Persistent remodeling and neurodegeneration in late-stage retinal degeneration

Human retinitis pigmentosa (RP) typically involves decades of progressive vision loss before some patients become blind, and prospective therapies target patients who have been blind for substantial time, even decades. Evaluations of molecular and cellular therapies have primarily employed short-lived mouse models lacking the scope of remodeling common in human RP. The Rho Tg P347L transgenic rabbit offers a unique opportunity to evaluate the primary degeneration event and subsequent progressive remodeling that ensues over a timespan that recapitulates the human disease phenotype. Retinas from a TgP347L rabbit model of human dominant RP and wild-type litter mates were harvested over an 8-year span and processed for transmission electron microscope connectomics, immunocytochemistry for a range of macromolecules, and computational molecular phenotyping for small molecules, including transport tracing with D-Asp. Early time points in the TgP347L rabbit recapitulate the established sequence of photoreceptor loss, retinal remodeling, and reprogramming, and also reveal progressive disruptions in Müller cell metabolism, where rather than observing a homogeneous glial population, chaotic metabolic signatures emerge. By 4 years, virtually all remnants of photoreceptors are gone and the neural retina manifests severe cell loss and near complete loss of glutamine synthetase, though glial glutamate transport persists. By 6 years, there is a global >90% neuronal loss. In some regions the retina is devoid of identifiable cells and replaced by unknown debris-like assemblies. Though the 6-year retina does have locations with recognizable neurons, all cell types are drastically reduced in number and some have altered metabolic phenotypes. These results are never seen in wt littermates, including rabbits which are 8 years old. Electron microscopic analysis using wide-field connectomics imaging of the 6-year TgP347L sample demonstrates some structurally normal synapses, indicating that survivor neurons in these regions are not quiescent despite the lack of sensory input for a substantial period of time. These results indicate that, although photoreceptor degeneration is the trigger, retinal remodeling ultimately gives way to neurodegeneration, which is a separate unrelenting disease process independent of the initial insult, closely resembling slow progressive CNS neurodegenerations. Indeed, both metabolic disruption and debris-related degeneration predicts the existence of a persistent neuropathy, and increases in ?-synuclein levels support a proteinopathy component. Remodeling and neurodegeneration progress until the retina is devoid of recognizable cells. There is no stable state into which the retina settles and no cell type is spared. This has profound implications for current therapeutics. There will likely be critical windows for implementation but, ultimately, suspension of neurodegenerative remodeling will be required for long-term success.

PERSISTENT REMODELING AND NEURODEGENERATION IN LATE-STAGE RETINAL DEGENERATION by Rebecca Lynne Pfeiffer A dissertation submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Neuroscience Interdepartmental Program in Neuroscience The University of Utah December 2017 Copyright © Rebecca Lynne Pfeiffer 2017 All Rights Reserved The University of Utah Graduate School STATEMENT OF DISSERTATION APPROVAL The dissertation of Rebecca Lynne Pfeiffer has been approved by the following supervisory committee members: Robert E. Marc , Chair April 13th, 2017 Date Approved Alessandra Angelucci , Member April 13th, 2017 Date Approved , Member David Krizaj April 13th, 2017 Date Approved Karen S. Wilcox , Member April 13th, 2017 Date Approved Megan E. Williams , Member April 13th, 2017 Date Approved and by Richard Dorsky , Chair/Dean of the Department/College/School of Interdepartmental Program in Neuroscience and by David B. Kieda, Dean of The Graduate School. ABSTRACT Human retinitis pigmentosa (RP) typically involves decades of progressive vision loss before some patients become blind, and prospective therapies target patients who have been blind for substantial time, even decades. Evaluations of molecular and cellular therapies have primarily employed short-lived mouse models lacking the scope of remodeling common in human RP. The Rho Tg P347L transgenic rabbit offers a unique opportunity to evaluate the primary degeneration event and subsequent progressive remodeling that ensues over a timespan that recapitulates the human disease phenotype. Retinas from a TgP347L rabbit model of human dominant RP and wild-type litter mates were harvested over an 8-year span and processed for transmission electron microscope connectomics, immunocytochemistry for a range of macromolecules, and computational molecular phenotyping for small molecules, including transport tracing with D-Asp. Early time points in the TgP347L rabbit recapitulate the established sequence of photoreceptor loss, retinal remodeling, and reprogramming, and also reveal progressive disruptions in Müller cell metabolism, where rather than observing a homogeneous glial population, chaotic metabolic signatures emerge. By 4 years, virtually all remnants of photoreceptors are gone and the neural retina manifests severe cell loss and near complete loss of glutamine synthetase, though glial glutamate transport persists. By 6 years, there is a global >90% neuronal loss. In some regions the retina is devoid of identifiable cells and replaced by unknown debris-like assemblies. Though the 6-year retina does have locations with recognizable neurons, all cell types are drastically reduced in number and some have altered metabolic phenotypes. These results are never seen in wt littermates, including rabbits which are 8 years old. Electron microscopic analysis using wide-field connectomics imaging of the 6-year TgP347L sample demonstrates some structurally normal synapses, indicating that survivor neurons in these regions are not quiescent despite the lack of sensory input for a substantial period of time. These results indicate that, although photoreceptor degeneration is the trigger, retinal remodeling ultimately gives way to neurodegeneration, which is a separate unrelenting disease process independent of the initial insult, closely resembling slow progressive CNS neurodegenerations. Indeed, both metabolic disruption and debris-related degeneration predicts the existence of a persistent neuropathy, and increases in α-synuclein levels support a proteinopathy component. Remodeling and neurodegeneration progress until the retina is devoid of recognizable cells. There is no stable state into which the retina settles and no cell type is spared. This has profound implications for current therapeutics. There will likely be critical windows for implementation but, ultimately, suspension of neurodegenerative remodeling will be required for long-term success. iv TABLE OF CONTENTS ABSTRACT ....................................................................................................................... iii LIST OF FIGURES ......................................................................................................... viii Chapters 1. INTRODUCTION .......................................................................................................... 1 1.1 Architecture of the Retina .................................................................................... 2 1.1.1 Photoreceptors....................................................................................... 3 1.1.2 Horizontal Cells .................................................................................... 4 1.1.3 Bipolar Cells ......................................................................................... 5 1.1.4 Amacrine Cells...................................................................................... 6 1.1.5 Ganglion Cells ...................................................................................... 7 1.1.6 Glia ........................................................................................................ 8 1.2 Retinal Degenerative Diseases............................................................................. 9 1.2.1 Retinitis Pigmentosa ............................................................................. 9 1.2.2 Age-Related Macular Degeneration ................................................... 10 1.3 Phases of Retinal Remodeling ........................................................................... 