A/J mice after the translocation experiment with 1 hour o f light exposure at 1000 lux. I found that at the brightest scot opic response of 0 Db, there was a significant difference in that the N48K KI mice still produce d a b wave ERG response (Figure 4 8 ) I am unsure as to what this response is telling us about a potential mechanism of retinal degeneration over time and how it is relevant to the biological and physiological processes that are occurring. There are a few potential theories that may be playing a role. 1 : The N48K KI mice fail to completely suppress r ho dopsin signaling as efficiently as WT mice and they the refore still respond to light, w hi ch makes sense in terms of the arrestin 1 delay phenotype 2 : N48K KI mice have prolonged synaptic signaling, whic h also could be related to the arrestin 1 phenotype with it remaining bound to the OPL. 3 : N48K KI mice are able to recover r ho dopsin signaling more rapidly than WT mice. 4 : N48K KI mouse cones respond to lower levels of light compared to WT mice so they are more responsive in the mesopic light range. And finally 5 : N48K KI mice have abnormal commu nication between the on/off rod cone bipolar cell signaling processes or their feedback mechanisms with horizontal cells. Given the complex nature of the cell types in the INL, it is difficult to determine which potential pathway is causative and it very well may be a combination of several mechanisms. I have considered possibly using synaptic inhibitors to block any potential recovery or collaborating with groups that can perform single cell recordings e x vivo
94 One other method I did attempt is a dual flash ERG recovery response, which would possibly show a more detailed difference in ERG recovery over time. Dual Flash ERG Recovery Response For the dual flash ERG recovery, the flash protocol is defined in Table 4 1 as well as described above. I ass es sed WT vs N48K KI mixed C57BL/6J albino mice at time points up to 45 minutes and I did not observe any statistical significance between the WT and KI mice (Figure 4 9 ). This would perhaps eliminate any dim cone response since the light intensity used was a mesopic mixed rod cone intensity, however this was performed on the Espion ERG system and not the LKC system as for the previous translocation ERG experiment and this may introduce some variability between the two experiments The lack of a dual flash E RG recovery phenotype may also eliminate any abnormal on/off rod cone bipolar cell signaling because the rods were completely bleached after the extended light exposure, and there was no significant difference in the rods ability to slo wly respond to light over time. I was unable to test this on the new A/J strain mice because they are highly sensitive to the anesthesia and are unable to remain anesthetized for the required 1 hour in order to run the experiment. Light Damage in WT, KO, and KI Mice Because I found a delay in arrestin 1 translocation upon light exposure, I wished to see if I could induce retinal degeneration after bright light exposure t hrough the use of light damage ( 223, 224 ) In order to test if my mice were susceptible to light damage, I f irst assessed WT mice at a ran ge of light intensities from 13 00, 1500, 2000, and 5000 lux as described above At all light intensities tested using a two way AN OVA there was significant degeneration post light damage in the WT mice (Figure 4 1 0 A Table 4 2 ) The most significant damage occurred at 5000 lux, which was expected (Figure 4 1 0 B
95 Table 4 2 ) At 2000 lux the WT mice did have statistically significant degeneration, particular ly in the superior retina, however, they still had a normal scotopic ERG response for both the a and b waves and only had a significantly reduced ERG at 5000 lux (Figure 4 1 0 Table 4 2 ). I therefore chose to analyze the Clrn1 / KO and N48K KI mice at 2000 lux for ONL thickness and scotopic ERG response. After light damage, only the s uperior retina showed slight degeneration in the Clrn1 / KO mice and the N48K KI mice had significant degeneration furthest from the optic nerve both in the superior and inferior retina (4 1 1 A B ) ( Table 4 3 ) Interestingl y, similarly to the WT mice, the Clrn1 / KO mice did not show any reduction in scotopic ERG amplitude post light damage for either the a or b waves (Figure 4 1 1 C, C This may not be the most reliable results b ecause the Clrn1 / KO mice already have a reduced b wave at 6 months old and they would therefore not necessarily show any significant diff erence since their ERG responses are already reduced without an treatment (Figure 4 1 1 C When I assayed for light damage on the N48K KI mice, I saw a significant decrease in scotopic ERG amplitudes for both the a and b waves for all light intensities (Figure 4 1 1 D, D This is intriguing because unlike the WT and Clrn1 / KO mice, the N48K KI mice are the only ones that display any significant loss of function following light damage. This can provide us with an additional phenotype to utilize in order to validate our AAV gene therapy vectors for an e ffective therapeutic treatment. Given there is a significant decrease in N48K KI ERG after light damage and a maintained ERG after bright light exposure, there may be a n underlying mechanism that is a ffec ted by both light conditions. T he combination of the se two experiments may be able to explain why patients do not lose vision initially, but go blind over a prolonged
96 per iod of time. This suggests the rate of vision loss may be related to the amount and level of light exposure over time. Perhaps initially PR cells can handle a minimal amount of light induced stress, but there is a limit ab ove which continued stress to the PR RPE visual system leads to PR degeneration. I can further test the Clrn1 / KO and N48K KI mice for pre conditioning prior to light damage which would be to expose the mice to a low level of light for one week prior to light damage to see if the loss of ONL thickness and ERG function can be reduced. This would allow me to assess whether the N48K KI mice have any abnormal neuroprotective features compared to the WT and Clrn1 / KO mice because they have a significantly reduced ERG response after light damage One complication with the A/J mouse strain is that they were shown to have abnormal RPE function and undergo r etinal degeneration as they age. Therefore the strain itself may possi bly be more susceptible to light damage due to background RPE complications The A/J mice have extensive inflammatory changes as they age with an increase of immune cell infiltrates in the retina which can have an impact on the rate of retinal degeneration due to light damage RPE flat mounts that the majority of RPE cells were multi nucleated with an abnormally large cell size indicating that the RPE cells were unhealthy and undergoing senescence (160).
97 Table 4 1. Dual f lash ERG r ecovery r esponse Steps Flash Intensity (cd.s/m 2 ) Interval Between Two Flashes Adaptation Time 1 1 0 ms 30 s 2 1 25 ms 30 s 3 1 50 ms 30 s 4 1 100 ms 30 s 5 1 250 ms 30 s 6 1 500 ms 30 s 7 1 750 ms 30 s 8 1 1 s 30 s 9 1 2 s 30 s 10 1 4 s 30 s 11 1 16 s 30 s N = 5 at 3 months old for all genotypes
98 Table 4 2. Two way ANOVA for WT light damage range of intensities Distance from Light Damage Brightness Intensity Optic Nerve (um) 1300 lux 1500 lux 2000 lux 5000 lux 600 um ns p < 0.05 ns p < 0.0001 **** 450 um ns ns ns p < 0.0001 **** 300 um ns p < 0.01 ** p < 0.05 p < 0.0001 **** 150 um p < 0.05 p < 0.0001 **** p < 0.0001 **** p < 0.0001 **** 0 um ns ns ns ns 150 um p < 0.05 p < 0.0001 **** p < 0.0001 **** p < 0.0001 **** 300 um p < 0.05 p < 0.01 ** p < 0.0001 **** p < 0.0001 **** 450 um ns p < 0.01 ** p < 0.0001 **** p < 0.0001 **** 600 um p < 0.001 *** p < 0.0001 **** p < 0.0001 **** p < 0.0001 **** N = 4 for all light intensities
99 Table 4 3. Two way ANOVA in Clarin 1 KO and N48K KI mice, 2000 lux light damage Distance from Optic Nerve (um) Clrn1 / KO N48K KI 600 um ns p < 0.001 *** 450 um ns p < 0.0001 **** 300 um ns p < 0.001 *** 150 um ns ns 0 um ns ns 150 um ns ns 300 um p < 0.05 p < 0.01 ** 450 um p < 0.001 *** p < 0.01 ** 600 um ns p < 0.0001 **** N = 10 for all light intensities
100 Figure 4 1. Normal l ocalization of a rrestin 1 and t ransducin. A, adapted localization of arrestin 1 and t ra nsducin in WT and Clrn1 / KO A/J mice. Goat 488 anti rabbit t r a nsducin and Goat 594 anti mouse arrestin 1 Fluorescence intensity plots of t ra nsducin in WT vs Clrn1 / KO A/J mice after 1 hour light exposure at 1000 lux. Goat 488 anti rabbit t ra nsducin and DAPI 358 VectaS hield. C of light adapted t ra nsducin localization in WT vs Clrn1 / KO A/ J mice. Goat 488 anti rabbit t ra nsducin and DAPI 358 vectashield.
101 Figure 4 2 Immunofluorescence and q uantification of a rrestin 1 in light a dapted Clarin 1 KO and WT retinal s ections signal intensity scanned across the photoreceptor layer. There is a significant amount of increased arrestin 1 signal in the OPL of Clrn1 / compared to WT OPL (A). Immunohistochemistry of a rr estin 1 distribution after light exposure translocation. Nucl ei are stained blue with DAPI. B) WT immunohistochemistry Clrn1 / KO immunohistochemistry C) Bar graph showing the relative signal intensi tie s of arrestin 1 in the OPL of Clrn1 / KO and WT mice, expressed as a percentage of the total signal intensity in the photoreceptor layer, shown for both Clrn1 / KO and N48K KI mice ( **p<0.01) relative signal intensities of arrestin 1 in the OPL of N48K KI and WT mice, expressed as a percentage of the total signal intensity in the photoreceptor layer, shown for both Clrn1 / KO and N48K KI mice (**p<0.01). Similar staining was seen for both the mixed C57BL/6J albino strain a nd the A/J strain.
102 Figure 4 3 GFAP e xpression in WT vs Clarin 1 KO C57BL/6J mixed a lbino m ice. A) Representative WT staining of GFAP. The staining appears to have relatively minimal GFAP expression. B) Representative Clrn1 / KO staining of GFAP. The staining appears to be relatively similar to the WT mice with perhaps a slight upregulation of GFAP expression, but it is not significantly higher across multiple samples.
103 Figure 4 4 Normal o lfac tory s tructure in b oth WT and N48K KI A/J mice A) WT olfacto ry and respiratory epithelium. N48K KI olfactory and respiratory epithelium. Disruption of CLRN 1 does not affect cilia formation in the nasal cavity. Representative confocal images of olfactory and respiratory epithelium from 7 month old con trol and mutant mice stained with the ciliary axoneme B) WT olfactory signaling. N48K K I olf actory signaling. There is normal active olfactory signaling in CLRN1 WT and N48K KI mice. Representative confocal im ages of olfactory bulb glomeruli from WT and N48K KI mice stained with tyrosine hydroxylase. Tyrosine hydroxylase is present in juxtaglomerular cells innervati ng glomer ul i of control and CLRN 1 mutants, indicative of active olfactory signaling. Scale,
104 Figure 4 5 RT PCR of WT A/J ol factory epithelial tissue l ysates A) There is a PCR product for all three primer sets from o lfactory epithelial lysates. F2 R4 can pick up the 250 AA isoform of Clrn1 F2 R5 can pick up isoform 1 and 2, and F3 R1 can pick up all three isoforms.
105 Figure 4 6 ERG d ifference in C57BL/6J vs WT and Clarin 1 KO A/J m ice. White bars are C57BL/6J mice, black bars are WT A/J mice, and red bars are Clrn1 / KO A/J mic e. A) a wave ERG, 0 Db is significantly decreased in A/J mice. B) b wave ERG, 0 Db is significantly decreased in A/J mice. C) Cone ERG, all light intensities are decreased in A/J vs C57BL/6J mice and 0 Db is decreased in A/J WT vs Clrn1 / KO mice. D) Oscillatory potentials in C57BL/6J vs A/J mice. Only the 0 Db OP1 is decreased in C57BL/6J vs A/J mice and all light intensities have reduced OPs in WT vs Clrn1 / KO A/J mice.
106 Figure 4 7 Validation of the ERG p henotype in A/J Clarin 1 KO m ice. A) WT vs Clrn1 / KO normal scotopic a wave ERG response. B) WT vs Clrn1 / KO ab normal scotopic b wave ERG response. C) WT vs N48K KI normal scotopic a wave ERG response. D) WT vs N48K KI normal scotopic b wave ERG response.
107 Figure 4 8 ERG p henotype post 1 hour light exposure in the N48K KI m ice. After 1 hour of light exposure at 1000 lux, the N48K KI mice are still able to generate an ERG response at a 0 Db light intensity.
108 Figure 4 9 Dual f lash ERG r ecovery response in the N48K KI m ic e. a and b wave recovery ERG response after light exposure. There is no significant difference in either a or b wave ERG between WT and N48K KI mice.
109 Figure 4 10 Light d amage o ptimization in WT A/J m ice. A) Comparison of light damage in A/J WT mice at varying light intensities using SD OCT. There is minimal difference in loss of PR cells between untreated controls, 1300, 1500, or 2000 lux intensitie s of light. The only extreme difference is between WT unt reated vs 500 0 lux of light. B) Comparison of untreated control vs 2000 lux treated WT mice. C) WT scotopic a wave wave ERG after 2000 lux light damage there is no significant decrease in ERG amplitude after 4 hours of light exposure at 2000 lux D) WT scotopic a wave ERG after 5000 lux light damage, there is a significant decrease in ERG amplitude after 4 hours of light exposure at 5000 lux.
110 Figure 4 1 1 SD OCT and ERG in Clarin 1 KO vs N48K KI A/J m ice p re vs post l ight d amage A) Clrn1 / KO and B) N48K KI ONL thickness pre and post light damage at 2000 lux. There is a significant decrease in both genotypes post light damage, mostly in the superior retina for the Clrn1 / KO mice and farthest from the optic nerve in both the superior and inferior retina in the Clrn1 / KO a and b wave ERG. There is no significant difference between pre vs post light damage N48K KI a wave ERG. There is a signi ficant difference between pre vs post light damage for 10 N48K KI b wave ERG. There is a significant d ifference between pre vs post light damage for all scotopic light intensities.
111 CHAPTER 5 ASSAYING FOR PHENOTYPIC RESCUE USING AAV MEDIATED GENE THERAPY 2 Background In order to develop a successful gene therapy treatment, the serot ype, capsid mutation(s) promoter, and cDNA, targeted to the desired cell type need to be optimized in order to avoid toxicity and achieve optimal expression levels This can be highly variable based on the promoter, gene of interest, and fluorescent tags with in the AAV plasmid. There is further variability of many different AAV capsid serotypes and additi onal capsid mutations allow ing for greater ce ll specificity and transduction efficiency in the retina All of these are elements that need to be considered in order to design a successful gene therapy treatment for USH3A Previously, D r Cosgrove showed delay s in transducin movement upon exposure to light in Myo7A and whirlin mice compared to WT controls and t hey further show ed that lentiviral delivery of WT MYO7A can rescue the transducin translocation phenotype (176, 219, 240) Given that my data is similar to D r I hypothesized th at I could also rescue the arrestin 1 translocation phenotype using an AAV mediated delivery of C LRN 1 to the retina. Adeno Associated virus AAV is the most prevalent method of viral gene therapy to date and has also had the most clinical success to date, particularly in the retina. AAV is a non pathogenic parvovirus that can infect both dividing and non dividing cells. AAV has the smallest packaging capacity of all the viral vectors with a maximum DNA size of only about 5kb. The AAV genome is a linear single strand of DNA that consi sts of two reading frames 2 Reprinted with permission from Di nculescu A and Stupay RM et al. AAV Mediated Clarin 1 Expression in the Mouse Retina: Implication s for USH3A Gene Therapy. PLoS One 2016.11(2):e0148874.
112 containing the replication and capsid coding genes. There are 3 transcriptional initiation sites that control expression for 4 replication proteins. There are 3 capsid proteins that are composed of two smaller proteins (VP1 and VP 2) and one larger protein (VP3). These capsid proteins are in a ratio of 1:1:10, with VP3 being the most abundant. Flanking these sequences are the inverted terminal repeat s equences that are necessary for replication and packaging (66, 108) The AAV constructs employed for gene therapy have all of the replication and capsid genes removed so that all that remains are the inverted terminal repeat sequences for packaging, in addition to a promoter, enhancers, gene of interest and a poly adenylation addition sequence that are inserted between the inverted terminal repeats (108) For my AAV gene therapy, there ar e two primary backbones that I have utilized, a pTR vector and a self complimentary (sc) vector. The pTR vector is a single stranded vector that requires the cellular machinery for second s trand DNA synthesis and therefore expresses the gene of int erest relatively slowly at lower levels initially. The sc vector is able to fold onto itself result ing in a double stranded DNA molecule immediately upon cellular infection. This allows for rapid transcription and expression of the gene (cDNA) of interest (66) Serotypes and Modifications M ultipl e different capsid serotypes have been identified as well as multiple different capsid serotype specific mutations that have optimize d the AAV vectors for better cell type specific infection. There have also been many studies looking at the success of AAV mediated gene therapy in the retina and the optimization of targeting specific cell types with different capsid mutants. at a simple AAV2 capsid serotype injected into the subretinal space. Using a rod opsin promot er driving a synthetic GFP protein, they saw exclusive expression in PR cells
113 with 20% pan retinal expression and 100% expression at the site of injection (83) They further validated that their recombinant AAV viral preps were free of Adenovirus an d WT AAV (83) Subsequent retinal examined cell type specific infection after subretinal and intravitreal injection as well as long term expre ssion, stability, and toxicity (71, 104, 108, 222) These studies initiated the development of successful gene replacement therapies for several animal models of retinitis pigmentosa The AAV2 capsid utilizes several charged surface residues that are responsible for AAV2 binding to its heparin su lfate receptor and allows for entry into the cell. On e complication with the AAV2 capsid is that on the capsid surface t here are many tyrosine residues that can be phosphorylated and therefore target the viral particle for ubiquitination and degradation wi thin the cell (179) Site specific mutation of these tyrosine residues specifically in the VP3 region of the capsid surface allow for AAV vectors to avoid degradation and achieve higher transduction efficiency (128, 179, 197) The initial capsid mutations studied were single point mut ations in the tyrosine residues on the AAV capsid surface for the AAV2, 8 and 9 serotypes (180) Further studies looked at the efficacy of combining multiple capsid mutations in different combinations of tyrosine t o phenylalanine both after subretinal and intravitreal delivery. They showed that multiple combinations of tyrosine phenylalanine (Y F) mutants displayed pan retinal expressio n of an AAV2 GFP injected virus (179, 180) Upon immunohistochemistry analysis, there was differential expression with each capsid mutation from only PR cell s to expression in all retinal cell types after subretinal delivery In contrast, after intravitreal delivery, the majority of viral expression was present in retinal g anglion cells for all capsid mutations analyzed, with only two
114 appearing to successf ully target PR cells after an intravitreal delivery (179, 180) A subsequent study examined un modified or modified capsid mutants for serotypes AAV1, 2, 5, and 8 in RPE cell and 661W cone c ell retinal cultures to assess for which additional mutations may be optimal (197) Vector expressed protein was detect ed by FACS sort ing for both cell lines and it was found that the addition of Y F mutations increased the transduction efficiency (197) These capsid mutations were further tested i n vivo to test which mutations would be optimal for intravitreal injection in order to avoid any surgical damage from subretinal injection. A dditional threonine to valine (T V) mutations were examined as well It was found that the previously identifie d AAV2 Q uadruple mutant with an additional T V mutation at aa 491 worked the best for intravitreal injection and produced the highes t level of transduced PR cells across the retina (128) T his vector is described i n the Methods below. Similar studies have been done looking at systemic delivery to the retina using various AAV9 capsid mutations in d ogs (33, 34, 60 62) For this study however the two mutants I have focused on are the AAV2 Q uad Y F and AAV2 Quad T491V vectors in addition to WT AAV2 capsid. Methods Clrn1 / KO and N48K KI mice were generated as described above (87, 88, 217) Animals were housed and maintained as noted previously. The AAV viral vectors were generated and purified as previously reported and viral titer s was measured by real ti me PCR (103, 115, 245) All viruses that were tested here are listed in table 5 1 with their capsid serotype and mutations, vector construct, an d viral titer listed. The AAV2 Q uadruple mutant contains 4 tyrosine to phenylalanine mutations at residues Y F 272, 444, 500, and 730 (AAV2 Q uad). The second mutant utilized was the AAV2 Q uad ca psid with an additional tyrosine to valin e mutation at aa 491 (AAV2 Q uad T491V).
115 Either a ubiquitous smCBA actin) or a PR specific GRK1 (G protein receptor kinase 1) promoter was used for all experiments. For direct vector product visualiz ation either a C te rminal Venus tagged human CLRN 1 (CLRN1 Venus, from Dr Imanishi at Case Western University) or a hemagglutinin ( HA ) tagged CLRN1 was used for all experiments. Both subretinal and intravitreal injections were pe rformed as previously des cribed (179, 180) Briefly, a 30.5 gauge needle was used to pie rce the nasal side of the cornea and a Hamilton syringe with a 33 gauge needle was used to inject the virus into the eye. Subretinal injections used a blunt tip needle to deposit the virus into the subretinal space betw een the RPE and PR cells, and intravi treal injections used a 45 o beveled tip needle to deposit the virus into the vitreous space. All eyes were injected with 1 ul of viral vector and the associated vector genome copies/ ml are defined in table 5 1 for each. Right eyes were injected and left e yes remained un injected. C57BL/6J mice were tested with buffer alone in the right eye to assess any damage resulting from the injection procedure I ntravitreal and subretinal vectors at a 1: 1000 dilution of CBA and GRK CLRN1 were also tested in right eye s against and buffer alone in left eye s further minimize any variability due to injection damage (data not shown) For vector toxicity assays, right eye s received dilutions of one microliter of either a full strength titer of approximately 10 13 vector geno me/ml (vg/mL) or serially diluted vectors of 10 12 10 11 or 10 10 vg/mL in the right eye All treated mouse eyes were first documented for PR function by ERG and for retinal structure by OCT at 4 months post injection with several were analyzed up to 1 year post injection. Later experiments were also analyzed by OCT at 1 week post injection to verify there was no injection damage prior to ERG analysis. At 1 month CLRN1 Venus and CLRN1 HA were visible
116 by immunohistochemistry using an anti GFP or anti HA antibodies (3F10, Roche Diagnostics, Indianapolis, IN). ERG, OCT translocation and immunohistochemistry experiments were performed as described above. Differences in maximum a and b wave amplitudes between injected and un injected eyes were analyzed by the Student t test (GraphPad Prism 6.0, GraphPad Software, San Diego, CA), and were considered statistically significant if p<0.05 or less All ERG amplitudes are presented as mean SEM. All tissue samples were processed and stained as described above. Results AAV Delivered Clarin 1 Retinal Localization I first wanted to assess AAV CLRN1 localization in a subretinal vs intravitreal vector delive ry to see which cell types express AAV injected CLRN1. For subretinal injection, v ector construct 3 Table 5 1 was delivered subretinally to WT C57BL/6J mice retinal sections were obtained at 1 month post injection, and sections were imaged on a spinning disc confocal microscope as described above (72) Sc sm CBA CLRN1 Venus localization was seen primarily in PR cells but also in RPE cells (Figure 5 1 A) with significant CLRN1 Venus expression throughout the PR cell body, particularly in the IS and some punctate staining at the CC. For intravitreal injections, vector construct 4, Table 5 1 was delivered intravitreally to WT C57BL/6J mice and analyzed the same as subretinal injections. CLRN1 Venus expression was seen in all retinal c ell types except for RPE cells with t he majority of AAV CLRN1 expression in the i nner retina and GCL. There particularly in their apical microvilli adjacent to the OLM (Figure 5 1 B) (72) with limited expression in PR cells, unlike the subretinal injection expression. Given that we believe C lrn 1 is expressed in PR cells, it seem s the optimal delivery method would be a subretinal injection, however,
117 a comp lication with this me thod is the risk of potential damage from injection due to the obligate retinal detachment. The previous human clini cal trials have seen this post injection, particularly i f the injection is delivered close to the fo vea (46 49, 102, 117) Because mice lack a fovea, this is not an issue for these preliminary studies Although the CLRN1 Venus construct allowed me to see where virally expressed protein was localized a Venus tagged construct is inappropriate for a gene therapy treatment in the clinic Therefore, I designed an AAV construct replac ing the Venus tag with an HA tag (Construct 1 Table 5 1) Ideally no additional tags should be on the injected construct in order to take the vector into the clinic, but prior to this study there was not an optimal antibody that could recognize endogenous or injected Clrn1/ CLRN1 in t he retina. By using an HA tag I can assess if there is any toxicity when there is no interferen ce with the PDZ binding domain. Additionally, now that I have identified an antibody that recognizes Clrn 1 in WT, Clrn1 / KO, and N48K KI mice, I can generate an AAV construct without any tags so that it can be optimized for the clinic. When I tested thi s new HA tagged vector subretinally I see similar significant expression in PR and RPE cells (Figure 5 2 A, B) Interestingly, there appea red to be significant cell death near the injection site, where the viral bleb was administered, and th is cell toxici ty was reduced towards the periphery of the retina (Figure 5 2 B, C) This suggests that the full viral titer is toxic to the retina using an HA tagged CLRN1 construct. This is most likel y because Clrn 1 is normally expressed at very low levels in the retin a, thus excess CLRN1 is likely toxic I believe that t he reason why I do not see this toxicity with a subretinally delivered CLRN1 Venus construct is that there is a PDZ binding domain on the CT tail o f CLRN1 and this is where both tags were placed. The
118 Venus tag is very large and most likely masks the binding ability of the PDZ domain, whereas the HA tag is very small, only 10 aa, and therefore leaves the PDZ binding domain open to interactions. This would lead to copious amounts of functional CLRN1 prot ein in the cell and most likely leads to the observed toxic effect. In addition to the large amounts of CLRN1 protein, CBA CLRN1 HA is also expres sed in the RPE cells, where Clrn 1 is not normally expressed and this may be a contributing fac tor to the toxi city because Clrn 1 is a transmembrane prot ein and it s membrane localization may interfere with the normal function of the RPE microvilli. Given that I cannot utilize the Venus tagged construct for a gene therapy treatment and that the HA tagged construct is extremely toxic, I needed to optimize the HA construct in order to minimize the toxicity or optimize an intravitreal delivery vector to target PR cells Optimized AAV Vector Capsid, Promoter, and Titer for Safe and Effective Gene Therapy Initially, con structs 7 and 8 Table 5 1 were tested in C57BL/6J WT control mice following subretinal or intravitreal injection, to assess whether t he toxicity could be reduced using a less potent AAV serotype, a simple AAV2 capsid. Although I did see expression via a su bretinal injection, there was almost no expression at all with the intravitreal injection. Therefore, I chose to optimize for minimal toxicity using the original HA tagged vector, construct 1 Table 5 1. As note d above, u sing a full viral titer of 8.43 x 10 12 vg/ml I saw significant PR cell loss via immunohistochemistry as well as a significant loss of PR cell function via ERG analysis (Figure 5 3 A, A ) I then a ssessed reduced viral titers sequentially to a 1:1000 dilution at 8.43 x 10 9 vg/ml T he only dilution that did not show significant cell death or loss of functi on was the 1:1000 dilution dose (Figure 5 (Figure 5 4) (72) I next compared PR cell function by ERG analysis
119 afte r a full titer subretinal injection and a full titer intravitreal injection. Similar to th e 1:1000 subretinal injection, intravitreal injection did not show any functional loss in ERG amplitu de in injected vs un injected eyes (Figure 5 5 A, B ) Using immunohistochemistry for CLRN1 HA in the intravitreally treated mice, there seemed to be ample expression in all retinal cell types as seen previously, but it appears to preferentially target Mller cells and RGCs (Figure 5 5 ) (72) T his is a complication because I believe that Clrn 1 should be expressed in PR cells and they do not appear to be targeted via an intravitreal injection with the CBA promoter. Because of this, I wanted to first optimize for toxicity and optimize for intravitreal delivery to target PR cells. Given the initial constructs are under a ubiquitous CBA promoter, I also wanted to see if the subretinal t oxicity could be mediated by u sing a photoreceptor specific GRK1 promoter, with construct 10 Table 5 1 injected in WT C57BL/6J mice. I also tested construct 11 Table 5 1 intravitreally to assess if I can both target PR cells spe cifically and avoid toxicity from the subretinal injection. First, using OCT analysis, I assessed the amount of PR cell loss i n vivo prior to ERG functional analysis. Using a full titer virus, there was a significant loss of PR cells in the retina at the injection si t e a nd cell loss was not observed in the half of the retina not receiving vector. H owever, there was extensive thickening of the ILM an d nerve fiber layer (Figure 5 6 At 3 months post injection, there was s ignificant degeneration in the retinal periphery at a 1: 10 dilution of 2.1 x 10 11 vg/ml, which correlated with th e injection site (Figure 5 6 B, I did not see any significant degeneration at a 1:100 dilution at 3 months post injection, but there was thickening of the ILM an d nerve fiber layer and there was n o significant retinal damage in the mice that were treated intravitreally (Figure 5 6
120 I next assessed functional loss via ERG analysis in all subretinal dilutions as well as for intravitreal injection s Given the extensive retinal degeneration with full titer injections, they were excluded from ERG analysis. Similar to the CBA promoter constructs, the full titer as well as a 1:10 dilution at 2.1 x 10 11 vg/ml of subretinal GRK1 CLRN 1 HA vector exhibited a significant loss of function by ERG analysis of both the a and b wave amplitudes compared to un injected eyes (Figure 5 7 A, B ) When I tested the same AAV2 Quad Y F GRK1 C LRN 1 HA vector using an intravitreal injection I saw no significant reduction in either the a or b wave amplitude s (Figure 5 7 C ) I still have significant CLRN1 HA expression even at a subretinal 1:1000 vector dilution, suggesting there is sufficient CLRN1 expression even at low vector dose levels. Given these results, it seems that a 1:1000 dilution is optimal for either the CB A or GRK1 promoter vectors (Figure 5 A t 8 months post treatment ther e is still significant GRK1 CLRN 1 HA e xpression in PR cells and there is no sig nificant cell loss. When I assess the scotopic ERG response at 10 months of age however, ther e is a significant loss of ERG amplitude for all light intensities both for the a and b wave responses (Figure 5 8 B, C) This indicates that there is still some toxicity over time with the 1:1000 dilution of subretinal GRK1 CLRN1 HA in control C57BL/6J mice Unfortunately, even though intravitreal GRK1 CLRN1 HA injected retinas do not degenerate or have a decrease in ERG response, I was unable to detect CLRN1 HA using immunohistochemistry in the se retinas. T his may be due to low leve ls of protein being expressed through the intravitreal vector delivery route because the intravitreal vector at this dose cannot penetrate to PR cells efficien tly enough to detect expression I additionally tested a novel AAV2 serotype that contains an add itional mutation that
121 changes a threonine to a valine and has been shown to allow greater transduction efficiency across the retina (AAV2 Quad Y F T491V) (27, 28, 128, 197) Even with the additional modification to the viral capsid, I am still unable to detect any anti CLRN1 HA staining post injection and it remains unclear why this is the case. Moving forward in terms of a gene therapy treatment, it seems that either the CBA intravitreal rou te or the 1:1000 CBA or GRK subretinal rout e are the best current options. Gene Therapy Rescue of Clarin1 Retinal Phenotypes Giv en my studies on identifying an optimal AAV CLRN1 vector design for gene therapy I first tested the AAV2 Quad Y F sm CBA C LRN 1 HA construct after an intravitreal injection and looked at the arrestin 1 translocation phenotype 1 month post injection in the original mixed C57BL/6J albino mouse strain. I performed the tra nslocation assay as described in Chapter 4 and mice were analyzed as described previously (Figure 4 2) (72) Looking at the initial immunohistochemistry analysis, there d id appear to be some decrease in arrestin 1 miss localization after light exposure in the N48K KI mice, however this was not very consistent across samples (Figure 5 9 A, This was quantified in the fluorescence intensity plots as previously described and the overall ratio of OS/OPL x 100 was calculated for all mice (Figure 5 (71) After averaging all mice assayed for all genotypes surprisingly the N48K KI mice no longer had a significantly higher amount of arrestin 1 in the OPL compared to WT mice. However, the percentage of arrestin 1 in the OPL is not significantly different from the untreated N48K KI retinas (Figure 5 9 C) There is also the problem of antigen masking because there is so much viral expression in the OPL post injection, which makes it possible that I am not picking up all of the arrestin 1 protein because the antibody
122 cannot bind. When I compare the WT to the Clrn1 / KO mice, there did not appear to be any potential rescue of the arrestin 1 translocation phenotype (Figure 5 9 D). I next assessed the potential to rescue the reduced b wave ERG phenotype in the new A/J strain of mice. Because only the Clrn1 / KO mice showed a reduced ERG phenotype, I only tested thos e mice for phenotypic rescue. I tested construct 2 Table 5 1 after intravitreal injections and measured scotopic ERGs for WT and Clrn1 / KO mice at 2 months and 4 months post injection. At two months post injection, there was no significant difference between the treated vs untreated Clrn1 / KO mice, however, there was a significant difference between WT and Clrn1 / KO untreated mice as seen previously (Figure 4 7) (217) At 4 months post injection there was a significant difference between the WT and Clrn1 / KO untreated eyes. I also saw a significant in crease in ERG response in the treated vs untreated Clrn1 / KO mice at 4 months of age (Figure 5 10) This suggests a potentially successful gene therapy in that there is a maintained ERG response after AAV gene therapy treatment in the Clrn1 / KO mice as they age Given the additional light damage and maintained translocation ERG phenotypes, there is a potential to use those assays for treatment in the N48K KI mice.