11 1.3.1 Phase 0: The Healthy Retina .................................................................. 12 1.3.2 Phase 1: Rod Degeneration .................................................................... 13 1.3.3 Phase 2: Cone Degeneration .................................................................. 14 1.3.4 Early Phase 3: Neurite Remodeling ....................................................... 14 1.3.5 Mid Phase 3: Global Remodeling .......................................................... 15 1.3.6 Late Phase 3: Plateau Remodeling ......................................................... 16 1.4 Therapeutic Approaches to Vision Restoration ................................................. 16 1.4.1 Optogenetics .......................................................................................... 17 1.4.2 Chemical Photoswitches ........................................................................ 18 1.4.3 Cellular Replacement Strategies ............................................................ 19 1.4.4 Bionic Implants ...................................................................................... 20 1.4.5 Genetic Approaches ............................................................................... 21 1.4.6 Requirements for Therapeutic Interventions to Be Effective ................ 22 1.5 References .......................................................................................................... 27 2. MATERIALS AND METHODS .................................................................................. 35 2.1 Computational Molecular Phenotyping (CMP) ................................................. 35 2.1.1 Molecular Trapping and IgG Labeling .................................................. 36 2.1.2 CMP Analysis: Pattern Recognition ...................................................... 40 2.2 P347L Transgenic Rabbit Model of Retinitis Pigmentosa ................................ 43 2.3 References .......................................................................................................... 48 3. MÜLLER CELL METABOLIC CHAOS DURING RETINAL DEGENERATION 52 3.1 Introduction ........................................................................................................ 53 3.1.1 Retinal Degeneration and Remodeling .................................................. 53 3.1.2 Müller Glia in Degenerating Retina ....................................................... 54 3.1.3 Müller Glia Functions and Homogeneity .............................................. 54 3.1.4 Transgenic P347L Rabbit Model of Autosomal-Dominant Retinitis Pigmentosa (adRP).......................................................................................... 54 3.1.5 Computational Molecular Phenotyping ................................................. 54 3.2 Methods ............................................................................................................. 54 3.2.1 Model Systems ....................................................................................... 54 3.2.2 Tissue Processing ................................................................................... 55 3.2.3 CMP and Nomenclature ......................................................................... 55 3.2.4 Analysis.................................................................................................. 55 3.3 Results................................................................................................................ 55 3.3.1 CMP Signatures ..................................................................................... 56 3.3.2 Clustering, Histograms, and Analysis .................................................... 56 3.4 Discussion .......................................................................................................... 58 3.5 Acknowledgements ............................................................................................ 60 3.6 References .......................................................................................................... 60 4. EVALUATION OF EXCITOTOXICITY AS A MECHANISM OF RETINAL DEGENERATION ........................................................................................................... 62 4.1 Introduction ........................................................................................................ 62 4.1.1 Excitotoxicity Hypothesis ...................................................................... 63 4.1.2 Evidence Disputing Excitotoxicity ........................................................ 65 4.2 Materials and Methods....................................................................................... 67 4.2.1 Model Systems ....................................................................................... 67 4.2.2 Tissue Processing ................................................................................... 68 4.2.3 CMP Analysis ........................................................................................ 68 4.3 Results................................................................................................................ 69 4.4 Conclusions and Discussion .............................................................................. 72 4.5 References .......................................................................................................... 79 5. LONG-TERM RETINAL DEGENERATION LEADS TO EXTENSIVE NEURODEGENERATION AND ALPHA SYNUCLEIN AGGREGATION ................ 82 5.1 Effects of Long-term Remodeling on Therapeutic Interventions ...................... 82 5.2 Animal Models of Retinal Remodeling and Possible Caveats .......................... 83 5.3 Materials and Methods....................................................................................... 84 5.3.1 Model Systems ....................................................................................... 84 vi 5.3.2 Tissue Processing ................................................................................... 84 5.3.3 Computational Molecular Phenotyping (CMP) ..................................... 85 5.3.4 Protein Variation Quantification ............................................................ 86 5.4 Characterization of Aged P347L Rabbit ............................................................ 87 5.5 Ubiquitin Expression in Retinal Degeneration .................................................. 88 5.6 Alpha Synuclein Aggregation in Older Transgenic Retina ............................... 89 5.7 Discussion .......................................................................................................... 91 5.8 References ........................................................................................................ 104 6. DISCUSSION AND FUTURE DIRECTIONS ......................................................... 107 6.1 Müller Cell Metabolic Chaos ........................................................................... 107 6.2 Excitotoxicity in the Retina ............................................................................. 110 6.3 Retinal Neurodegeneration .............................................................................. 112 6.4 Implications for Therapeutic Interventions...................................................... 113 6.5 Future Directions ............................................................................................. 114 6.6 References ........................................................................................................ 121 vii LIST OF FIGURES Figures 1.1 Schematic of the major histological and cellular classifications of the retina ............ 25 1.2 Remodeling of the mammalian retina in retinal degenerations .................................. 26 2.1 K-means generation of theme maps ............................................................................ 46 3.1 Comparison of WT amino acid levels to that of Tg P347L ........................................ 56 3.2 Representative τQE and theme maps of WT retina vs degenerate retina ................... 57 3.3 Univariate amino acid probability density distributions (histograms) for each of the amino acids displayed in Fig 3.2....................................................................................... 58 3.4 τQE  rgb mappings of registered amino acid signals in serial 100 nm sections ..... 59 3.5 Cumulative distribution frequencies (CDFs) of the entire Müller cell cohorts in the regions displayed in 3.4C-F (red Normal Distribution (ND) line) in addition to 10 4x4 µm regions from WT and degenerate retina (MC1-10 CDFs) as gray lines. .......................... 