123 Table 5 1. AAV c onstructs Construct Plasmid Serotype Titer Volume Injection 1 Sc smCBA h CLRN 1 HA AAV2 Quad Y F 8.43x10 12 1 ul Subretinal 2 Sc smCBA h CLRN1 HA AAV2 Quad Y F 8.43x10 12 1 ul Intravitreal 3 sc smCBA h CLRN 1 Venus AAV2 Q uad Y F 1.54x10 12 1 ul Subretinal 4 sc sm CBA h CLRN 1 Venus AAV2 Q uad Y F 1.54x10 12 1 ul Intravitreal 5 sc sm CBA h CLRN1 HA AAV2 Q uad Y F 1.68x10 12 1 ul Subretinal 6 sc sm CBA h CLRN1 HA AAV2 Q uad Y F 1.68x10 12 1 ul Intravitreal 7 sc sm CBA h CLRN1 HA AAV2 2.27x10 12 1 ul Subretinal 8 sc sm CBA h CLRN1 HA AAV2 2.27x10 12 1 ul Intravitreal 9 sc sm CBA h CLRN1 HA AAV8 Y733F 1.23x10 13 1 ul Subretinal 10 pTR GRK1 h CLRN1 HA AAV2 Q uad 2.1x10 12 1 ul Subretinal 11 pTR GRK1 h CLRN1 HA AAV2 Q uad 2.1x10 12 1 ul Intravitreal 12 pTR GRK1 h CLRN1 HA AAV2 Q uad Y F T491V 2.31x10 13 1 ul I ntravitreal
124 Figure 5 1. Localization of v ector e xpressed CL A R I N 1 Venus f ollowing s ubretinal and intravitreal d elivery A) CLRN1 Venus fusion fluorescence was detected on the apical side of RPE cell membrane s (arrows), and in PR IS, ONL a nd OPL. strong CLRN1 Venus fluorescence in specific regions within photoreceptor cells: IS, cell body membrane and OPL. Nuclei are stain ed blue with DAPI B) Representative image of retinal cross section showing CLRN1 Venus fusion fluorescence. Strong expression is seen at the OPL, the stratified dendrites within the IPL, and inner retinal neurons. CLRN1 Venus expression is shown by intense fluorescence at the Mller cells and their apical processes at the OLM
125 Figure 5 2 Toxicity and p hotoreceptor c ell d eath post s ubretinal i njection of CLARIN 1 A) A cross section of a whole eye showing pan retinal expression of AAV deliver ed CLRN 1 (green). C57BL/6J mice were injected subretinally with an AAV2 Q uad Y F sc sm CBA h C LRN 1 HA virus at a full titer of 8. 43x10 12 vg/ml and showed at 6 weeks post injection. B) R etinal section adjacent to injection site, there is minim al PR cell death peripheral to t he injection site. C) R etinal section at the site of injection, there is significant PR cell death. Localizatio n of vector expressed CLRN1 HA protein by immunostaining with an anti HA antibody (40X magnification).
126 Figure 5 3. Photoreceptor c ell death and r etinal m orphology in f ull titer vs 1:1000 dilution of s ubretinally d elivered CLARIN 1 A) Average ERG amplitudes of 7 mice injected subretinally with a full titer of AAV2 Q uad Y F s c sm CBA h CLRN1 HA at a viral titer of 8.43x10 12 vg/ml Black bars are un injected left eyes and white bars are injected right eyes. The full titer shows a signif icant decrease in ERG amplitude at all scotopic light intensities. Immunohistochemistry staining for anti HA CLRN1 (green) and anti arrestin 1 (red). Retinal cross sections show significant retinal damage and PR cell death at the site of injection. B) Average ERG amplitudes of 5 mice injected subretinally with a 1:1000 dilution of AAV2 Q uad Y F s c sm CBA h C LRN 1 HA at a viral titer of 8.43x10 9 vg/ml Black bars are un injected left eyes and white bars are injected right eyes. The 1:1000 dilution titer shows no significant decrease in ERG amplitude at all scotopic light intensities. Immunohistochemistry staining for anti HA CLRN1 (green) and anti arrestin 1 (red). Retinal cross sections show significant preserv ation of retinal morphology and minimal PR cell death at the site of injection.
127 Figure 5 4. Evaluation of r etinal f unction and retinal morphology in CL A R I N 1 HA injected mice following s ubretinal d elivery. A). Bar graphs show the average maximum ERG b wave amplitudes in scotopic, dark adapted conditions of un injected wild type contr ol eyes, compared to AAV injected eyes that received decreasing doses of AAV CLRN1 HA vector (2 months post injection). Max b wave amplitudes in AAV injected eyes were dos e dependent, and were sig nificantly lower than un injected controls at 10 10 10 9 and 10 8 vg (* p<0.05, ** p<0.001). B). Localization of CLRN1 HA protein following subretinal delivery of the AAV2 Quad Y F sc smCBA hC LRN1 HA vector. CLRN1 HA fluorescence (g reen) was detected by immunohistochemistry in photoreceptor IS region, ONL and OPL, and was absent from the outer segments (labelled with a rhodopsin antibody, red). The eyes received diluted vector (10 7 vg).
128 Figure 5 5 ERG and Immunohistochemistry a n alysis of s ub retinal vs i ntravitreal AAV2 Qua d Y F smCBA h C LARIN 1 HA in C57BL/6J mice 6 w eeks post i njection All mice are control C57BL/6J strain measured 1 month post inj ection. Black bars are un injected eyes and white bars are injected eyes. A) Subretinal injected AAV2 Q uad Y F sc sm C BA h CLRN 1 HA. ERGs for all scotopic light intensities are significantly decre ased in injected vs un injected eyes for both the a and b waves P values= *<0.05. **<0.01. B) Intravitreally injected AAV2 Q uad Y F smBA h CLRN 1 HA. ERGs are not significantly decreased in injected vs un injected eyes for all light intensities C ) Immunohistochemistry staining for anti HA CLRN1 (green). There is significant expression in the GCL and INL. There also seems to be significant ex pression in Mller cells, with some possible expression in PR cells C arrestin 1 (red) staining for PR cells merged with CLRN1 HA (green). C arrestin 1 (red), Clrn HA (green ), and D API (blue) co staining, merged
129 Figure 5 6 SD OCT of subretinally injected AAV2 Quad Y F pTR GRK 1 h CLARIN 1 HA All mice tested were C57BL/6J controls injected subretinally with AAV2 Quad Y F pTR GRK1 h C LRN 1 HA. A) Full titer of 2.10x10 12 vg/ml 1 year post Image A taken adjacent to the optic nerve. B) 1:10 dilution of 2.10x10 11 vg/ml. 3 months post injection taken at the Image B taken adjacent to the optic nerve C) 1:100 dilution of 2.10x10 10 vg/ml. 3 months post injection taken at the optic nerve C taken adjacent to the optic nerve
130 Figure 5 7 ERG analysis of subretinal vs i ntravitreal AAV2 Quand Y F pTR GRK1 h CLARIN 1 HA 3 m onths p ost i nj ection. White bars are un injected and black bars are injected eyes A) Subretinal 1:10 dilution a wave ERG. All light intensities are significantly decrease d in injected v ersu s un injected eyes, p values= *<0.05, **<0.01. B) Subretinal 1:10 dilution b wave ERG. All light intensities are significantly decreased in injected v ersu s un inj ected eyes p values= *<0.05, **<0.01, ***<0.001. C) Intravitreal full titer a wave ERG. All intensities are not significantly different between injected and un injected eyes D) Intravitreal full titer b wave ERG. All intensities are not significantly dif ferent between injected and un injected eyes
131 Figure 5 8 Evaluation of retinal f unction and GRK1 h CL A R I N 1 H A e xpression following subretinal d elivery. A ). Localization of CLRN1 HA protein following subretinal delivery of the AAV2 Q uad pTR GRK1 h C LRN 1 HA vector. CLRN1 HA fluorescence (green) was detected by immunohistochemistry in photoreceptor IS region, ONL and OPL, and was absent from the outer segments staining with CLRN1, DAPI, and rhodopsin for OS mark er. The eyes received a diluted vector at a dilution of 1:1000 At 10 months pf age there did not appear to be any reduction in ONL thickness. B) a wave ERG at 10 months post injection. C) b wave ERG at 10 months post injection. Over time there is a signif icant loss in scotopic ERG amplitude for both a and b waves over time.
132 Figure 5 9 Arrestin 1 t ranslocation q uantification after i ntravitreal i njection A) Immunohistochemistry staining of the untreated left eye of an N48K KI mouse after 1 hour of light exposure. There is still a significant amount of arrestin 1 remaining at the OPL synapses. A rrestin 1 (red) and DAPI (blue). Immunohistochemistry staining of the treated righ t eye of an N48K KI mouse after 1 hour of light exposure. There is no longer a significant amount of arrestin 1 remaining at the OPL synapses in mice treated intravitreally with AAV2 Q uad YF sc smCBA h C LRN 1 HA. Arrestin 1 (red), CLRN1 HA (green), and DAPI (blue). B) Fluorescence intensity plot of the un trea ted left fluorescence intensity ratio of the OPL/OS in C57BL/6J WT vs N48K KI mice. There appears to be a decrease in the amount o f arrestin 1 at the OPL synapses vs the OS, but the amount in the treated vs untreated N48K KI eyes is not significant even though there is no longer a difference between the KI and C57BL/6J WT mice. D) Average fluorescence intensity ratio of the OPL/OS in C57BL/6J Albino WT vs Albino Clrn1 / KO mice. There appears to be no decrease in the amount of arrestin 1 at the OPL synapses vs the OS in the treated vs untreated Clrn1 / KO eyes, and the ratio of arrestin 1 in the OPL/OS in the WT vs Clrn1 / KO re mained the same after treatment.
133 Figure 5 10 Evaluation of retinal f unction in WT vs Cl a r i n 1 / KO untreated and treated retinas f ollowing intravitreal d elivery Bar graph show s the average maximum ERG b wave amplitudes in scotopic, dark adapted c onditions of untreated WT and Clrn1 / KO control eyes, compared to AAV treated eyes that received the full titer of AAV2 Quad sc smCBA h C LRN 1 HA vector (2 months and 4 months post injection ). Maximum b wave amplitudes were not significantly lower than un treated controls at 2 or 4 months in the WT mice. Clrn1 / KO mice at 2 months were slightly higher in the treated vs untreated eyes and the untreated Clrn1 / KO mice had a significantly reduced ERG compared to WT, as shown above. At 4 months the Clrn1 / KO treated eyes were significantly higher than the untreated eyes and the ERG from the untreated eyes remained significantly lower than the WT.
134 CHAPER 6 CONCLUSIONS A primary complication with t he USH mouse models is the majority have normal retinal function and morphology Many studies for USH proteins localize them to the PR CC and calyceal processes in the retina and the stereocilia bundle on cochl ear hair cells in the ear (5, 198) The USH3 protein, CLRN1, is the only USH protein not proven to be a component of these USH networks (3, 147, 194, 211) CLRN1 is believed to be a tetraspanin transmembrane protein that has a C terminal PDZ binding domain with a suggested role in sensory synap ses (4, 132) One issue with the previous Clrn1 /Clrn1 data, is the retinal localization data i s conflicting and may due to the different techniques utilized The i mmunohistochemistry data localized C lrn 1 at the CC in PR cel ls, and t he i n situ data showed C lrn 1 expression in the INL, most likel y in Mller cells, with expression loss at 1 month old (53, 87, 241) Another group using retinal transcript ion profiling showed Clrn1 expression only in starburst amacrine cells (207) Given this conflicting data, it was unclear which cell types in the retina express Clrn1 / C lrn 1. More recently, i n zebrafish C lrn 1 was shown to localize to inner retinal cells and PR cells inc luding cell cell contacts between PR and Mller glial cells at the OLM with some faint expression in the GCL (93, 182) C lrn 1 localization in the cochlea show ed staining in the cochlear hair cells in the inner ear (93, 146, 182, 241) Recently, it has also been shown that Pch15, the USH 1F protein, binds to the C terminal tail of C lrn 1 and that C lrn 1 is essential for proper subcellular localization of PCH15 in the cochlea r hair cells (169) The zebrafish CC staining supports the subcellular localization of C lrn 1 that I s aw at the CC in PR cell s with the NB CLRN1 antibody in mice (Figure 3 2 ) Additionally, this antibody colocalizes with AAV injected CLRN1 both vi a the subretinal or intravitreal
135 route (Figure 3 6) c onfirming the NB CLRN1 antibody recognize s C lrn 1 T he N48K KI mice, have C lrn 1 localizing at the OS distal tips and/or RPE microvilli (Figure 3 4 ) T he Clrn1 / KO mice appear to have normal synapses but N48K KI mice w ere not studied (4, 72, 87, 89, 169, 241, 242) Since other USH proteins were shown to play a role in sensory synapses, it would be useful to pursue in these m ice (3, 135, 191 194) Here I show that the Clrn1 / ; r d1 / double KO mouse has no expression of the 250 aa isoform in WT or Clrn1 / KO mice lack ing PR cells, but the WT Clrn1 + / + ; rd1 / mice lacking PR cells still express the 232 aa isoform (Figure 3 1C, D) The C57BL/6J mice contain ing PR cells express the 250 aa isoform, suggesting the 250 aa isoform is most likely expressed by PR cells and the 232 aa isoform is expressed b y the inner retina (Figure 3 1) Additionally, quant it ative RT PCR could be optimized for isoform specific primers This would inform us whether the original RT PCR for the double C lrn 1 /r d1 WT and KO data is correct because I could compare whole retina to only inner retina and measure if it is really isoform 1 in the INL, isoform 2 in PR cells, and where isoform 3 may be expressed, if at all. In addition to identifying where endoge nous Clrn1 /C lrn 1 is expressed in the retina, I further attempted to ident ify a retinal phenotype in multiple mouse strains for Clrn1 / KO and N48K KI mice. I identified a delay in arrestin 1 translocation from the OPL to the OS upon exposure to light (Figure 4 2 ) This is significant because other USH mouse models have report ed a light induced delay in transducin movement but not arrestin 1 (176, 219) I further showed that N48K KI mice have a maintained ERG response after 1 hour of light exposure indicating that under a 0 dB mesopic single flash light intensity N48K KI mic e are still responsive to light (Figure 4 8 ) Given the
136 delay in arrestin 1 movement and maintained ERG response, I test ed the ERG recovery response rate to see if with age there was a significant difference. This was previously shown in an arrestin 1 doub le KO mouse (235) where the stud y identified a novel prolonged p hotoresponse with altered rhodopsin recovery in these mice, but these mice have a completely normal ONL similar to my mice (235) Unfortunately, I was unable to see any difference in ERG re covery in th e WT vs N48K KI A/J mice (Figure 4 9 ) To further test whether or not the arrestin 1 delay and maintained ERG response indicate a potential mechanism for disease progression I also looked at the ability to induce degeneration and PR cell death after a 4 hour bright light exposure In WT A/J mice there was some significant PR cell loss in the ONL after 4 hours of light exposure at 2000 lux however there was no decrease in ERG response and this was therefore used as the control light intensity (Figure 4 1 0 ) When I assessed for both retinal degeneration and ERG loss after 2000 lux of light damage, the Clrn1 / KO mice showed minimal degeneration in the superior retina whereas the N48K KI mice showed degeneration in both the superior and infe rior retina b ut neither was significantly different from the WT mice after light damage When I looked at a loss of function by ERG analysis after light dam age the Clrn1 / KO mice showed no significant decrease in ERG response; however, the N48K KI mice showed a sig nificant decrease in ERG amplit ude for the 10 and 0 Db a wave intensities and at every scotopic b wave light intensity This now further provides a quantifiable outcome parameter that I can assess for an AAV mediated gene therapy phenotype rescue in the N48K KI mice in the future. I also validated the A/J mouse strain has a reduced ERG compared to C57BL/6J mice, and Clrn1 / KO A/J mice have a significant decrease in ERG amplitude compared
137 to WT A/J mice as they age as previously published (Figure 4 6 4 7 ) (160, 217) Given thes e phenotypes between the Clrn1 / KO and N48K KI mice, although they both possess a delay in arrestin 1 translocation, it seems that the Clrn1 / KO phenotype is more related to a decreased b wave ERG at dim light intensities, which may be due to an inne r retinal and not a PR cell phenotype. In contrast, the N48K KI phenotype is due to a maintained ERG respo nse after light exposure, and loss of retinal function after light damage with a significantly reduced a and b wave ERG at all light intensities (Figure 4 This suggests the N48K KI phenotype is due to malfunctions in PR cells. This is significant because it means that even though both mutations are in the same protein, there may be diffe rent biological mechanisms inducing loss of retinal function and degeneration depending on the mutation. This is p articularly relevant in the case of the N48K KI mutation since it s the most prevalent mutation in USH 3A patient s I next assessed the gene therapy potential for Clrn1/ CLRN1 First I looked at AAV CLRN1 expression in the retina after subretinal or intravitreal injections with an AAV2 Quad smCBA h C LRN 1 Venus tagged vector Subretinal delivery showed AAV CLRN1 in PR and RPE cells with the majority in the IS and ONL of PR cells and disti nct punctate staining at the CC (Figure 5 1) (72) I ntravitreal ly delivered AAV CLRN1 was expressed in every retinal cell type particularly in the INL and GCL. There is also significant staining in Mller cells based on the staining in their apical processes at the OLM (Figure 5 1) I originally tested the Venus tagged construct because at the time there was not a Clrn1/ CLRN1 antibody available h owever, a Venus tagged constru ct would not be able to move forward for gene therapy treatment in the clinic.
138 I generated a C terminal HA Tagged AAV CLRN1 vector in order to follow viral expres sion levels without altering Clrn1/ CLRN1 protein structure Unfortu nately, when I t ested the s ubretinal AAV2 Quad Y F smCBA h ClLRN1 HA construct the virus was extremely toxic and killed the majority of PR cells at the injection site (Figre5 2) This was significantly reduced at a vector dilution of 1:1000 from the original titer and protein expression was still observed using an anti HA antibody (Figure 5 3, 5 4) Furthermore, when I compared the full titer subretinal and intravitreal injection s there appeared to be no toxicity in the int ravitreally treated eyes both functionally via ERG amplitudes, and structurally with no loss of PR cells (Figure 5 5) Given the CBA HA vector toxicity, and the ability for the AAV CLRN1 constructs to transfect multiple cell types, I needed to optimize fo r PR cell expression, a delivery method that is not toxic and a capsid serotype and design that will target PR cells through both injection methods if possible T o specifically target PR cells by subretinal and intravitreal delivery, I generate d a constr uct driven by a PR specific GRK1 promoter with an HA tag. First I performed OCT analysis 3 months post treatment to assay for subretinal injection damage. The full titer GRK1 vector was extremely toxic, with complete loss of PR at the injection site. At a 1:10 dilution, there was still some cell death, and only at the 1:100 and 1:1000 diluti ons was there no significant degeneration (Figure 5 6) W hen I assess the ERG response for each injection, the 1:10 dilution w as extreme ly toxic to the retina and only the 1:100 and 1:1000 intravitreal injections did not show any significant loss in retinal function (Figure 5 7) With age the 1:1000 dilution at 10 months old (8 months post treatment) still show ed no significant retinal degeneration however, there was a sig nificant loss of ERG amplitude for al l scotopic light intensities in both the a and b
139 wave s (Figure 5 8) This suggests that even though the initial treatment appeared safe and non toxic, over time there is still some functional toxicity that results from the AAV2 Quad Y F pTR GRK1 h C LRN 1 HA vector treatment. Given the possibility that there are different isoforms of Clrn1/ CLRN1 present in different cell types, it seem s that the currently preferred method for my gene therapy approach is either a subret inal del i very of CBA h CLRN1 HA vector at a 1:1000 dilution dose of 8.43 x 10 9 vg/ml or an intravitreal delivery of CBA h CLRN1 HA vector at a full titer dose of 8.43 x 10 12 vg/ml. To test for a potential treatment, I intravitreally injected the AAV2 Quad smCBA hCLRN1 HA construct into WT, Clrn1 / KO, and N48K KI mice (Table 5 1, construct 11) I performed the arrestin 1 translocation experiment 1 month post injection and the N48K mice showed some significant rescue o f the arrestin 1 delay after light exposure compared to WT mice but the untreated versus treated Clrn1 / KO eyes did not show a significant difference (Figure 5 9) Because the arrestin 1 localization is not significantly different in WT vs N48K KI trea ted eyes, this is still a useful phenotype to screen for the success of a gene therapy treatment because they are behaving more like the WT mice. I fu rther wanted to attempt to rescue the reduced ERG b wave phenotype in the Clrn1 / KO mice using an intravitreal injection of the AAV2 Quad smCBA h C LRN 1 HA construct. A t 2 months post injection there was no significant ERG rescue in treated vs untreated retinas, but at 4 months post injection there wa s a significantly higher ERG amplitude in the treated vs untreated retinas (Figure 5 10) This is pro mising because it means that the AAV treatment was able to maintain a stable ERG response compared to the untr eated control. A similar vector was used by a collaborator to test for AAV mediated CLRN1 expressio n in cochlear hair cells in a new mouse model Using in situ
140 hybridization t hey observed CLRN1 expression in inner and outer cochlear hair cells as well as ganglion cells until 1 week post natally (90) To test for a therapeutic rescue of the hearing loss phenotype in Clrn1 / KO mice they observed ABR (auditory brainstem response) thresholds in WT, Clrn1 / KO untreated, and Clrn1 / KO treated mice. Mice were injected at P3 and analyzed 22 weeks post injection. At all three ABR intensities tested (8, 16, and 32 kHz) they saw a significant rescue in ABR thresholds post treatment in the Clrn1 / KO mice (90) This AAV treatment can further be validated after retinal injections. Additionally, because the N48K KI mice also showed degeneration and a significant loss of ERG functi on post light damage this can also be a potential assay to measure the success of a gene therapy treatment. If these phenotypic rescue experiments prove to be reliable and reproducible they can further be used to validate my viral vectors in order to tran slate my gene therapy into the clinic to treat patients. In conclusion, I have shown that my novel NB CLRN1 antibody stains the CC in PR cells and that it recognizes my injected viral construct through both a subretinal and intravitreal delivery I have also identified a novel retinal phenotype of a delay of arrestin 1 movement upon exposure to light in both the KO and KI mice. I confirmed that the Clrn1 / KO display a reduced ERG b wave amplitude phenotype, in agreement with a recent study (217) I further showed that N48K KI mice a re more suscep tible to light damage compared to the WT and Clrn1 / KO mice due to a loss of a and b wave ERG amplitude after light exposure All of these previously unidentified phenotypes can now be employed to further screen my optim ized AAV vector de signs to gain a statistically valid and reproducible approach to develop a gene therapy treatment for USH3A.
141 LIST OF REFERENCES 1. Adams NA, Awadein A, and Toma HS. The retinal ciliopathies. Ophthalmic Genet 28: 113 125, 2007. 2. Adamus G, Zam ZS, Arendt A, Palczewski K, McDowell JH, and Hargrave PA. Anti rhodopsin monoclonal antibodies of defined specificity: characterization and application. Vision Res 31: 17 31, 1991. 3. Adato A, Michel V, Kikkawa Y, Reiners J, Alagramam KN, Weil D, Yonekawa H, Wolfrum U, El Amraoui A, and Petit C. Interactions in the network of Usher syndrome type 1 proteins. Hum Mol Genet 14: 347 356, 2005. 4. Adato A, Vreugde S, Joensuu T, Avidan N, Hamalainen R, Belenkiy O, Olender T, Bonne Tamir B, Ben Asher E, Espinos C, Milln JM, Lehesjoki AE, Flannery JG, Avraham KB, Pietrokovski S, Sankila EM, Beckmann JS, and Lancet D. USH3A transcripts encode clarin 1, a four transmembrane domain protein with a possible role in sensory synap ses. Eur J Hum Genet 10: 339 350, 2002. 5. Ahmed ZM, Frolenkov GI, and Riazuddin S. Usher proteins in inner ear structure and function. Physiol Genomics 45: 987 989, 2013. 6. Ahmed ZM, Goodyear R, Riazuddin S, Lagziel A, Legan PK, Behra M, Burgess SM, Lill ey KS, Wilcox ER, Griffith AJ, Frolenkov GI, Belyantseva IA, Richardson GP, and Friedman TB. The tip link antigen, a protein associated with the transduction complex of sensory hair cells, is protocadherin 15. J Neurosci 26: 7022 7034, 2006. 7. Ahmed ZM, K jellstrom S, Haywood Watson RJ, Bush RA, Hampton LL, Battey JF, Riazuddin S, Frolenkov G, Sieving PA, and Friedman TB. Double homozygous waltzer and Ames waltzer mice provide no evidence of retinal degeneration. Mol Vis 14: 2227 2236, 2008. 8. Ahmed ZM, Ri azuddin S, Ahmad J, Bernstein SL, Guo Y, Sabar MF, Sieving P, Griffith AJ, Friedman TB, Belyantseva IA, and Wilcox ER. PCDH15 is expressed in the neurosensory epithelium of the eye and ear and mutant alleles are responsible for both USH1F and DFNB23. Hum M ol Genet 12: 3215 3223, 2003. 9. Ahmed ZM, Riazuddin S, Bernstein SL, Ahmed Z, Khan S, Griffith AJ, Morell RJ, Friedman TB, and Wilcox ER. Mutations of the protocadherin gene PCDH15 cause Usher syndrome type 1F. Am J Hum Genet 69: 25 34, 2001. 10. Ahmed ZM Riazuddin S, and Wilcox ER. The molecular genetics of Usher syndrome. Clin Genet 63: 431 444, 2003.
142 11. Akoury E, El Zir E, Mansour A, Mgarban A, Majewski J, and Slim R. A novel 5 bp deletion in Clarin 1 in a family with Usher syndrome. Ophthalmic Gene t 32: 245 249, 2011. 12. Armengot M, Salom D, Diaz Llopis M, Millan JM, Milara J, Mata M, and Cortijo J. Nasal ciliary beat frequency and beat pattern in retinal ciliopathies. Invest Ophthalmol Vis Sci 53: 2076 2079, 2012. 13. Athanasiou D, Aguila M, Bevilacqua D, Novoselov SS, Parfitt DA, and Cheetham ME. The cell stress machinery and retinal degeneration. FEBS Lett 587: 2008 2017, 2013. 14. Bahloul A, Michel V, Hardelin JP, Nouaille S, Hoos S, Houdusse A, England P, and Petit C. Cadherin 23, myosin V IIa and harmonin, encoded by Usher syndrome type I genes, form a ternary complex and interact with membrane phospholipids. Hum Mol Genet 19: 3557 3565, 2010. 15. Barrong SD, Chaitin MH, Fliesler SJ, Possin DE, Jacobson SG, and Milam AH. Ultrastructure of c onnecting cilia in different forms of retinitis pigmentosa. Arch Ophthalmol 110: 706 710, 1992. 16. Beltran WA, Cideciyan AV, Guziewicz KE, Iwabe S, Swider M, Scott EM, Savina SV, Ruthel G, Stefano F, Zhang L, Zorger R, Sumaroka A, Jacobson SG, and Aguirre GD. Canine retina has a primate fovea like bouquet of cone photoreceptors which is affected by inherited macular degenerations. PLoS One 9: e90390, 2014. 17. Belyantseva IA, Boger ET, Naz S, Frolenkov GI, Sellers JR, Ahmed ZM, Griffith AJ, and Friedman TB Myosin XVa is required for tip localization of whirlin and differential elongation of hair cell stereocilia. Nat Cell Biol 7: 148 156, 2005. 18. Berbari NF, O'Connor AK, Haycraft CJ, and Yoder BK. The primary cilium as a complex signaling center. Curr Bi ol 19: R526 535, 2009. 19. Berger A, Cavallero S, Dominguez E, Barbe P, Simonutti M, Sahel JA, Sennlaub F, Raoul W, Paques M, and Bemelmans AP. Spectral domain optical coherence tomography of the rodent eye: highlighting layers of the outer retina using si gnal averaging and comparison with histology. PLoS One 9: e96494, 2014. 20. Bermingham McDonogh O, Corwin JT, Hauswirth WW, Heller S, Reed R, and Reh TA. Regenerative medicine for the special senses: restoring the inputs. J Neurosci 32: 14053 14057, 2012. 21. Bermingham McDonogh O and Reh TA. Regulated reprogramming in the regeneration of sensory receptor cells. Neuron 71: 389 405, 2011.