60 4.1 Müller cell metabolic variation across retinal regions and species............................. 75 4.2 Müller cell metabolic variation within Müller cells ................................................... 76 4.3 GS/Q/E variability in WT compared to degenerate retina .......................................... 77 4.4 D-Aspartate loading in retinal chips ........................................................................... 78 5.1 Representative registered images of amino acid immunoreactivity in 100nm sections of WT, Phase 2, and Phase 3 remodeling in rabbit retina ................................................. 94 5.2 Representative images from WT rabbit (A-C), 2yr (d-F), 4yr (G-I), 5(yr (J-L), and 6yr (M-O) Tg P347L rabbit retina........................................................................................... 95 5.3 Theme map overlay of TEM image of retina.............................................................. 98 5.4 Grayscale ubiquitin labeling in rabbit retinas ............................................................. 99 5.5 Endogenous signaling of phosphorylated α-synuclein (α-syn), and GFAP .............. 100 5.6 Western Blot: Beta actin demonstrated some increase in Tg retinas when compared to WT. ................................................................................................................................. 102 5.7 Antibody inhibition experiment ................................................................................ 103 6.1 Retinal degeneration, a combination of remodeling and neurodegenerative processes ......................................................................................................................... 117 6.2 Workflow for generation of pathoconnectome ......................................................... 119 6.3 Preliminary RPC1 data ............................................................................................. 120 ix CHAPTER 1 INTRODUCTION Diseases such as diabetic retinopathy, age-related macular degeneration, glaucoma, and retinitis pigmentosa, cause irreversible vision loss affecting millions of people worldwide.1 Of these diseases, degenerations that onset with the loss of photoreceptors (age-related macular degeneration and retinitis pigmentosa) have received special attention with the goal of restoring vision through therapeutic interventions, primarily after photoreceptors have degenerated.2-13 Ultimately this dissertation aims to describe the long term effects of retinal degeneration, factors contributing to degeneration following the loss of photoreceptors, identify potential complications in implementing therapeutics longterm, and propose windows of degeneration in which therapeutic interventions have the greatest likelihood of success. This chapter aims to introduce the reader to the structure of the healthy retina and its general roles in vision. In addition, this chapter will give a broad overview of retinal degenerative diseases, phases of degeneration, therapeutic interventions currently in development, and describe probable requirements for long-term vision restoration, which will be recurrent themes throughout this dissertation. 2 1.1 Architecture of the Retina The retina is a thin multilaminar strip of nervous tissue along the posterior surface of the eye, which developmentally originates from the frontal cortex, making it part of the central nervous system. The retina is subdivided into six major histological layers: the photoreceptor layer (which may be further subdivided into the outersegment, innersegment, and outer nuclear layer), the outer plexiform layer (OPL), the inner nuclear layer (INL), the inner plexiform layer (IPL), the ganglion cell layer (GCL), and the optic fiber layer (OFL) (Figure 1.1). It is composed of five major neuronal cell types: photoreceptors, horizontal cells, bipolar cells, amacrine cells, and ganglion cells.14 These cells working together play important roles in processing color, motion, intensity, and directionality from visual stimulus.15 In addition to neurons, the retina also has numerous glial cells: Müller cells, astrocytes, and microglia of immune system origin, providing critical functions in the retina. The retinal circuitry is primarily divided into two major pathways: vertical and horizontal. The vertical pathway is the path from the photoreceptors, bipolar cells, and ganglion cells, which ultimately project axons into the brain. It is primarily thought of as an amplification pathway, though the presence of ON and OFF cells permits the first level of light-detection processing, and ON-OFF motifs are preserved from bipolar cells through processing in the brain. The horizontal pathway tunes signals from the vertical pathway from being extremely broad regions of light detection to more defined regions of higher light intensity with less light intense regions surrounding. The higher intensity regions of focus are termed centers while the surrounding regions are aptly named the surround. Together, the vertical and horizontal pathways provide initial processing of visual information, prior to action potentials relaying signals to the brain 3 1.1.1 Photoreceptors In most species, photoreceptors are the neurons of the retina found closest to the back of the eye. They are the first cells to receive visual stimulus via light, and consist of two major types: cones and rods. Rods and cones are responsible for translating light information (photons) into neurochemical signals, which is how most information is passed throughout the nervous system. In photoreceptors, the neurotransmitter released is glutamate. Mammals tend to have two to three types of cones and a single type of rod. The opsins (light-sensitive proteins) for both classes of photoreceptors are contained in the outersegment portion of the photoreceptor. Outer segments are composed of bilipid membrane stacked disks, which are continuously shed distally, and replaced proximal to the cell body. Rod outer segments are stacked in thin rod shaped columns, while cone outer segments are conical. Cones provide responses to bright light and have different opsins (light-sensing molecules) tuned to different wavelengths. The mixing of information from cones containing different opsins is the mechanism by which we see color. In contrast to cones, there is only one type of rod in mammals. Rods contain the opsin rhodopsin, and the rod transduction pathway makes it extremely sensitive to light, facilitating vision at low levels of light (e.g., starlight). A fundamental principle associated with all neurons is that when they are depolarized, the rate of synaptic vesicle fusion leading to neurotransmitter release is increased. When neurons are hyperpolarized the opposite is true. In response to light, photoreceptors hyperpolarize, meaning the amount of glutamate released decreases as the intensity of light increases. In darkness, photoreceptors are depolarized and constitutively release glutamate, leading to a large concentration of glutamate in the synaptic cleft. Though this may seem counterintuitive, the balance of glutamate being 4 released versus being removed from the synapse via transporters creates a concentration gradient making the system very sensitive to even small variations in light. The hyperpolarization and depolarization responses to photon capture are not an all-or-none response (e.g., an action potential), but rather a graded potential, allowing small and large changes in available photons to be rapidly relayed to neurons of the inner retina. The postsynaptic partners of photoreceptors (horizontal cells and bipolar cells) have evolved mechanisms to detect and respond to slight changes in glutamate release by photoreceptors. 