143 22. Bilotta J and Saszik S. The zebrafish as a model visual system. Int J Dev Neurosci 19: 621 629, 2001. 23. Bonnet C an d El Amraoui A. Usher syndrome (sensorineural deafness and retinitis pigmentosa): pathogenesis, molecular diagnosis and therapeutic approaches. Curr Opin Neurol 25: 42 49, 2012. 24. Bonnet C, Grati M, Marlin S, Levilliers J, Hardelin JP, Parodi M, Niasme G rare M, Zelenika D, Dlpine M, Feldmann D, Jonard L, El Amraoui A, Weil D, Delobel B, Vincent C, Dollfus H, Eliot MM, David A, Calais C, Vigneron J, Montaut Verient B, Bonneau D, Dubin J, Thauvin C, Duvillard A, Francannet C, Mom T, Lacombe D, Duriez F, D rouin Garraud V, Thuillier Obstoy MF, Sigaudy S, Frances AM, Collignon P, Challe G, Couderc R, Lathrop M, Sahel JA, Weissenbach J, Petit C, and Denoyelle F. Complete exon sequencing of all known Usher syndrome genes greatly improves molecular diagnosis. Or phanet J Rare Dis 6: 21, 2011. 25. Bonnet C, Riahi Z, Chantot Bastaraud S, Smagghe L, Letexier M, Marcaillou C, Lefvre GM, Hardelin JP, El Amraoui A, Singh Estivalet A, Mohand Sad S, Kohl S, Kurtenbach A, Sliesoraityte I, Zobor D, Gherbi S, Testa F, Simo Vidmar M, Zupan A, Battelino S, Martorell Sampol L, Claveria MA, Catala Mora J, Dad S, Mller LB, Rodriguez Jorge J, Hawlina M, Auricchio A, Sahel JA, Marlin S, Zrenner E, Audo I, and Petit C. An innovative strateg y for the molecular diagnosis of Usher syndrome identifies causal biallelic mutations in 93% of European patients. Eur J Hum Genet 24: 1730 1738, 2016. 26. Bork JM, Peters LM, Riazuddin S, Bernstein SL, Ahmed ZM, Ness SL, Polomeno R, Ramesh A, Schloss M, S risailpathy CR, Wayne S, Bellman S, Desmukh D, Ahmed Z, Khan SN, Kaloustian VM, Li XC, Lalwani A, Bitner Glindzicz M, Nance WE, Liu XZ, Wistow G, Smith RJ, Griffith AJ, Wilcox ER, Friedman TB, and Morell RJ. Usher syndrome 1D and nonsyndromic autosomal rec essive deafness DFNB12 are caused by allelic mutations of the novel cadherin like gene CDH23. Am J Hum Genet 68: 26 37, 2001. 27. Boyd RF, Sledge DG, Boye SL, Boye SE, Hauswirth WW, Komromy AM, Petersen Jones SM, and Bartoe JT. Photoreceptor targeted gene delivery using intravitreally administered AAV vectors in dogs. Gene Ther 23: 223 230, 2016. 28. Boye SE, Alexander JJ, Witherspoon CD, Boye SL, Peterson JJ, Clark ME, Sandefer KJ, Girkin CA, Hauswirth WW, and Gamlin PD. Highly Efficient Delivery of Adeno Associated Viral Vectors to the Primate Retina. Hum Gene Ther 27: 580 597, 2016. 29. Boye SE, Boye SL, Lewin AS, and Hauswirth WW. A comprehensive review of retinal gene therapy. Mol Ther 21: 509 519, 2013.
144 30. Boye SE, Boye SL, Pang J, Ryals R, Everhart D, Umino Y, Neeley AW, Besharse J, Barlow R, and Hauswirth WW. Functional and behavioral restoration of vision by gene therapy in the guanylate cyclase 1 (GC1) knockout mouse. PLoS One 5: e11306, 2010. 31. Boye SL, Peshenko IV, Huang WC, Min SH, McDoom I, Kay CN, Liu X, Dyka FM, Foster TC, Umino Y, Karan S, Jacobson SG, Baehr W, Dizhoor A, Hauswirth WW, and Boye SE. AAV mediated gene therapy in the guanylate cyclase (RetGC1/RetGC2) double knockout mouse model of Leber congenital amaurosis. Hum Gene Ther 24: 189 202, 2013. 32. Boye SL, Peterson JJ, Choudhury S, Min SH, Ruan Q, McCullough KT, Zhang Z, Olshevskaya EV, Peshenko IV, Hauswirth WW, Ding XQ, Dizhoor AM, and Boye SE. Gene Therapy Fully Restores Vision to the All Cone Nrl( / ) Gucy2e( / ) Mouse Model of Leber Congenital Amaurosis 1. Hum Gene Ther 26: 575 592, 2015. 33. Bruewer AR, Mowat FM, Bartoe JT, Boye SL, Hauswirth WW, and Petersen Jones SM. Evaluation of lateral spread of transgene expression following subretinal AAV mediated gene delivery in dog s. PLoS One 8: e60218, 2013. 34. Byrne LC, Lin YJ, Lee T, Schaffer DV, and Flannery JG. The expression pattern of systemically injected AAV9 in the developing mouse retina is determined by age. Mol Ther 23: 290 296, 2015. 35. Caberlotto E, Michel V, Foucher I, Bahloul A, Goodyear RJ, Pepermans E, Michalski N, Perfettini I, Alegria Prvot O, Chardenoux S, Do Cruzeiro M, Hardelin JP, Richardson GP, Avan P, Weil D, and Petit C. Usher type 1G protein sans is a critical component of the tip link complex, a structure controlling actin polymerization in stereocilia. Proc Natl Acad Sci U S A 108: 5825 5830, 2011. 36. Calvert PD, Govardovskii VI, Arshavsky VY, and Makino CL. Two temporal phases of light adaptation in retinal rods. J Gen Physiol 119: 129 145, 20 02. 37. Calvert PD and Makino CL. The time course of light adaptation in vertebrate retinal rods. Adv Exp Med Biol 514: 37 60, 2002. 38. Calvert PD, Schiesser WE, and Pugh EN. Diffusion of a soluble protein, photoactivatable GFP, through a sensory cilium. J Gen Physiol 135: 173 196, 2010. 39. Calvert PD, Strissel KJ, Schiesser WE, Pugh EN, and Arshavsky VY. Light driven translocation of signaling proteins in vertebrate photoreceptors. Trends Cell Biol 16: 560 568, 2006.
145 40. Cao D, Pokorny J, and Grassi MA. Isolated mesopic rod and cone electroretinograms realized with a four primary method. Doc Ophthalmol 123: 29 41, 2011. 41. Cardenas Rodriguez M and Badano JL. Ciliary biology: understanding the cellular and genetic basis of human ciliopathies. Am J Med Gen et C Semin Med Genet 151C: 263 280, 2009. 42. Chamling X, Seo S, Bugge K, Searby C, Guo DF, Drack AV, Rahmouni K, and Sheffield VC. Ectopic expression of human BBS4 can rescue Bardet Biedl syndrome phenotypes in Bbs4 null mice. PLoS One 8: e59101, 2013. 43 Chang B, Hawes NL, Hurd RE, Davisson MT, Nusinowitz S, and Heckenlively JR. Retinal degeneration mutants in the mouse. Vision Res 42: 517 525, 2002. 44. Chang B, Hawes NL, Hurd RE, Wang J, Howell D, Davisson MT, Roderick TH, Nusinowitz S, and Heckenlivel y JR. Mouse models of ocular diseases. Vis Neurosci 22: 587 593, 2005. 45. Chih B, Liu P, Chinn Y, Chalouni C, Komuves LG, Hass PE, Sandoval W, and Peterson AS. A ciliopathy complex at the transition zone protects the cilia as a privileged membrane domain. Nat Cell Biol 14: 61 72, 2011. 46. Cideciyan AV, Aguirre GK, Jacobson SG, Butt OH, Schwartz SB, Swider M, Roman AJ, Sadigh S, and Hauswirth WW. Pseudo fovea formation after gene therapy for RPE65 LCA. Invest Ophthalmol Vis Sci 56: 526 537, 2014. 47. Cidec iyan AV, Hauswirth WW, Aleman TS, Kaushal S, Schwartz SB, Boye SL, Windsor EA, Conlon TJ, Sumaroka A, Pang JJ, Roman AJ, Byrne BJ, and Jacobson SG. Human RPE65 gene therapy for Leber congenital amaurosis: persistence of early visual improvements and safety at 1 year. Hum Gene Ther 20: 999 1004, 2009. 48. Cideciyan AV, Hauswirth WW, Aleman TS, Kaushal S, Schwartz SB, Boye SL, Windsor EA, Conlon TJ, Sumaroka A, Roman AJ, Byrne BJ, and Jacobson SG. Vision 1 year after gene therapy for Leber's congenital amauro sis. N Engl J Med 361: 725 727, 2009. 49. Cideciyan AV, Jacobson SG, Beltran WA, Sumaroka A, Swider M, Iwabe S, Roman AJ, Olivares MB, Schwartz SB, Komromy AM, Hauswirth WW, and Aguirre GD. Human retinal gene therapy for Leber congenital amaurosis shows a dvancing retinal degeneration despite enduring visual improvement. Proc Natl Acad Sci U S A 110: E517 525, 2013.
146 50. Cideciyan AV, Roman AJ, Jacobson SG, Yan B, Pascolini M, Charng J, Pajaro S, and Nirenberg S. Developing an Outcome Measure With High Luminance for Optogenetics Treatment of Severe Retinal Degenerations and for Gene Therapy of Cone Diseases. Invest Ophthalmol Vis Sci 57: 3211 3221, 2016. 51. Colella P, Sommella A, Marrocco E, Di Vicino U, Polishch uk E, Garcia Garrido M, Seeliger MW, Polishchuk R, and Auricchio A. Myosin7a deficiency results in reduced retinal activity which is improved by gene therapy. PLoS One 8: e72027, 2013. 52. Colombo L, Sala B, Montesano G, Pierrottet C, De Cill S, Maltese P Bertelli M, and Rossetti L. Choroidal Thickness Analysis in Patients with Usher Syndrome Type 2 Using EDI OCT. J Ophthalmol 2015: 189140, 2015. 53. Cosgrove D and Zallocchi M. Clarin 1 protein expression in photoreceptors. Hear Res 259: 117, 2010. 54. Co sgrove D and Zallocchi M. Usher protein functions in hair cells and photoreceptors. Int J Biochem Cell Biol 46: 80 89, 2014. 55. Cremers FP, Kimberling WJ, Klm M, de Brouwer AP, van Wijk E, te Brinke H, Cremers CW, Hoefsloot LH, Banfi S, Simonelli F, Flei schhauer JC, Berger W, Kelley PM, Haralambous E, Bitner Glindzicz M, Webster AR, Saihan Z, De Baere E, Leroy BP, Silvestri G, McKay GJ, Koenekoop RK, Millan JM, Rosenberg T, Joensuu T, Sankila EM, Weil D, Weston MD, Wissinger B, and Kremer H. Development o f a genotyping microarray for Usher syndrome. J Med Genet 44: 153 160, 2007. 56. Czarnecki PG and Shah JV. The ciliary transition zone: from morphology and molecules to medicine. Trends Cell Biol 22: 201 210, 2012. 57. D'Amico F, Skarmoutsou E, and Stivala F. State of the art in antigen retrieval for immunohistochemistry. J Immunol Methods 341: 1 18, 2009. 58. da Cruz L, Dorn JD, Humayun MS, Dagnelie G, Handa J, Barale PO, Sahel JA, Stanga PE, Hafezi F, Safran AB, Salzmann J, Santos A, Birch D, Spencer R, C ideciyan AV, de Juan E, Duncan JL, Eliott D, Fawzi A, Olmos de Koo LC, Ho AC, Brown G, Haller J, Regillo C, Del Priore LV, Arditi A, Greenberg RJ, and Group AIS. Five Year Safety and Performance Results from the Argus II Retinal Prosthesis System Clinical Trial. Ophthalmology 123: 2248 2254, 2016. 59. Dagnelie G, Christopher P, Arditi A, da Cruz L, Duncan JL, Ho AC, Olmos de Koo LC, Sahel JA, Stanga PE, Thumann G, Wang Y, Arsiero M, Dorn JD, Greenberg RJ, and Group AIS. Performance of real world functional vision tasks by blind subjects improves after implantation with the Argus II retinal prosthesis system. Clin Exp Ophthalmol 45: 152 159, 2017.
147 60. Dalkara D, Byrne LC, Klimczak RR, Visel M, Yin L, Merigan WH, Flannery JG, and Schaffer DV. In vivo directed evolution of a new adeno associated virus for therapeutic outer retinal gene delivery from the vitreous. Sci Transl Med 5: 189ra176, 2013. 61. Dalkara D, Byrne LC, Lee T, Hoffmann NV, Schaffer DV, and Flannery JG. Enhanced gene delivery to the neonatal re tina through systemic administration of tyrosine mutated AAV9. Gene Ther 19: 176 181, 2012. 62. Dalkara D, Kolstad KD, Caporale N, Visel M, Klimczak RR, Schaffer DV, and Flannery JG. Inner limiting membrane barriers to AAV mediated retinal transduction fro m the vitreous. Mol Ther 17: 2096 2102, 2009. 63. Dalke C, Lster J, Fuchs H, Gailus Durner V, Soewarto D, Favor J, Neuhuser Klaus A, Pretsch W, Gekeler F, Shinoda K, Zrenner E, Meitinger T, Hrab de Angelis M, and Graw J. Electroretinography as a screeni ng method for mutations causing retinal dysfunction in mice. Invest Ophthalmol Vis Sci 45: 601 609, 2004. 64. Davenport JR and Yoder BK. An incredible decade for the primary cilium: a look at a once forgotten organelle. Am J Physiol Renal Physiol 289: F115 9 1169, 2005. 65. Day TP, Byrne LC, Schaffer DV, and Flannery JG. Advances in AAV vector development for gene therapy in the retina. Adv Exp Med Biol 801: 687 693, 2014. 66. Daya S and Berns KI. Gene therapy using adeno associated virus vectors. Clin Micro biol Rev 21: 583 593, 2008. 67. DE ROBERTIS E. Electron microscope observations on the submicroscopic organization of the retinal rods. J Biophys Biochem Cytol 2: 319 330, 1956. 68. Delprat B, Michel V, Goodyear R, Yamasaki Y, Michalski N, El Amraoui A, Pe rfettini I, Legrain P, Richardson G, Hardelin JP, and Petit C. Myosin XVa and whirlin, two deafness gene products required for hair bundle growth, are located at the stereocilia tips and interact directly. Hum Mol Genet 14: 401 410, 2005. 69. Di Palma F, H olme RH, Bryda EC, Belyantseva IA, Pellegrino R, Kachar B, Steel KP, and Noben Trauth K. Mutations in Cdh23, encoding a new type of cadherin, cause stereocilia disorganization in waltzer, the mouse model for Usher syndrome type 1D. Nat Genet 27: 103 107, 2 001. 70. Di Palma F, Pellegrino R, and Noben Trauth K. Genomic structure, alternative splice forms and normal and mutant alleles of cadherin 23 (Cdh23). Gene 281: 31 41, 2001.