1.1.2 Horizontal Cells Horizontal cells are found immediately below the outer plexiform layer and, as the name would suggest, are oriented horizontal to the vertical pathway of the retina. The primary function of horizontal cells in most mammals is feedback onto cone photoreceptors helping to generate center-surround response curves and assist in tuning luminosity, though in nonmammalian vertebrates, they also help to tune chromatic information. Horizontal cells are extensively coupled to one another via large gap junctions, facilitating wide spreading of depolarization and hyperpolarization events throughout the retina. A common theme in neuroscience is sign-conserving versus sign-inverting synapses. Sign-conserving synapses are cases where the postsynaptic neuron has the same polarity response as the presynaptic neuron, though not necessarily the same kinetics. An example of this is the synapse between cones and horizontal cells. Horizontal cells are predominately activated via ionotropic glutamate receptors, which depolarize in response to glutamate released by cones and hyperpolarize when glutamate release is decreased, thereby matching the polarization state of the cones that directly drive them. In contrast, the "synapses" back 5 onto cones from horizontal cells are sign-inverting: as horizontal cells hyperpolarize, they trigger depolarization of the photoreceptor, albeit using nonconventional signaling pathways though to involve ephaptic mechanisms or alkalization of the perisynaptic space. These sign-inverting and sign-conserving feedback loops are a common motif that, combined with graded potentials, are fundamental to initial visual processing performed by the retina. The number and types of horizontal cells is variable between animals, but in mammals it is generally accepted that there are two types: type A (axon-less) and type B (axon-bearing). Despite the two morphological types, horizontal cells are often referred to as having three functional types; this is because the rod-associated axon terminals and cone-associated somas of type B cells function independently of one another. It is now thought that horizontal cell axon terminals may provide some feedback function for rods as well. The three functional types of horizontal cell's feedback with photoreceptors create the first level of center-surround tuning. 1.1.3 Bipolar Cells Bipolar cells are specialized interneurons, which pass signals from photoreceptors onto the rest of the retina. In mammals bipolar cells are subdivided into three major classes: ON-cone, OFF-cone, and rod. As described above, photoreceptors are persistently depolarized in darkness, and hyperpolarize in response to light. Bipolar cells respond to increases or decreases in light by either sign-conserving or sign-inverting responses via ionotropic glutamate receptors (iGluRs) or group III metabotropic glutamate receptors (mGluRs), respectively, on dendrites that synapse with photoreceptors within the OPL. Metabotropic glutamate receptors transduce sign-inverting signaling, which close cation 6 permeable channels in the plasma membrane in response to glutamate, decreasing inward currents and causing hyperpolarization. Alternatively, ionotropic glutamate receptors gate opening of cation channels in response to glutamate, increasing inward currants and initiate depolarization. Thus iGluRs mimick the response of the photoreceptors, leading to signconserving signaling. ON-cone and rod bipolar cells express mGluR6 metabotropic glutamate receptors, causing them to depolarize in the presence of light. OFF-cone bipolar cells have ionotropic glutamate receptors, causing them to be depolarized in the absence of light. Bipolar cells, like photoreceptors, are glutamatergic in their neurotransmission to synaptic partners, with glutamate packaged into synaptic vesicles and associated with ribbon synapses in bipolar cell axons. Synaptic ribbons perform important tasks in amplifying signals from photoreceptors and pass the information to amacrine and ganglion cells either via AMPA receptors or a mixture of AMPA and NMDA receptors. Cone bipolar cells synapse directly onto ganglion cell dendrites, while rod bipolar cell axons do not.16 Rod signals reach ganglion cells indirectly via the AII amacrine cell network. The layers of the retina where bipolar cells make their synaptic connections is mostly segregated: ONbipolar cells synapse in the proximal IPL (closer to the GCL), and OFF-bipolar cells synapse in the distal IPL (closer to the INL)17; although this is a generalization and some synaptic crossover does occur.18 1.1.4 Amacrine Cells Amacrine cells play multiple roles in visual processing in the retina. They are the most diverse cell type in the retina, with the mammalian retina possessing at least 30 classes of amacrine cells. The main neurotransmitters associated with amacrine cells are GABA 7 and glycine.19-21 These neurotransmitters are inhibitory and they act on synapses in a signinverting manner. In addition to chemical synapses, amacrine cells also couple with one another via a large network of gap junctions, permitting the transfer of graded potentials electrically. GABAergic amacrine cells are predominantly thought of as gain modulators of the retina helping to organize vision beyond ON versus OFF with a complex network of feedback synapses onto bipolar cells and feed forward synapses with ganglion cells. Beyond the synaptic contacts amacrine cells make onto bipolar and ganglion cells, one type of amacrine cell, the AII amacrine cell, plays a unique role in the vertical pathway as a sign-conserving interneuron. The AII amacrine cell receives sign-conserving synaptic input from the rod-bipolar cell axons causing it to depolarize. This depolarization then spreads to nearby ON-bipolar cells via gap junctions, leading to the depolarization and release of glutamate by these cells onto ganglion cells. Simultaneously, the depolarization of AII amacrine cells causes the release of the neurotransmitter glycine onto OFF-bipolar cells, thereby inhibiting them.16, 22 These two paths inject rod-ON and rod-OFF signals into ON and OFF ganglion cells, respectively. 1.1.5 Ganglion Cells Ganglion cells are the only neuron of the retina to directly synapse within the brain. In mammals, ganglion cells project to the lateral geniculate nucleus, superior colliculus, suprachiasmatic nucleus, and accessory optic nuclei in the midbrain.23 The dendrites of ganglion cells arborize in the IPL and are postsynaptic to both bipolar cells and amacrine cells. Depending on the combination of bipolar and amacrine cell synapses onto ganglion cells and the spatial properties of the amacrine cells, different features of the visual world 8 are encoded into a sequence of action potentials to discreet nuclei in the brain. In addition to ganglion cells firing depending on whether they receive ON or OFF input, thereby preserving the initial encoding by bipolar cells, some ganglions cells are specially tuned to respond only to signals moving in certain directions, or are temporally responsive, giving transient or sustained responses to the same visual stimulus. Other ganglion cells are also intrinsically photosensitive due to their expression of melanopsin; these ganglion cells play important roles in the regulation of pupil size and maintenance of circadian rhythm. 1.1.6 Glia Glia are traditionally referred to as the support cells of the nervous system. In the retina, the predominant glial cell type is the Müller cell. Müller glia have been shown to play important structural roles within the neural retina24 including sealing the retina off from the vitreous and separating the ONL from the INL. They play a key role in retinal homeostasis through removal of neural waste products such as CO2 and NH3, redistribution of K+, and are important in the glutamate-glutamine cycle by metabolizing glutamate to glutamine via the enzyme glutamine synthetase.25-27 The other macroglia found in the retina is astrocytes; they are predominantly restricted to the OFL, though they make some vascular contacts in the GCL, contributing to the blood-retinal-barrier (BRB). One distinction between brain astrocytes and retinal astrocytes is the absence of glutamine synthetase in astrocytes of the retina. In addition to macroglia, microglia may be found distributed throughout the retina, though they are typically found in the ONL, OPL, IPL, and GCL in healthy retina. In conditions of stress microglia may be found in different activation states throughout the retina. 9 In summary, the retina is a complex neural system, consisting of over 70 neural types (in mammals), which form chemical and electrical connections responsible for initial visual processing. The precise patterning of these connections, in addition to the metabolic system requirements to allow for constant signaling, work together to provide the brain with the information it needs to interpret the visual world. 1.2 Retinal Degenerative Diseases Retinal degenerative diseases including retinitis pigmentosa (RP) and age-related macular degeneration (AMD) cause irreversible vision impairment, by directly or indirectly leading to photoreceptor loss. Though the initial progression rate of photoreceptor loss may vary between diseases, the ultimate outcome is the same: global loss of photoreceptors across the retina, leaving a deafferented neural retina. 