148 71. Dinculescu A, Glushakova L, Min SH, and Hauswirth WW. Adeno associated virus vectored gene therapy for retinal disease. Hum Gene Ther 16: 649 663, 2005. 72. Dinculescu A, Stupay RM, Deng WT, Dyka FM, Min SH, Boye SL, Chiodo VA, Abrahan CE, Zhu P, Li Q, Strettoi E, Novelli E, Nagel Wolfrum K, Wolfrum U, Smith WC, and Hauswirth WW. AAV Mediated Clarin 1 Expression in the Mouse Retina: Implications for USH3A Gene Therapy. PLoS One 11: e0148874, 2016. 73. Donoso LA, Gregerson DS, Smith L, Robertson S, Knospe V, Vrabec T, and Kalsow CM. S antigen: preparation and characterization of sit e specific monoclonal antibodies. Curr Eye Res 9: 343 355, 1990. 74. Duncan JL, Richards TP, Arditi A, da Cruz L, Dagnelie G, Dorn JD, Ho AC, Olmos de Koo LC, Barale PO, Stanga PE, Thumann G, Wang Y, and Greenberg RJ. Improvements in vision related quality of life in blind patients implanted with the Argus II Epiretinal Prosthesis. Clin Exp Optom 100: 144 150, 2017. 75. Eckmiller MS. Defective cone photoreceptor cytoskeleton, alignment, feedback, and energetics can lead to energy depletion in macular degene ration. Prog Retin Eye Res 23: 495 522, 2004. 76. Eckmiller MS. Microtubules in a rod specific cytoskeleton associated with outer segment incisures. Vis Neurosci 17: 711 722, 2000. 77. Ekstrm P, Sanyal S, Narfstrm K, Chader GJ, and van Veen T. Accumulati on of glial fibrillary acidic protein in Mller radial glia during retinal degeneration. Invest Ophthalmol Vis Sci 29: 1363 1371, 1988. 78. Eley L, Yates LM, and Goodship JA. Cilia and disease. Curr Opin Genet Dev 15: 308 314, 2005. 79. Elias RV, Sezate SS, Cao W, and McGinnis JF. Temporal kinetics of the light/dark translocation and compartmentation of arrestin and alpha transducin in mouse photoreceptor cells. Mol Vis 10: 672 681, 2004. 80. Falk N, Lsl M, Schrder N, and Giel A. Specialized Cilia in M ammalian Sensory Systems. Cells 4: 500 519, 2015. 81. Ferguson LR, Balaiya S, Grover S, and Chalam KV. Modified protocol for in vivo imaging of wild type mouse retina with customized miniature spectral domain optical coherence tomography (SD OCT) device. B iol Proced Online 14: 9, 2012. 82. Flannery JG and Visel M. Adeno associated viral vectors for gene therapy of inherited retinal degenerations. Methods Mol Biol 935: 351 369, 2013.
149 83. Flannery JG, Zolotukhin S, Vaquero MI, LaVail MM, Muzyczka N, and Hausw irth WW. Efficient photoreceptor targeted gene expression in vivo by recombinant adeno associated virus. Proc Natl Acad Sci U S A 94: 6916 6921, 1997. 84. Frolenkov GI, Belyantseva IA, Friedman TB, and Griffith AJ. Genetic insights into the morphogenesis o f inner ear hair cells. Nat Rev Genet 5: 489 498, 2004. 85. Garcia Gonzalo FR and Reiter JF. Scoring a backstage pass: mechanisms of ciliogenesis and ciliary access. J Cell Biol 197: 697 709, 2012. 86. Garg A, Yang J, Lee W, and Tsang SH. Stem Cell Therapi es in Retinal Disorders. Cells 6, 2017. 87. Geller SF, Guerin KI, Visel M, Pham A, Lee ES, Dror AA, Avraham KB, Hayashi T, Ray CA, Reh TA, Bermingham McDonogh O, Triffo WJ, Bao S, Isosomppi J, Vstinsalo H, Sankila EM, and Flannery JG. CLRN1 is nonessentia l in the mouse retina but is required for cochlear hair cell development. PLoS Genet 5: e1000607, 2009. 88. Geng R, Geller SF, Hayashi T, Ray CA, Reh TA, Bermingham McDonogh O, Jones SM, Wright CG, Melki S, Imanishi Y, Palczewski K, Alagramam KN, and Flann ery JG. Usher syndrome IIIA gene clarin 1 is essential for hair cell function and associated neural activation. Hum Mol Genet 18: 2748 2760, 2009. 89. Geng R, Melki S, Chen DH, Tian G, Furness DN, Oshima Takago T, Neef J, Moser T, Askew C, Horwitz G, Holt JR, Imanishi Y, and Alagramam KN. The mechanosensory structure of the hair cell requires clarin 1, a protein encoded by Usher syndrome III causative gene. J Neurosci 32: 9485 9498, 2012. 90. Geng R, Omar A, Gopal SR, Chen DH, Stepanyan R, Basch ML, Dincule scu A, Furness DN, Saperstein D, Hauswirth W, Lustig LR, and Alagramam KN. Modeling and Preventing Progressive Hearing Loss in Usher Syndrome III. Sci Rep 7: 13480, 2017. 91. Geruschat DR, Richards TP, Arditi A, da Cruz L, Dagnelie G, Dorn JD, Duncan JL, H o AC, Olmos de Koo LC, Sahel JA, Stanga PE, Thumann G, Wang V, and Greenberg RJ. An analysis of observer rated functional vision in patients implanted with the Argus II Retinal Prosthesis System at three years. Clin Exp Optom 99: 227 232, 2016. 92. Glckle N, Kohl S, Mohr J, Scheurenbrand T, Sprecher A, Weisschuh N, Bernd A, Rudolph G, Schubach M, Poloschek C, Zrenner E, Biskup S, Berger W, Wissinger B, and Neidhardt J. Panel based next generation sequencing as a reliable and efficient technique to detect m utations in unselected patients with retinal dystrophies. Eur J Hum Genet 22: 99 104, 2014.
150 93. Gopal SR, Chen DH, Chou SW, Zang J, Neuhauss SC, Stepanyan R, McDermott BM, and Alagramam KN. Zebrafish Models for the Mechanosensory Hair Cell Dysfunction in U sher Syndrome 3 Reveal That Clarin 1 Is an Essential Hair Bundle Protein. J Neurosci 35: 10188 10201, 2015. 94. Gorlin RJ, Tilsner TJ, Feinstein S, and Duvall AJ. Usher's syndrome type III. Arch Otolaryngol 105: 353 354, 1979. 95. Govardovskii VI, Calvert PD, and Arshavsky VY. Photoreceptor light adaptation. Untangling desensitization and sensitization. J Gen Physiol 116: 791 794, 2000. 96. Greenberg KP, Lee ES, Schaffer DV, and Flannery JG. Gene delivery to the retina using lentiviral vectors. Adv Exp Med Biol 572: 255 266, 2006. 97. Gurevich VV, Hanson SM, Song X, Vishnivetskiy SA, and Gurevich EV. The functional cycle of visual arrestins in photoreceptor cells. Prog Retin Eye Res 30: 405 430, 2011. 98. Hanson SM, Cleghorn WM, Francis DJ, Vishnivetskiy SA, Raman D, Song X, Nair KS, Slepak VZ, Klug CS, and Gurevich VV. Arrestin mobilizes signaling proteins to the cytoskeleton and redirects their activity. J Mol Biol 368: 375 387, 2007. 99. Hao W, Wenzel A, Obin MS, Chen CK, Brill E, Krasnoperova NV, Eversole Cire P, Kleyner Y, Taylor A, Simon MI, Grimm C, Rem CE, and Lem J. Evidence for two apoptotic pathways in light induced retinal degeneration. Nat Genet 32: 254 260, 2002. 100. Harris BZ and Lim WA. Mechanism and role of PDZ domains in signaling complex a ssembly. J Cell Sci 114: 3219 3231, 2001. 101. Hata Y, Nakanishi H, and Takai Y. Synaptic PDZ domain containing proteins. Neurosci Res 32: 1 7, 1998. 102. Hauswirth WW, Aleman TS, Kaushal S, Cideciyan AV, Schwartz SB, Wang L, Conlon TJ, Boye SL, Flotte TR, Byrne BJ, and Jacobson SG. Treatment of leber congenital amaurosis due to RPE65 mutations by ocular subretinal injection of adeno associated virus gene vector: short term results of a phase I trial. Hum Gene Ther 19: 979 990, 2008. 103. Hauswirth WW, Lewi n AS, Zolotukhin S, and Muzyczka N. Production and purification of recombinant adeno associated virus. Methods Enzymol 316: 743 761, 2000.
151 104. Hauswirth WW, Li Q, Raisler B, Timmers AM, Berns KI, Flannery JG, LaVail MM, and Lewin AS. Range of retinal dis eases potentially treatable by AAV vectored gene therapy. Novartis Found Symp 255: 179 188; discussion 188 194, 2004. 105. Haywood Watson RJ, Ahmed ZM, Kjellstrom S, Bush RA, Takada Y, Hampton LL, Battey JF, Sieving PA, and Friedman TB. Ames Waltzer deaf m ice have reduced electroretinogram amplitudes and complex alternative splicing of Pcdh15 transcripts. Invest Ophthalmol Vis Sci 47: 3074 3084, 2006. 106. Helga Kolb EF and Ralph N. Webvision : the organization of the retina and visual system : [Bethesda, Md.] : National Library of Medicine : [National Center for Biotechnology Information], 2007., 2007. 107. Herrera W, Aleman TS, Cideciyan AV, Roman AJ, Banin E, Ben Yosef T, Gardner LM, Sumaroka A, Windsor EA, Schwartz SB, Stone EM, Liu XZ, Kimberling WJ, a nd Jacobson SG. Retinal disease in Usher syndrome III caused by mutations in the clarin 1 gene. Invest Ophthalmol Vis Sci 49: 2651 2660, 2008. 108. Herzog RW and Zolotukhin S. Guide to Human Gene Therapy: World Scientific, 2010. 109. Hildebrandt F, Benzing T, and Katsanis N. Ciliopathies. N Engl J Med 364: 1533 1543, 2011. 110. Huang D, Swanson EA, Lin CP, Schuman JS, Stinson WG, Chang W, Hee MR, Flotte T, Gregory K, and Puliafito CA. Optical coherence tomography. Science 254: 1178 1181, 1991. 111. Hudspeth AJ. Integrating the active process of hair cells with cochlear function. Nat Rev Neurosci 15: 600 614, 2014. 112. Insinna C and Besharse JC. Intraflagellar transport and the sensory outer segment of vertebrate photoreceptors. Dev Dyn 237: 1982 1992, 2008. 113. Ishikawa H and Marshall WF. Ciliogenesis: building the cell's antenna. Nat Rev Mol Cell Biol 12: 222 234, 2011. 114. Isosomppi J, Vstinsalo H, Geller SF, Heon E, Flannery JG, and Sankila EM. Disease causing mutations in the CLRN1 gene alter normal C LRN1 protein trafficking to the plasma membrane. Mol Vis 15: 1806 1818, 2009. 115. Jacobson SG, Acland GM, Aguirre GD, Aleman TS, Schwartz SB, Cideciyan AV, Zeiss CJ, Komaromy AM, Kaushal S, Roman AJ, Windsor EA, Sumaroka A, Pearce Kelling SE, Conlon TJ, C hiodo VA, Boye SL, Flotte TR, Maguire AM, Bennett J, and Hauswirth WW. Safety of recombinant adeno
152 associated virus type 2 RPE65 vector delivered by ocular subretinal injection. Mol Ther 13: 1074 1084, 2006. 116. Jacobson SG, Cideciyan AV, Aguirre GD, Roma n AJ, Sumaroka A, Hauswirth WW, and Palczewski K. Improvement in vision: a new goal for treatment of hereditary retinal degenerations. Expert Opin Orphan Drugs 3: 563 575, 2015. 117. Jacobson SG, Cideciyan AV, Ratnakaram R, Heon E, Schwartz SB, Roman AJ, P eden MC, Aleman TS, Boye SL, Sumaroka A, Conlon TJ, Calcedo R, Pang JJ, Erger KE, Olivares MB, Mullins CL, Swider M, Kaushal S, Feuer WJ, Iannaccone A, Fishman GA, Stone EM, Byrne BJ, and Hauswirth WW. Gene therapy for leber congenital amaurosis caused by RPE65 mutations: safety and efficacy in 15 children and adults followed up to 3 years. Arch Ophthalmol 130: 9 24, 2012. 118. Jacobson SG, Matsui R, Sumaroka A, and Cideciyan AV. Retinal Structure Measurements as Inclusion Criteria for Stem Cell Based Thera pies of Retinal Degenerations. Invest Ophthalmol Vis Sci 57: ORSFn1 9, 2016. 119. Jacobson SG, Sumaroka A, Luo X, and Cideciyan AV. Retinal optogenetic therapies: clinical criteria for candidacy. Clin Genet 84: 175 182, 2013. 120. Jansen F, Kalbe B, Scholz P, Mikosz M, Wunderlich KA, Kurtenbach S, Nagel Wolfrum K, Wolfrum U, Hatt H, and Osterloh S. Impact of the Usher syndrome on olfaction. Hum Mol Genet 25: 524 533, 2016. 121. Jenkins PM, McEwen DP, and Martens JR. Olfactory cilia: linking sensory cilia fu nction and human disease. Chem Senses 34: 451 464, 2009. 122. Jin H, White SR, Shida T, Schulz S, Aguiar M, Gygi SP, Bazan JF, and Nachury MV. The conserved Bardet Biedl syndrome proteins assemble a coat that traffics membrane proteins to cilia. Cell 141: 1208 1219, 2010. 123. Joensuu T, Blanco G, Pakarinen L, Sistonen P, Kriinen H, Brown S, Chapelle A, and Sankila EM. Refined mapping of the Usher syndrome type III locus on chromosome 3, exclusion of candidate genes, and identification of the putative mo use homologous region. Genomics 38: 255 263, 1996. 124. Joensuu T, Hmlinen R, Lehesjoki AE, de la Chapelle A, and Sankila EM. A sequence ready map of the Usher syndrome type III critical region on chromosome 3q. Genomics 63: 409 416, 2000. 125. Joiner AM, Green WW, McIntyre JC, Allen BL, Schwob JE, and Martens JR. Primary Cilia on Horizontal Basal Cells Regulate Regeneration of the Olfactory Epithelium. J Neurosci 35: 13761 13772, 2015.
153 126. Kalloniatis M, Nivison Smith L, Chua J, Acosta ML, and Fletche r EL. Using the rd1 mouse to understand functional and anatomical retinal remodelling and treatment implications in retinitis pigmentosa: A review. Exp Eye Res 150: 106 121, 2016. 127. Katsanis N. The continuum of causality in human genetic disorders. Geno me Biol 17: 233, 2016. 128. Kay CN, Ryals RC, Aslanidi GV, Min SH, Ruan Q, Sun J, Dyka FM, Kasuga D, Ayala AE, Van Vliet K, Agbandje McKenna M, Hauswirth WW, Boye SL, and Boye SE. Targeting photoreceptors via intravitreal delivery using novel, capsid mutat ed AAV vectors. PLoS One 8: e62097, 2013. 129. Kazmierczak P, Sakaguchi H, Tokita J, Wilson Kubalek EM, Milligan RA, Mller U, and Kachar B. Cadherin 23 and protocadherin 15 interact to form tip link filaments in sensory hair cells. Nature 449: 87 91, 2007 130. Kee HL, Dishinger JF, Blasius TL, Liu CJ, Margolis B, and Verhey KJ. A size exclusion permeability barrier and nucleoporins characterize a ciliary pore complex that regulates transport into cilia. Nat Cell Biol 14: 431 437, 2012. 131. Khan MI, Kerst en FF, Azam M, Collin RW, Hussain A, Shah ST, Keunen JE, Kremer H, Cremers FP, Qamar R, and den Hollander AI. CLRN1 mutations cause nonsyndromic retinitis pigmentosa. Ophthalmology 118: 1444 1448, 2011. 132. Khan SH, Javed MR, Qasim M, Shahzadi S, Jalil A, and Rehman SU. Domain analyses of Usher syndrome causing Clarin 1 and GPR98 protein models. Bioinformation 10: 491 495, 2014. 133. Kim YK, Kim JH, Yu YS, and Ko HW. Localization of primary cilia in mouse retina. Acta Histochem 115: 789 794, 2013. 134. Kol stad KD, Dalkara D, Guerin K, Visel M, Hoffmann N, Schaffer DV, and Flannery JG. Changes in adeno associated virus mediated gene delivery in retinal degeneration. Hum Gene Ther 21: 571 578, 2010. 135. Kremer H, van Wijk E, Mrker T, Wolfrum U, and Roepman R. Usher syndrome: molecular links of pathogenesis, proteins and pathways. Hum Mol Genet 15 Spec No 2: R262 270, 2006. 136. Krishnamoorthy V, Cherukuri P, Poria D, Goel M, Dagar S, and Dhingra NK. Retinal Remodeling: Concerns, Emerging Remedies and Future Prospects. Front Cell Neurosci 10: 38, 2016. 137. Ku CA, Chiodo VA, Boye SL, Goldberg AF, Li T, Hauswirth WW, and Ramamurthy V. Gene therapy using self complementary Y733F capsid mutant AAV2/8 restores vision in a model of early onset Leber congenital amau rosis. Hum Mol Genet 20: 4569 4581, 2011.
154 138. Lagziel A, Ahmed ZM, Schultz JM, Morell RJ, Belyantseva IA, and Friedman TB. Spatiotemporal pattern and isoforms of cadherin 23 in wild type and waltzer mice during inner ear hair cell development. Dev Biol 28 0: 295 306, 2005. 139. Lagziel A, Overlack N, Bernstein SL, Morell RJ, Wolfrum U, and Friedman TB. Expression of cadherin 23 isoforms is not conserved: implications for a mouse model of Usher syndrome type 1D. Mol Vis 15: 1843 1857, 2009. 140. Lam S, Cao H Wu J, Duan R, and Hu J. Highly efficient retinal gene delivery with helper dependent adenoviral vectors. Genes Dis 1: 227 237, 2014. 141. Langham ME and Kramer T. Decreased choroidal blood flow associated with retinitis pigmentosa. Eye (Lond) 4 ( Pt 2): 374 381, 1990. 142. Libby RT, Kitamoto J, Holme RH, Williams DS, and Steel KP. Cdh23 mutations in the mouse are associated with retinal dysfunction but not retinal degeneration. Exp Eye Res 77: 731 739, 2003. 143. Lindemann CB and Lesich KA. Flagellar and ciliary beating: the proven and the possible. J Cell Sci 123: 519 528, 2010. 144. Lopes VS, Boye SE, Louie CM, Boye S, Dyka F, Chiodo V, Fofo H, Hauswirth WW, and Williams DS. Retinal gene therapy with a large MYO7A cDNA using adeno associated virus. Gene Ther 20: 824 833, 2013. 145. Luby Phelps K, Fogerty J, Baker SA, Pazour GJ, and Besharse JC. Spatial distribution of intraflagellar transport proteins in vertebrate photoreceptors. Vision Res 48: 413 423, 2008. 146. Lukacs GL. Proteostasis: Chaperoning for hearing loss. Nat Chem Biol 12: 388 389, 2016. 147. Maerker T, van Wijk E, Overlack N, Kersten FF, McGee J, Goldmann T, Sehn E, Roepman R, Walsh EJ, Kremer H, and Wolfrum U. A novel Usher protein network at the periciliary reloading point between molecular transport machineries in vertebrate photoreceptor cells. Hum Mol Genet 17: 71 86, 2008. 148. Mathur P and Yang J. Usher syndrome: Hearing loss, retinal degeneration and ass ociated abnormalities. Biochim Biophys Acta 1852: 406 420, 2015. 149. Mburu P, Mustapha M, Varela A, Weil D, El Amraoui A, Holme RH, Rump A, Hardisty RE, Blanchard S, Coimbra RS, Perfettini I, Parkinson N, Mallon AM, Glenister P, Rogers MJ, Paige AJ, Moir L, Clay J, Rosenthal A, Liu XZ, Blanco G, Steel KP, Petit C, and Brown SD. Defects in whirlin, a PDZ domain molecule involved in stereocilia elongation, cause deafness in the whirler mouse and families with DFNB31. Nat Genet 34: 421 428, 2003.
155 150. McCall MA and Gregg RG. Comparisons of structural and functional abnormalities in mouse b wave mutants. J Physiol 586: 4385 4392, 2008. 151. McEwen DP, Jenkins PM, and Martens JR. Olfactory cilia: our direct neuronal connection to the external world. Curr Top Dev Biol 85: 333 370, 2008. 152. McGee J, Goodyear RJ, McMillan DR, Stauffer EA, Holt JR, Locke KG, Birch DG, Legan PK, White PC, Walsh EJ, and Richardson GP. The very large G protein coupled receptor VLGR1: a component of the ankle link complex required for the normal development of auditory hair bundles. J Neurosci 26: 6543 6553, 2006. 153. McIntyre JC, Davis EE, Joiner A, Williams CL, Tsai IC, Jenkins PM, McEwen DP, Zhang L, Escobado J, Thomas S, Szymanska K, Johnson CA, Beales PL, Green ED, Mullikin JC, Sa bo A, Muzny DM, Gibbs RA, Atti Bitach T, Yoder BK, Reed RR, Katsanis N, Martens JR, and Program NCS. Gene therapy rescues cilia defects and restores olfactory function in a mammalian ciliopathy model. Nat Med 18: 1423 1428, 2012. 154. McIntyre JC, William s CL, and Martens JR. Smelling the roses and seeing the light: gene therapy for ciliopathies. Trends Biotechnol 31: 355 363, 2013. 155. McLean WJ, Yin X, Lu L, Lenz DR, McLean D, Langer R, Karp JM, and Edge AS. Clonal Expansion of Lgr5 Positive Cells from Mammalian Cochlea and High Purity Generation of Sensory Hair Cells. Cell Rep 18: 1917 1929, 2017. 156. Menini A, Lagostena L, and Boccaccio A. Olfaction: from odorant molecules to the olfactory cortex. News Physiol Sci 19: 101 104, 2004. 157. Michalski N, Michel V, Bahloul A, Lefvre G, Barral J, Yagi H, Chardenoux S, Weil D, Martin P, Hardelin JP, Sato M, and Petit C. Molecular characterization of the ankle link complex in cochlear hair cells and its role in the hair bundle functioning. J Neurosci 27: 6478 6488, 2007. 158. Milln JM, Aller E, Jaijo T, Blanco Kelly F, Gimenez Pardo A, and Ayuso C. An update on the genetics of usher syndrome. J Ophthalmol 2011: 417217, 2011. 159. Mouro A, Christensen ST, and Lorentzen E. The intraflagellar transport machiner y in ciliary signaling. Curr Opin Struct Biol 41: 98 108, 2016. 160. Mustafi D, Maeda T, Kohno H, Nadeau JH, and Palczewski K. Inflammatory priming predisposes mice to age related retinal degeneration. J Clin Invest 122: 2989 3001, 2012. 161. Nachury MV, S eeley ES, and Jin H. Trafficking to the ciliary membrane: how to get across the periciliary diffusion barrier? Annu Rev Cell Dev Biol 26: 59 87, 2010.
156 162. Nair KS, Hanson SM, Kennedy MJ, Hurley JB, Gurevich VV, and Slepak VZ. Direct binding of visual arre stin to microtubules determines the differential subcellular localization of its splice variants in rod photoreceptors. J Biol Chem 279: 41240 41248, 2004. 163. Nair KS, Hanson SM, Mendez A, Gurevich EV, Kennedy MJ, Shestopalov VI, Vishnivetskiy SA, Chen J Hurley JB, Gurevich VV, and Slepak VZ. Light dependent redistribution of arrestin in vertebrate rods is an energy independent process governed by protein protein interactions. Neuron 46: 555 567, 2005. 164. Najafi M, Maza NA, and Calvert PD. Steric volume exclusion sets soluble protein concentrations in photoreceptor sensory cilia. Proc Natl Acad Sci U S A 109: 203 208, 2012. 165. Naso MF, Tomkowicz B, Perry WL, and Strohl WR. Adeno Associated Virus (AAV) as a Vector for Gene Therapy. BioDrug s 2017. 166. Nishio SY, Takumi Y, and Usami SI. Laser capture micro dissection combined with next generation sequencing analysis of cell type specific deafness gene expression in the mouse cochlea. Hear Res 348: 87 97, 2017. 167. Noben Trauth K, Zheng QY, and Johnson KR. Association of cadherin 23 with polygenic inheritance and genetic modification of sensorineural hearing loss. Nat Genet 35: 21 23, 2003. 168. O'Connor AK, Malarkey EB, Berbari NF, Croyle MJ, Haycraft CJ, Bell PD, Hohenstein P, Kesterson RA and Yoder BK. An inducible CiliaGFP mouse model for in vivo visualization and analysis of cilia in live tissue. Cilia 2: 8, 2013. 169. Ogun O and Zallocchi M. Clarin 1 acts as a modulator of mechanotransduction activity and presynaptic ribbon assembly. J Cell Biol 207: 375 391, 2014. 170. Overlack N, Maerker T, Latz M, Nagel Wolfrum K, and Wolfrum U. SANS (USH1G) expression in developing and mature mammalian retina. Vision Res 48: 400 412, 2008. 171. Pakarinen L, Karjalainen S, Simola KO, Laippala P, and Kaitalo H. Usher's syndrome type 3 in Finland. Laryngoscope 105: 613 617, 1995. 172. Pakarinen L, Tuppurainen K, Laippala P, Mntyjrvi M, and Puhakka H. The ophthalmological course of Usher syndrome type III. Int Ophthalmol 19: 307 311, 1995. 173. Pan B, Askew C, Galvin A, Heman Ackah S, Asai Y, Indzhykulian AA, Jodelka FM, Hastings ML, Lentz JJ, Vandenberghe LH, Holt JR, and Gloc GS. Gene therapy restores auditory and vestibular function in a mouse model of Usher syndrome type 1c. Nat Biotechnol 35: 264 272, 2017.
157 174. Parmeggiani F, De Nadai K, Piovan A, Binotto A, Zamengo S, and Chizzolini M. Optical coherence tomography imaging in the management of the Argus II retinal prosthesis system. Eur J Ophthalmol 27: e16 e21, 2017. 175. Peachey NS and Ball SL. Electrophysiological analysis of visual function in mutant mice. Doc Ophthalmol 107: 13 36, 2003. 176. Peng YW, Zallocchi M, Wang WM, Delimont D, and Cosgrove D. Moderate light induced degeneration of rod photoreceptors with delayed transducin translocati on in shaker1 mice. Invest Ophthalmol Vis Sci 52: 6421 6427, 2011. 177. Pennesi ME, Michaels KV, Magee SS, Maricle A, Davin SP, Garg AK, Gale MJ, Tu DC, Wen Y, Erker LR, and Francis PJ. Long term characterization of retinal degeneration in rd1 and rd10 mic e using spectral domain optical coherence tomography. Invest Ophthalmol Vis Sci 53: 4644 4656, 2012. 178. Petit C. Usher syndrome: from genetics to pathogenesis. Annu Rev Genomics Hum Genet 2: 271 297, 2001. 179. Petrs Silva H, Dinculescu A, Li Q, Deng WT, Pang JJ, Min SH, Chiodo V, Neeley AW, Govindasamy L, Bennett A, Agbandje McKenna M, Zhong L, Li B, Jayandharan GR, Srivastava A, Lewin AS, and Hauswirth WW. Novel properties of tyrosine mutant AAV2 vectors in the mouse retina. Mol Ther 19: 293 301, 2011. 180. Petrs Silva H, Dinculescu A, Li Q, Min SH, Chiodo V, Pang JJ, Zhong L, Zolotukhin S, Srivastava A, Lewin AS, and Hauswirth WW. High efficiency transduction of the mouse retina by tyrosine mutant AAV serotype vectors. Mol Ther 17: 463 471, 2009. 181. P etrs Silva H and Linden R. Advances in gene therapy technologies to treat retinitis pigmentosa. Clin Ophthalmol 8: 127 136, 2014. 182. Phillips JB, Vstinsalo H, Wegner J, Clment A, Sankila EM, and Westerfield M. The cone dominant retina and the inner ear of zebrafish express the ortholog of CLRN1, the causative gene of human Usher syndrome type 3A. Gene Expr Patterns 13: 473 481, 2013. 183. Pietola L, Aarnisalo AA, Abdel Rahman A, Vstinsalo H, Isosomppi J, Lppnen H, Kentala E, Johansson R, Valtonen H, Vasama JP, Sankila EM, and Jero J. Speech recognition and communication outcomes with cochlear implantation in Usher syndrome type 3. Otol Neurotol 33: 38 41, 2012. 184. Pinto LH, Invergo B, Shimomura K, Takahashi JS, and Troy JB. Interpretation of the mou se electroretinogram. Doc Ophthalmol 115: 127 136, 2007.
158 185. Plantinga RF, Kleemola L, Huygen PL, Joensuu T, Sankila EM, Pennings RJ, and Cremers CW. Serial audiometry and speech recognition findings in Finnish Usher syndrome type III patients. Audiol Neu rootol 10: 79 89, 2005. 186. Plantinga RF, Pennings RJ, Huygen PL, Sankila EM, Tuppurainen K, Kleemola L, Cremers CW, and Deutman AF. Visual impairment in Finnish Usher syndrome type III. Acta Ophthalmol Scand 84: 36 41, 2006. 187. Rachel RA, Li T, and Swa roop A. Photoreceptor sensory cilia and ciliopathies: focus on CEP290, RPGR and their interacting proteins. Cilia 1: 22, 2012. 188. Ramamurthy V and Cayouette M. Development and disease of the photoreceptor cilium. Clin Genet 76: 137 145, 2009. 189. Ratnam K, Vstinsalo H, Roorda A, Sankila EM, and Duncan JL. Cone structure in patients with usher syndrome type III and mutations in the Clarin 1 gene. JAMA Ophthalmol 131: 67 74, 2013. 190. Reidel B, Goldmann T, Giessl A, and Wolfrum U. The translocation of si gnaling molecules in dark adapting mammalian rod photoreceptor cells is dependent on the cytoskeleton. Cell Motil Cytoskeleton 65: 785 800, 2008. 191. Reiners J, Mrker T, Jrgens K, Reidel B, and Wolfrum U. Photoreceptor expression of the Usher syndrome t ype 1 protein protocadherin 15 (USH1F) and its interaction with the scaffold protein harmonin (USH1C). Mol Vis 11: 347 355, 2005. 192. Reiners J, Nagel Wolfrum K, Jrgens K, Mrker T, and Wolfrum U. Molecular basis of human Usher syndrome: deciphering the meshes of the Usher protein network provides insights into the pathomechanisms of the Usher disease. Exp Eye Res 83: 97 119, 2006. 193. Reiners J, Reidel B, El Amraoui A, Boda B, Huber I, Petit C, and Wolfrum U. Differential distribution of harmonin isoforms and their possible role in Usher 1 protein complexes in mammalian photoreceptor cells. Invest Ophthalmol Vis Sci 44: 5006 5015, 2003. 194. Reiners J, van Wijk E, Mrker T, Zimmermann U, Jrgens K, te Brinke H Overlack N, Roepman R, Knipper M, Kremer H, and Wolfrum U. Scaffold protein harmonin (USH1C) provides molecular links between Usher syndrome type 1 and type 2. Hum Mol Genet 14: 3933 3943, 2005. 195. Reiter JF, Blacque OE, and Leroux MR. The base of the cilium: roles for transition fibres and the transition zone in ciliary formation, maintenance and compartmentalization. EMBO Rep 13: 608 618, 2012.