1.2.1 Retinitis Pigmentosa Retinitis pigmentosa (RP) refers to a range of hereditary disorders, which onset with degeneration of the photoreceptors. The name retinitis pigmentosa originates from the ophthalmoscopic observation of exposed pigmentation in the neural retina, arising from migration of pigmented RPE cells into the neural retina.28 The prevalence of RP is approximately 1 in 4000, with varying modes of inheritance: 50-60% of cases are inherited autosomal-recessive, 30-40% autosomal-dominant, and 5-15% X-linked. In fact, RP is known to arise from almost 300 gene defects,* which are subclassified by their genetic * https://sph.uth.edu/Retnet/ (Accessed March 5th,2017) 10 component, mode of inheritance, and whether the defect is constrained to the eye or also affects nonocular tissues. Examples of associated nonocular diseases are Bardet-Biedl syndrome and Usher syndrome, which are both autosomal-recessive. Bardet-Biedl syndrome, in addition to photoreceptor loss, is associated with polydactyly, truncal obesity, hypogonadism, renal dysfunction, and learning difficulties.29 Usher syndrome is associated with hearing-loss in addition to RP.30 Despite the many modes of inheritance and genes involved with RP, outer segments of rod degenerate in a patchy fashion across the retina, followed by cones, and gradually lead to loss of vision in patients. The speed of vision loss and the mechanism causing the photoreceptor degeneration varies widely across RP diseases, along with the prevalence of that specific genetic defect. Some patients will present with vision loss early in life, while others will remain asymptomatic until mid-adulthood. In most cases, RP patients are legally blind by their mid-forties, and photoreceptor degeneration continues until the retina is completely devoid of photoreceptors, leaving patients completely unable to perceive light for years to decades of life.31 1.2.2 Age-Related Macular Degeneration As the name would imply, Age-Related Macular Degeneration (AMD) predominately affects older people, and is the leading cause of vision loss in individuals over the age of 50.† Unlike RP, AMD is only partially associated with genetic inheritance. Rather, AMD is also associated with environmental factors such as smoking, which can † https://nei.nih.gov/health/maculardegen/armd_facts 11 greatly increase an individual's risk.32 Clinically, patients often complain about blurriness in their central vision, which progresses to a loss of central vision which then advances outward towards the periphery.33 Histologically, dry AMD presents with drusen (lipid deposits under the retina) separating the RPE from the choroid, metabolic alterations of RPE cells,34 and subsequent stress and loss of photoreceptors. Wet AMD arises from new blood vessel formation (neovascularization) in the retina and subsequent leakage of blood and serum into the neural retina. 1.3 Phases of Retinal Remodeling Early studies evaluating the progression of advanced RP in humans noted clear gliosis, cellular disorganization, and displaced RPE cells scattered around the neural retina.16 Other early ophthalmologists have also noted RP associated atrophy in humans, with an emphasis on an apparent loss of ganglion cells.35-36 These early observations however have not been recapitulated in animal models, and therefore have largely not been factored into strategies for therapeutic interventions.37 This was in part due to limitations of conventional histology and the result of short-lived experimental animal models. Recent post mortem studies of humans and the further extensive characterization of animal models38 have shown that neurons and glia of the retina undergo metabolic and morphologic remodeling as a key component of retinal degeneration, antecedent and subsequent to photoreceptor loss in both humans and animal models of RP.33, 37-40 Remodeling includes neuronal rewiring, neurite growth, and reprogramming of neurons' characteristic gene expression. Some remodeling begins as soon as photoreceptor stress is initiated, while other remodeling processes emerge following the complete regional loss of 12 photoreceptors. The phases of remodeling follow the loss of photoreceptors, so different regions of the retina are often in different phases of remodeling because photoreceptor loss appears in a patchy fashion in most instances. The description presented below assumes a rod-initiated RP-like process, but the long term progression, particularly in phase 3 (described below), is similar in all forms.28 Ultimately, remodeling leads to large-scale corruption of neural networks in the retina, both through neural rewiring and cell death, making it unclear whether a retina which has undergone these alterations is capable of supporting therapeutic interventions to vision loss.41 In the interest of clarity, our lab has classified the extent of remodeling into 3 distinct phases (Figure 1.2).28, 40 1.3.1 Phase 0: The Healthy Retina The healthy retina is described more completely in section 1.1 of this chapter. In brief, the retina is composed of two major components that will be referenced throughout this section: the sensory retina and the neural retina. The sensory retina is composed of the photoreceptors (both their outer segments and nuclei), which are responsible for the primary detection of light. The neural retina then performs extensive processing of the light information before transmitting the information to the brain via ganglion cell action potential. The neural retina is composed of five main cell types: horizontal cells (HCs), bipolar cells (BCs), amacrine cells (ACs), ganglion cells (GCs), and glia (predominantly Müller cells (Müller cells), but also including microglia and astrocytes). 13 1.3.2 Phase 1: Rod Degeneration Rod degenerations are triggered through a number of gene defects including those listed in section 1.2 of this chapter, or through some sort of trauma (retinal detachment or light-induced retinal degeneration). Regardless of the initiating insult, rod outer segments become shortened, followed by rod stress, deconstruction, and ultimately cell death. Prior to this, some rods will extend their axons and terminate in the GCL, rather than the conventional termination in the OPL.42 During the initial stress of the outer segments, rod bipolar cells begin to retract their dendrites from their previously associated rod terminal (spherule).43 Some of these rod bipolar cells will initiate rewiring and extend their dendrites toward cone terminals (pedicles).44 In addition, most or all rod bipolar cells change their expressed receptors from metabotropic to ionotropic, causing them to functionally change from being ON-type cells to OFF-type. Beyond the neural responses seen from the rod bipolar cells, glia also respond to the rod degeneration. Microglia begin to invade the subretinal space, presumably to aid in the removal of debris of rod degeneration. In the neural retina, Müller cells also respond to the stress signals by beginning to become hypertrophic and increasing expression of glial fibrillary acidic protein (GFAP), which is typically maintained at very low to undetectable levels in Müller cells of the healthy retina. In addition to GFAP increase, Müller cells also show a variable increase in glutathione levels, which is thought to be an indicator of oxidative stress.45 During the later period of rod degeneration, cones also begin to truncate their outer segments, even if the precipitating event was rod-specific. The mechanisms leading to the cone degeneration are currently unclear, though several models of bystander killing have been proposed.28, 33, 37-38, 46 14 1.3.3 Phase 2: Cone Degeneration As phase 1 remodeling transitions into phase 2, rods are continually degenerating, leading to a collapse of the outer nuclear layer and concurrently cone stress and degeneration. The main feature distinguishing phase 2 remodeling is the widespread death of photoreceptors.28 This process is likely the result of a number of factors, including microglial invasion of the subretinal space. RPE dysfunction caused by the rapid disassembly of rods, leading to debris accumulation in the subretinal space, is another theory for why cones die. The end result is the same; cones will degenerate and die, even in cases where the precipitating gene mutation is rod specific. This process may be fast or slow; there are forms of RP in which rod death causes cone outer segment loss, but the cone nuclei remain for extended periods of time, these are termed cone-sparing RP,47 and delay most late-stage remodeling as long as cones are present. When cones deconstruct, BCs retract their dendrites.43 HCs become hypertrophic and some extend multiple axonlike processes into the neural retina, in some cases, reaching the GCL.48 Simultaneously, Müller cells elaborate a seal of their distal microvilli across the neural retina, gluing it together with intermediate junctions and fully separating the remnant cones from the RPE.37-38 Though there is some cell death in the neural retina during this period, it is a small and not seen in regular histological analysis. 1.3.4 Early Phase 3: Neurite Remodeling The primary feature signaling the initiation of phase 3 remodeling is the loss of all photoreceptors. The preservation of even a small portion of cone cell bodies in a region of retina is sufficient to prevent much of the gross remodeling in that local region.38 As the 15 last of the cone cell bodies disappear, the glial seal separating the neural retina from the RPE completes, causing a total separation of the neural retina from the RPE and choroidal vasculature. In addition to the seal, Müller cells also hypertrophy along their trunks leading to the formation of large columns that fractionate the IPL,49 though the impacts of these elaborated columns on IPL connectivity is unknown, they appear to be conduits for neuronal migration.38 Cell death begins to become detectable within the neural retina by analysis of cell loss in phase 3, although the rate is so low that it cannot be tracked by techniques such as TUNEL labeling. At this point, anomalous neurites are produced by all classes of the remaining neurons. 1.3.5 Mid Phase 3: Global Remodeling Following the extension of neurites by amacrine, bipolar, and ganglion cells into new neurite fascicles, many new neuron processes form de novo clusters of neurites that form new synapses between one another.38 These collections of synapses between ectopic neurites are termed microneuromas and contain pre- and postsynaptic structures from all cell types, though whether they maintain the selectivity seen in healthy retina is unknown. Some neurons translocate their cell bodies to ectopic locations throughout the retina (e.g., ganglion cells migrate to the distal retina and bipolar and amacrine cells migrate into the ganglion cell layer), though most of these translocations occur along the elaborated Müller columns that appear earlier in remodeling.38 Not all neurons survive this global remodeling, and there is a noticeable cell loss throughout the retina, with no apparent preference among any neural type. Throughout all of these changes, surviving neurons appear to maintain their core molecular signatures, regardless of positional location within the retina.38 16 1.3.6 Late Phase 3: Plateau Remodeling As phase 3 continues, cell death leads to a depleted neural retina, with all layers being reduced in cell number. Microneuromas persist, though they may decrease in volume due to loss of neurons.40 Müller cells break through the inner limiting membrane and form an epi-retinal membrane between the vitreous and the neural retina. The RPE begins to break apart and RPE cells can be found dispersed throughout the neural retina.38 In summary, regardless of the primary insult, the neural retina undergoes a series of increasingly deleterious changes in response to loss of photoreceptors. These processes will likely have an impact on any therapeutic scheme introduced to restore/preserve vision once remodeling is initiated. Further understanding of these components will be essential to recommending the best potential therapeutic for different stages of remodeling.41 1.4 Therapeutic Approaches to Vision Restoration A primary interest for current vision research is the advancement of new therapies for the treatment of photoreceptor-associated vision loss. Therapies including small electrode array implants to stimulate remaining neurons,2-4, 50 optogenetic approaches involving the transfection of specialized opsins into remnant neurons,6-7 and a number of genetic and cellular based approaches are currently being investigated.13, 51 Although these therapies may hold promise for future treatment of retinal degenerative diseases, it is difficult to fully assess their probability of success or improve upon the initial concept without better understanding the remodeling associated with retinal degenerative diseases. 17 1.4.1 Optogenetics The fundamental concept behind optogenetics as a treatment for retinal degenerative disorders is that during and following the loss of outer segments of photoreceptors, if another cell type could be genetically encoded to detect light, vision could be rescued. Early studies demonstrated that when the algal channelrhodopsin-2 (ChR2) is inserted into the membrane of mammalian cells, a depolarization event can be initiated via light52 and Archean halorhodopsin (NpHR) can induce hyperpolarization.53-55 From this, the hypothesis was proposed that by genetically inserting these apoproteins into neurons, any neuron could be made light sensitive.56 Ultimately, the idea that neurons of the retina could be given new light detecting capabilities following the loss of photoreceptors57 was born. From here the discussion quickly turned to determining which cell type of the retina is optimal for targeting to induce, or restore, light-sensitivity. One early avenue was to restore light-sensitivity to cone somas, following the loss of outer segments with the goal of preserving native processing in the retina.5, 7 Though this has the caveat that in many forms of retinal degeneration, cone cell death occurs relatively early in the progression of the disease in humans, this may yet be a potential treatment in cases of cone-sparing RP. Subsequently, the neurons of the inner retina were targeted to compensate for the loss of photoreceptors, either through a broadly delivered intravitreal injection approach, which predominantly targets ganglion cells58-61 or through targeted delivery to specific cells of the inner retina.8, 12 The discussion is still ongoing which cell is the best for inducing light sensitivity, in what combination, and which are the best promoters for cell-specific targeting. A major problem with all the algal and protozoan opsins is their very low photosensitivity, requiring extremely bright light displays to generate responses.41 18 Some groups have demonstrated that native opsins, such as rhodopsin, transduced into mouse bipolar cells offer improved performance.62 Others have successfully transduced melanopsin into retinals, but the time course of melanopsin signaling is, so far, much too slow to provide functional vision.63 And of course virtually all of these studies are performed in retinas of mice in the very early stages of retinal degeneration, usually midphase 2. 1.4.2 Chemical Photoswitches A method for temporarily inducing photosensitivity in neurons involves the application of chemical photoswitches. These photosensitive small molecules regulate neuronal electrical activity by transiently blocking ion channels. An early study used the photochemical AAQ (acrylamide-azobenzene-quaternary ammonium), which in its trans form blocks K+ channels, thereby increasing excitability, and in its cis form unblocks K+ channels. AAQ photoisomerizes from its trans to cis conformation in the presence of short wavelength (~380nm) light, which unblocks K+ channels and decreases excitability. Relaxation from cis to trans occurs slowly in darkness, or reisomerization can be rapidly induced using long wavelength (~540nm) light,64 allowing bi-directional optical control. Polosukhina et al. unexpectedly found that 380nm light elicited action potentials, though K+ channels become unblocked at that wavelength. This turned out to be the result of AAQ primarily effecting amacrine cells, causing ganglion cells to fire due to disinhibition. A distinct problem with AAQ is the necessity of high intensity UV light for isomerization. Another chemical photoswitch that has been investigated is DENAQ (diethylaminoazobenzene-quaternary ammonium).65 DENAQ is also a K+ channel photoswitch that 19 blocks K+ channels in its trans confirmation and unblocks in cis. A main difference is that DENAQ is red-shifted and photoisomerizes trans to cis within the visible spectrum (450550nm) then rapidly relaxes back to the trans conformation in the dark (within ~600msec). Also, DENAQ has been shown to be responsive to lower intensity light than either AAQ or many optogenetic strategies. Recently