159 196. Ribeiro JC, Oliveiros B, Pereira P, Antnio N, Hummel T, Paiva A, and Silva ED. Accelerated age related olfactory decline among type 1 Usher patients. Sci Rep 6: 28309, 2016. 197. Ryals RC, Boye SL, Dinculescu A, Hauswirth WW, and Boye SE. Quantifying transduction efficiencies of unmodified and tyrosine capsid mutant AAV vectors in vitro using two ocular ce ll lines. Mol Vis 17: 1090 1102, 2011. 198. Sahly I, Dufour E, Schietroma C, Michel V, Bahloul A, Perfettini I, Pepermans E, Estivalet A, Carette D, Aghaie A, Ebermann I, Lelli A, Iribarne M, Hardelin JP, Weil D, Sahel JA, El Amraoui A, and Petit C. Locali zation of Usher 1 proteins to the photoreceptor calyceal processes, which are absent from mice. J Cell Biol 199: 381 399, 2012. 199. Saihan Z, Webster AR, Luxon L, and Bitner Glindzicz M. Update on Usher syndrome. Curr Opin Neurol 22: 19 27, 2009. 200. San kila EM, Pakarinen L, Kriinen H, Aittomki K, Karjalainen S, Sistonen P, and de la Chapelle A. Assignment of an Usher syndrome type III (USH3) gene to chromosome 3q. Hum Mol Genet 4: 93 98, 1995. 201. Schultz JM, Bhatti R, Madeo AC, Turriff A, Muskett J A, Zalewski CK, King KA, Ahmed ZM, Riazuddin S, Ahmad N, Hussain Z, Qasim M, Kahn SN, Meltzer MR, Liu XZ, Munisamy M, Ghosh M, Rehm HL, Tsilou ET, Griffith AJ, Zein WM, Brewer CC, and Friedman TB. Allelic hierarchy of CDH23 mutations causing non syndromic deafness DFNB12 or Usher syndrome USH1D in compound heterozygotes. J Med Genet 48: 767 775, 2011. 202. Sedmak T and Wolfrum U. Intraflagellar transport molecules in ciliary and nonciliary cells of the retina. J Cell Biol 189: 171 186, 2010. 203. Sedmak T a nd Wolfrum U. Intraflagellar transport proteins in ciliogenesis of photoreceptor cells. Biol Cell 103: 449 466, 2011. 204. Senften M, Schwander M, Kazmierczak P, Lillo C, Shin JB, Hasson T, Gloc GS, Gillespie PG, Williams D, Holt JR, and Mller U. Physic al and functional interaction between protocadherin 15 and myosin VIIa in mechanosensory hair cells. J Neurosci 26: 2060 2071, 2006. 205. Seo S, Mullins RF, Dumitrescu AV, Bhattarai S, Gratie D, Wang K, Stone EM, Sheffield V, and Drack AV. Subretinal gene therapy of mice with Bardet Biedl syndrome type 1. Invest Ophthalmol Vis Sci 54: 6118 6132, 2013. 206. Shichida Y and Morizumi T. Mechanism of G protein activation by rhodopsin. Photochem Photobiol 83: 70 75, 2007.
160 207. Siegert S, Cabuy E, Scherf BG, Kohle r H, Panda S, Le YZ, Fehling HJ, Gaidatzis D, Stadler MB, and Roska B. Transcriptional code and disease map for adult retinal cell types. Nat Neurosci 15: 487 495, S481 482, 2012. 208. Siemens J, Kazmierczak P, Reynolds A, Sticker M, Littlewood Evans A, an d Mller U. The Usher syndrome proteins cadherin 23 and harmonin form a complex by means of PDZ domain interactions. Proc Natl Acad Sci U S A 99: 14946 14951, 2002. 209. Simons DL, Boye SL, Hauswirth WW, and Wu SM. Gene therapy prevents photoreceptor death and preserves retinal function in a Bardet Biedl syndrome mouse model. Proc Natl Acad Sci U S A 108: 6276 6281, 2011. 210. Slepak VZ and Hurley JB. Mechanism of light induced translocation of arrestin and transducin in photoreceptors: interaction restrict ed diffusion. IUBMB Life 60: 2 9, 2008. 211. Sorusch N, Wunderlich K, Bauss K, Nagel Wolfrum K, and Wolfrum U. Usher syndrome protein network functions in the retina and their relation to other retinal ciliopathies. Adv Exp Med Biol 801: 527 533, 2014. 212 Sumaroka A, Matsui R, Cideciyan AV, McGuigan DB, Sheplock R, Schwartz SB, and Jacobson SG. Outer Retinal Changes Including the Ellipsoid Zone Band in Usher Syndrome 1B due to MYO7A Mutations. Invest Ophthalmol Vis Sci 57: OCT253 261, 2016. 213. Sun LW, J ohnson RD, Langlo CS, Cooper RF, Razeen MM, Russillo MC, Dubra A, Connor TB, Han DP, Pennesi ME, Kay CN, Weinberg DV, Stepien KE, and Carroll J. Assessing Photoreceptor Structure in Retinitis Pigmentosa and Usher Syndrome. Invest Ophthalmol Vis Sci 57: 242 8 2442, 2016. 214. Sung CH and Chuang JZ. The cell biology of vision. J Cell Biol 190: 953 963, 2010. 215. Sung CH and Tai AW. Rhodopsin trafficking and its role in retinal dystrophies. Int Rev Cytol 195: 215 267, 2000. 216. Szymanska K and Johnson CA. The transition zone: an essential functional compartment of cilia. Cilia 1: 10, 2012. 217. Tian G, Lee R, Ropelewski P, and Imanishi Y. Impairment of Vision in a Mouse Model of Usher Syndrome Type III. Invest Ophthalmol Vis Sci 57: 866 875, 2016. 218. Tia n G, Zhou Y, Hajkova D, Miyagi M, Dinculescu A, Hauswirth WW, Palczewski K, Geng R, Alagramam KN, Isosomppi J, Sankila EM, Flannery JG, and Imanishi Y. Clarin 1, encoded by the Usher Syndrome III causative gene, forms a membranous microdomain: possible rol e of clarin 1 in organizing the actin cytoskeleton. J Biol Chem 284: 18980 18993, 2009.
161 219. Tian M, Wang W, Delimont D, Cheung L, Zallocchi M, Cosgrove D, and Peng YW. Photoreceptors in whirler mice show defective transducin translocation and are suscepti ble to short term light/dark changes induced degeneration. Exp Eye Res 118: 145 153, 2014. 220. Tian M, Zallocchi M, Wang W, Chen CK, Palczewski K, Delimont D, Cosgrove D, and Peng YW. Light induced translocation of RGS9 mouse rod photorecept ors. PLoS One 8: e58832, 2013. 221. Toth CA, Narayan DG, Boppart SA, Hee MR, Fujimoto JG, Birngruber R, Cain CP, DiCarlo CD, and Roach WP. A comparison of retinal morphology viewed by optical coherence tomography and by light microscopy. Arch Ophthalmol 11 5: 1425 1428, 1997. 222. Trapani I, Puppo A, and Auricchio A. Vector platforms for gene therapy of inherited retinopathies. Prog Retin Eye Res 43: 108 128, 2014. 223. Ueki Y, Chollangi S, Le YZ, and Ash JD. gp130 activation in Mller cells is not essential for photoreceptor protection from light damage. Adv Exp Med Biol 664: 655 661, 2010. 224. Ueki Y, Wang J, Chollangi S, and Ash JD. STAT3 activation in photoreceptors by leukemia inhibitory factor is associ ated with protection from light damage. J Neurochem 105: 784 796, 2008. 225. Vstinsalo H, Jalkanen R, Bergmann C, Neuhaus C, Kleemola L, Jauhola L, Bolz HJ, and Sankila EM. Extended mutation spectrum of Usher syndrome in Finland. Acta Ophthalmol 91: 325 3 34, 2013. 226. Vstinsalo H, Jalkanen R, Dinculescu A, Isosomppi J, Geller S, Flannery JG, Hauswirth WW, and Sankila EM. Alternative splice variants of the USH3A gene Clarin 1 (CLRN1). Eur J Hum Genet 19: 30 35, 2011. 227. Wachtmeister L. Oscillatory poten tials in the retina: what do they reveal. Prog Retin Eye Res 17: 485 521, 1998. 228. Wei Q, Ling K, and Hu J. The essential roles of transition fibers in the context of cilia. Curr Opin Cell Biol 35: 98 105, 2015. 229. Weil D, El Amraoui A, Masmoudi S, Mus tapha M, Kikkawa Y, Lain S, Delmaghani S, Adato A, Nadifi S, Zina ZB, Hamel C, Gal A, Ayadi H, Yonekawa H, and Petit C. Usher syndrome type I G (USH1G) is caused by mutations in the gene encoding SANS, a protein that associates with the USH1C protein, har monin. Hum Mol Genet 12: 463 471, 2003. 230. Wheway G, Parry DA, and Johnson CA. The role of primary cilia in the development and disease of the retina. Organogenesis 10: 69 85, 2014.
162 231. Williams CL, Li C, Kida K, Inglis PN, Mohan S, Semenec L, Bialas NJ Stupay RM, Chen N, Blacque OE, Yoder BK, and Leroux MR. MKS and NPHP modules cooperate to establish basal body/transition zone membrane associations and ciliary gate function during ciliogenesis. J Cell Biol 192: 1023 1041, 2011. 232. Williams CL, Uyting co CR, Green WW, McIntyre JC, Ukhanov K, Zimmerman AD, Shively DT, Zhang L, Nishimura DY, Sheffield VC, and Martens JR. Gene Therapeutic Reversal of Peripheral Olfactory Impairment in Bardet Biedl Syndrome. Mol Ther 2017. 233. Williams DS. Usher syndrome: animal models, retinal function of Usher proteins, and prospects for gene therapy. Vision Res 48: 433 441, 2008. 234. Wolfrum U and Schmitt A. Rhodopsin transport in the membrane of the connecting cilium of mammalian photoreceptor cells. Cell Motil Cytosk eleton 46: 95 107, 2000. 235. Xu J, Dodd RL, Makino CL, Simon MI, Baylor DA, and Chen J. Prolonged photoresponses in transgenic mouse rods lacking arrestin. Nature 389: 505 509, 1997. 236. Yan D and Liu XZ. Genetics and pathological mechanisms of Usher syn drome. J Hum Genet 55: 327 335, 2010. 237. Yang J, Liu X, Zhao Y, Adamian M, Pawlyk B, Sun X, McMillan DR, Liberman MC, and Li T. Ablation of whirlin long isoform disrupts the USH2 protein complex and causes vision and hearing loss. PLoS Genet 6: e1000955, 2010. 238. Yaqoob Z, Wu J, and Yang C. Spectral domain optical coherence tomography: a better OCT imaging strategy. Biotechniques 39: S6 13, 2005. 239. Yoshimura H, Oshikawa C, Nakayama J, Moteki H, and Usami S. Identification of a novel CLRN1 gene mutation in Usher syndrome type 3: two case reports. Ann Otol Rhinol Laryngol 124 Suppl 1: 94S 99S, 2015. 240. Zallocchi M, Binley K, Lad Y, Ellis S, Widdowson P, Iqball S, Scripps V, Kelleher M, Loader J, Miskin J, Pe ng YW, Wang WM, Cheung L, Delimont D, Mitrophanous KA, and Cosgrove D. EIAV based retinal gene therapy in the shaker1 mouse model for usher syndrome type 1B: development of UshStat. PLoS One 9: e94272, 2014. 241. Zallocchi M, Meehan DT, Delimont D, Askew C Garige S, Gratton MA, Rothermund Franklin CA, and Cosgrove D. Localization and expression of clarin 1, the Clrn1 gene product, in auditory hair cells and photoreceptors. Hear Res 255: 109 120, 2009.
163 242. Zallocchi M, Meehan DT, Delimont D, Rutledge J, Gr atton MA, Flannery J, and Cosgrove D. Role for a novel Usher protein complex in hair cell synaptic maturation. PLoS One 7: e30573, 2012. 243. Zheng QY, Ding D, Yu H, Salvi RJ, and Johnson KR. A locus on distal chromosome 10 (ahl4) affecting age related hea ring loss in A/J mice. Neurobiol Aging 30: 1693 1705, 2009. 244. Zhong H, Eblimit A, Moayedi Y, Boye SL, Chiodo VA, Chen Y, Li Y, Nichols RM, Hauswirth WW, Chen R, and Mardon G. AAV8(Y733F) mediated gene therapy in a Spata7 knockout mouse model of Leber co ngenital amaurosis and retinitis pigmentosa. Gene Ther 22: 619 627, 2015. 245. Zolotukhin S, Potter M, Zolotukhin I, Sakai Y, Loiler S, Fraites TJ, Chiodo VA, Phillipsberg T, Muzyczka N, Hauswirth WW, Flotte TR, Byrne BJ, and Snyder RO. Production and puri fication of serotype 1, 2, and 5 recombinant adeno associated viral vectors. Methods 28: 158 167, 2002. 246. Zou J, Mathur PD, Zheng T, Wang Y, Almishaal A, Park AH, and Yang J. Individual USH2 proteins make distinct contributions to the ankle link complex during development of the mouse cochlear stereociliary bundle. Hum Mol Genet 24: 6944 6957, 2015.
164 BIOGRAPHICAL SKETCH Rachel Michelle Stupay was born and raised in Frankfort, Illinois and graduated from Lincoln Way East High School in the spring of 2004. Growing up Rachel trained as a professional ballet dancer with The School of Ballet Chicago for over a decade. She performed with the Joffre y Ballet of Ch icago for four consecutive years and was a rom 2004 2005 Rachel was a Post Graduate with Ballet Met in Columbus, OH. She was then offered a company contract with the Alabama ballet from 2006 2008. During this time Rachel was also a Sales Lead/ Key holder and Credit Card Coach for Ann Taylor Inc. In 2007, Rachel returned to undergraduate school at the University of Alabama at Birmingham In 2009 she was awarded a Genetics Summer R esearch Internship and a Summer Internshi p in Biomedical Sciences in 2010 where she worked in Dr. Bradley K. Rachel first became interested in studying Ciliopathies through her w here her research consisted of exe cuting a Caenorhabiditis elegans mutagenesis screen to identify novel ciliopathy disease causing alleles ( Stupay RM et al, 2009 https://www.uab.edu/inquiro/images/Archives/Volume_3.pd f pages 37 45 ; Stupay RM et al, 2010 https://www.uab.edu/inquiro/images/Archives/Volume_4.pdf pages 55 61 ) (231) Rachel graduated from UAB in the spring of 2011 with a major i n Molecular Biology and a minor in Biochemistry. Rachel Interdisciplinary Program in Biomedical Sciences and began g raduate school in the fall of 2011 In 2012 Rachel joined the lab of Dr. William W. Hauswirth and was appointed a Graduate Fellow on an NIH Op hthalmology Training Grant Her thesis project involve d studying an Usher Syndrome Type 3A (USH3A) mouse model in order to define a retinal phenotype and
165 establish a therap eutic approach for treatment (72) Rachel has extensive experience with working with mice on several ophthalmologic assays including: ERG (LKC and Espion systems), optical coherence tomography (OCT) (Heidelburg Spectralis and Bioptigen systems), fundoscopy, and optokinetics as well as exte nsive breeding and genotyping. Rachel is also experienced in Cryostat and paraffin microtome fixation and sectioning for histology, qPCR, PCR, western blotting, and AAV vector cloning. Rachel graduated with her doctoral degree in the fall of 2017 and subs equently moved to Cleveland, Ohio to work as a post doctoral fellow with Dr. Brian D. Perkins in the Cole Eye Institute at the Cleveland Clinic Her postdoctoral studies are focus ed on using zebrafish as a model organism to study photoreceptor degeneration in ciliopathy models that mimic patient phenotypes and how the cilia protein mutations disrupt normal cellular pathway s.