investigators have begun exploring new photoswitches capable of acting in AMPA receptors, via the photoswitchable excitatory amino acid (ATA).66 ATA is active in the dark, and photoisomerizes from trans to cis in blue light (480nm). The distribution of AMPA receptors causes a group of RGCs to respond to blue light, while another group fires in response to darkness.67 A combination of specific photoswitches may facilitate restoring ON-OFF responses, though remodeling leading to altered glutamate receptor expression levels will likely be of concern.68 Photoswitches also demonstrate a shorter lifespan than optogenetics, which may make them a potential diagnostic tool.41 1.4.3 Cellular Replacement Strategies In the majority of the central nervous system of mammals, once a neuron dies it is not replaced, which is the case in the mammalian retina. The advent of stem cell techniques brought with it the capabilities of differentiating new cells in vitro, with the hope of replacement of cells or entire organs grown specifically for an individual. Since the death of photoreceptors is the precipitating feature of all retinal degenerative diseases, photoreceptor replacement strategies are being investigated as a potential treatment. Photoreceptor transplants have had early success with surviving, for at least a few weeks, and have been found to generate synapses in the neural retina.69-70 It has been shown that 20 following introduction of embryonic stem cell derived photoreceptors, pupillary responses are increased in visually deficient mice,71 though it is unclear whether that is a response of photoreceptors or a change in the response of intrinsically-photosensitive retinal ganglion cells (ipRGCs). Further research is still ongoing to determine the long-term viability of transplanted cells and their functionality in the context of the total retina. In addition, our group has found that following the transplantation of fetal retina there is subsequent remodeling that occurs as fast as, or potentially faster than, that seen in degenerate retina without intervention.72 1.4.4 Bionic Implants Beyond the biologic interventions currently being proposed, another methodology aims to intervene using principles similar to the widely successful cochlear implant; in this case, the retinal implant. Certain retinal implants, such as the Argus II epiretinal implant that is placed next to the ganglion cell layer in the vitreous, stimulate nearby neurons to fire action potentials in a pattern controlled by a video camera headset. This approach has the potential to provide for coarse vision, allowing increased independence and mobility among individuals affected by retinal degenerations.50, 73-74 There are two types of retinal implants: subretinal and epiretinal.3 A subretinal implant is surgically implanted under the retina between the photoreceptor layer (or former photoreceptor layer in the case of retinal degenerations) and the RPE, and uses thousands of microdiodes to generate light responses to be processed through the neural retina. An epiretinal implant is surgically placed along the ganglion cell layer of the retina, which does not require detachment of the retina from the RPE and generates action potentials in the associated ganglion cells based on images 21 from an external camera, bypassing neural retina processing. Both of these techniques have certain advantages. The subretinal implant has the capacity to utilize any circuitry remaining in the neural retina; however, as discussed in the remodeling section, it is likely that the circuitry of late retinal degenerations is corrupted. The epiretinal implant is currently in clinical trials (Argus II), which are showing distinct promise,73-75 although it is unable to mimic the wide variation in signals normally encoded by ganglions cells 41 (direction selectivity, ON versus OFF versus ON-OFF signals, and only minimal centersurround responses). 1.4.5 Genetic Approaches Among the therapeutic interventions for vision loss, genetic approaches are especially favored in cases where a specific gene defect is known to be the direct cause of vision loss. One such disorder is Leber's Congenital Amaurosis (LCA), in which one form is specifically caused by a loss of function mutation in the RPE specific gene RPE-65. LCA is an early onset retinal degeneration, where affected individuals typically experience severe vision loss in the first few years of life. Gene-replacement therapy delivered via adeno-associated viruses (AAVs) has been used to introduce functional RPE-65, and has shown promise in restoring proper RPE function in animal models, especially the Briard dog76, and patients.77-81 Other gene therapies for replacing defective genes have been tested with x-linked mutations leading to RP with successful prevention of much of the early photoreceptor loss.82 A primary difficulty for the exclusive use of gene therapy to prevent or slow vision loss is the wide array of genetic defects that can lead to inherited retinal degenerations and the multiple environmental factors that contribute to the generation of 22 AMD.32 More importantly, intercession in most cases of recessive disease must occur perinatally, posing severe challenges for the medical ethics of early genetic manipulation. Also, many diseases are autosomal dominant in which a single defective gene can override the function of a normal gene and lead to photoreceptor death. Approaches to these disorders require suppression of the defective gene with methods such as shRNA targeting or new CRISPR/Cas9 gene editing methods. Even with these tools, there is currently no therapeutic path for human autosomal disease. While autosomal diseases tend to progress more slowly, extending into adulthood (via unknown mechanisms), it is not clear that the critical defect leading to widespread degeneration is operation of the defective gene per se or accumulation of toxic protein, in which case late intervention will be of little use. It is also important to note that despite the success of RPE65 gene therapy, the treated individuals still experience progressive cell death and slow rod kinetics83 which will require further research before the gene therapy can be deemed a full rescue. 1.4.6 Requirements for Therapeutic Interventions to Be Effective The generation of therapeutic interventions and treatment for photoreceptorinitiated vision loss is a lofty goal with a number of avenues of exploration. All therapeutic interventions require a functional neural retina. Therapies that hope to utilize the neural processing afforded by the inner retina have a number of difficult problems to overcome. Cell-replacement therapies are unlikely to result in long-term vision restoration or preservation unless the initial insult leading to photoreceptor degeneration is also directly treated. It also has not been shown that the synapses made by newly introduced photoreceptors produce the same visual acuity as 23 native photoreceptors (not even within an order of magnitude), nor have they been shown to persist. For most therapeutics targeted at stimulation of bipolar cells to be effective, regardless of whether via optogenetics, photoswitches, or subretinal implant, the primary circuitry of the retina must remain intact in addition to the metabolic support of surrounding glia. The basic idea that an epiretinal implant only stimulates ganglion cells, or that ganglion cells transduced by optogenetics or photoswitches could function independently is complicated by extensive coupling between ganglion cells and amacrine cells, as well as penetration of currents deep into retinal tissue. Previous data indicate that alterations in circuitry are present early in remodeling and it is unclear to what extent this affects the signals that are communicated to ganglion cells. We certainly know that retinal ganglion cells begin to display massive uncorrelated spiking in late phase 2.84-87 In addition, the early subtle variation of Müller cell metabolic phenotypes leads to questions regarding how widely Müller cells will vary with respect to one another, and how it will affect the metabolic support provided by Müller cells, particularly in the case of glutamate because excitotoxicity has been proposed as a potential contributing factor to retinal degeneration.25, 88-89 Clarifying these mechanisms is a goal of this dissertation. Retinal degenerative diseases have predominately been evaluated in rodent models, which have short life-spans, or shortly after the total loss of photo-receptors in large-eye model organisms. In contrast, humans live for decades with these disorders and are often functionally blind for years to decades of life. Many proposed interventions must target late-stage patients that are unable to perceive light, and some target ganglion cells with the assumption they will be the last remaining targets of intervention. We do not know the complete order or extent to which the neural retina deteriorates years after most 24 photoreceptors have died. Again, this dissertation will clarify much of that uncertainty. In summary, this dissertation aims to characterize the structural, biochemical, and metabolic changes in late-stage retinal remodeling years after all photoreceptors have degenerated, and evaluate their implications for therapeutic interventions. 25 Figure 1.1 Schematic of the major histological and cellular classifications of the retina. Adapted from (http://webvision.med.utah.edu/book/part-i-foundations/simple-anatomyof-the-retina/) 26 Figure 1.2 Remodeling of the mammalian retina in retinal degenerations. (From Marc et al. 200337) Normal retina demonstrating proper lamination and organization of the mammalian retina. Some slight variations, such as vasculature, may vary with species. Phase 1 (a) shortening of rod outer segments. (b) Shortening of cone outer segments. (c) Initial death of rod soma. (d) Ectopic rod axons descending beyond opl. Phase 2 (e) Widespread rod death. (f) Loss of cone outer segments. (g) Ectopic neurite sprouting by amacrine cells. (h) Cone cell death. (i)Müller cell soma translocation. (j)Early stage of the glial seal. (k) Horizontal cell hypertrophy. (l) Horizontal cell neurite sprouting. (m+n) Bipolar cell dendritic retraction. Early Phase 3 (o) Fully formed glial seal. (p) Increased Müller cell hypertrophy. (q) Initial neural retina neuronal loss. (r) Neurite fascicle formation. (s) Early microneuroma formation. Mid Phase 3 (t) Increased multicellular fascicle formation. (u) Microneuroma growth. (v+w) Neuronal translocation. (x) Progressive cell death. Late Phase 3 (y) Regression of microneuromas due to cell death. (z) Progressive cell death. (aa) Continued Müller cell hypertrophy. (bb) Hypertrophy and vessel invasion. (cc) RPE cell death. (dd) Invasion of the RPE into the neural retina. 27 1.5 References 1. Foster, A.; Resnikoff, S., The Impact of Vision 2020 on Global Blindness. Eye (Lond) 2005, 19 (10), 1133-1135. 2. Stingl, K.; Bartz-Schmidt, K. U.; Besch, D.; Braun, A.; Bruckmann, A.; Gekeler, F.; Greppmaier, U.; Hipp, S.; Hortdorfer, G.; Kernstock, C.; Koitschev, A.; Kusnyerik, A.; Sachs, H.; Schatz, A.; Stingl, K. T.; Peters, T.; Wilhelm, B.; Zrenner, E., Artificial Vision with Wirelessly Powered Subretinal Electronic Implant Alpha-IMS. Proc Biol Sci 2013, 280 (1757), 20130077. 3. Zrenner, E., The Subretinal Implant: Can Microphotodiode Arrays Replace Degenerated Retinal Photoreceptors to Restore Vision? Ophthalmologica 2002, 216 8-20; discussion 52-23. 4. Zrenner, E., Fighting Blindness with Microelectronics. Sci Transl Med 2013, 5 (210), 210. 5. Busskamp, V.; Duebel, J.; Balya, D.; Fradot, M.; Viney, T. J.; Siegert, S.; Groner, A. C.; Cabuy, E.; Forster, V.; Seeliger, M.; Biel, M.; Humphries, P.; Paques, M.; Mohand-Said, S.; Trono, D.; Deisseroth, K.; Sahel, J. A.; Picaud, S.; Roska, B., Genetic Reactivation of Cone Photoreceptors Restores Visual Responses in Retinitis Pigmentosa. Science 2010, 329 (5990), 413-417. 6. Busskamp, V.; Picaud, S.; Sahel, J. A.; Roska, B., Optogenetic Therapy for Retinitis Pigmentosa. Gene Ther 2012, 19 (2), 169-175. 7. Busskamp, V.; Roska, B., Optogenetic Approaches to Restoring Visual Function in Retinitis Pigmentosa. Curr Opin Neurobiol 2011, 21 (6), 942-946. 8. Lagali, P. S.; Balya, D.; Awatramani, G. B.; Munch, T. A.; Kim, D. S.; Busskamp, V.; Cepko, C. L.; Roska, B., Light-Activated Channels Targeted to ON Bipolar Cells Restore Visual Function in Retinal Degeneration. Nat Neurosci 2008, 11 (6), 667-675. 9. Nirenberg, S.; Pandarinath, C., Retinal Prosthetic Strategy with the Capacity to Restore Normal Vision. Proc Natl Acad Sci U S A 2012, 109 (37), 15012-15017. 10. Margalit, E.; Maia, M.; Weiland, J. D.; Greenberg, R. J.; Fujii, G. Y.; Torres, G.; Piyathaisere, D. V.; O'Hearn, T. M.; Liu, W.; Lazzi, G.; Dagnelie, G.; Scribner, D. A.; de Juan, E., Jr.; Humayun, M. S., Retinal Prosthesis for the Blind. Surv Ophthalmol 2002, 47 (4), 335-356. 11. Jiang, L.; Zhang, H.; Dizhoor, A. M.; Boye, S. E.; Hauswirth, W. W.; Frederick, J. M.; Baehr, W., Long-Term RNA Interference Gene Therapy in a Dominant Retinitis Pigmentosa Mouse Model. Proc Natl Acad Sci U S A 2011, 108 (45), 18476-18481. 28 12. Doroudchi, M. M.; Greenberg, K. P.; Liu, J.; Silka, K. A.; Boyden, E. S.; Lockridge, J. A.; Arman, A. C.; Janani, R.; Boye, S. E.; Boye, S. L.; Gordon, G. M.; Matteo, B. C.; Sampath, A. P.; Hauswirth, W. W.; Horsager, A., Virally Delivered Channelrhodopsin-2 Safely and Effectively Restores Visual Function in Multiple Mouse Models of Blindness. Mol Ther 2011, 19 (7), 1220-1229. 13. Petrs-Silva, H.; Linden, R., Advances in Gene Therapy Technologies to Treat Retinitis Pigmentosa. Clin Ophthalmol 2014, 8, 127-136. 14. Marc, R. E., The Structure of Vertebrate Retinas. In The Retinal Basis of Vision, Toyoda, J.-I.; Murakami, M.; Kaneko, A.; Saito, T., Eds. Elsevier: Amsterdam, 1999; pp 3-19. 15. Masland, R. H., The Neuronal Organization of the Retina. Neuron 2012, 76 (2), 266-280. 16. Kolb, H.; Gouras, P., Electron Microscopic Observations of Human Retinitis Pigmentosa, Dominantly Inherited. Invest Ophthalmol 1974, 13 (7), 487-498. 17. Famiglietti, E. V., Jr.; Kaneko, A.; Tachibana, M., Neuronal Architecture of On and Off Pathways to Ganglion Cells in Carp Retina. Science 1977, 198 (4323), 12671269. 18. Lauritzen, J. S.; Anderson, J. R.; Jones, B. W.; Watt, C. B.; Mohammed, S.; Hoang, J. V.; Marc, R. E., ON Cone Bipolar Cell Axonal Synapses in the OFF Inner Plexiform Layer of the Rabbit Retina. J Comp Neurol 2013, 521 (5), 977-1000. 19. Marc, R. E., Retinal Neurotransmitters. In The Visual Neurosciences, Chalupa, L. M.; Werner, J., Eds. MIT Press: Cambridge MA, 2004; pp 315-330. 20. Marc, R. E., Information Processing: Amacrine Cells. In Encyclopedia of the Eye, Dartt, D., Ed. Elsevier: 2010; Vol. 2, pp 330-337. 21. Marc, R. E., Functional Neuroanatomy of the Retina. In Duane's Ophthalmology, Tasman, W.; Jaeger, E. A., Eds. Lippincott Williams & Wilkins: Philadelphia, 2009. 22. Dowling, J. E.; Werblin, F. S., Organization of Retina of the Mudpuppy, Necturus Maculosus. I. Synaptic Structure. J Neurophysiol 1969, 32 (3), 315-338. 23. Rodieck, R. W., Visual Pathways. Annu Rev Neurosci 1979, 2, 193-225. 24. Byrne, L. C.; Khalid, F.; Lee, T.; Zin, E. A.; Greenberg, K. P.; Visel, M.; Schaffer, D. V.; Flannery, J. G., AAV-Mediated, Optogenetic Ablation of Muller Glia Leads to Structural and Functional Changes in the Mouse Retina. PLoS One 2013, 8 (9), e76075. 25. Reichenbach, A.; Bringmann, A., New Functions of Muller Cells. Glia 2013, 61 (5), 651-678. 29 26. Riepe, R. E.; Norenburg, M. D., Muller Cell Localisation of Glutamine Synthetase in Rat Retina. Nature 1977, 268 (5621), 654-655. 27. Pow, D. V.; Robinson, S. R., Glutamate in Some Retinal Neurons is Derived Solely from Glia. Neuroscience 1994, 60 (2), 355-366. 28. Jones, B. W.; Marc, R. E., Retinal Remodeling During Retinal Degeneration. Exp Eye Res 2005, 81 (2), 123-137. 29. Medicine, U. S. N. L. o. Bardet-Biedl Syndrome. https://ghr.nlm.nih.gov/condition/bardet-biedl-syndrome (accessed March, 22). 30. Koenig, R., Bardet-Biedl Syndrome and Usher Syndrome. Dev Ophthalmol 2003, 37, 126-140. 31. Hartong, D. T.; Berson, E. L.; Dryja, T. P., Retinitis Pigmentosa. Lancet 2006, 368 (9549), 1795-1809. 32. Armstrong, R. A.; Mousavi, M., Overview of Risk Factors for Age-Related Macular Degeneration (AMD). J Stem Cells 2015, 10 (3), 171-191. 33. Jones, B. W.; Marc, R. E.; Pfeiffer, R. L. Retinal Remodeling and Plasticity. http://webvision.med.utah.edu/. 34. Jones, B. W.; Pfeiffer, R. L.; Ferrell, W. D.; Watt, C. B.; Tucker, J. F.; Marc, R. E., Retinal Remodeling And Metabolic Alterations in Human AMD. Front Cell Neurosci 2016, 10, 103. 35. Wolff, E., A Pathology of the Eye. 1st edition ed.; Blakiston: Philadelphia, 1935. 36. Humayun, M. S.; Prince, M.; de Juan, E., Jr.; Barron, Y.; Moskowitz, M.; Klock, I. B.; Milam, A. H., Morphometric Analysis of the Extramacular Retina from Postmortem Eyes with Retinitis Pigmentosa. Invest Ophthalmol Vis Sci 1999, 40 (1), 143-148. 37. Marc, R. E.; Jones, B. W.; Watt, C. B.; Strettoi, E., Neural Remodeling in Retinal Degeneration. Prog Retin Eye Res 2003, 22 (5), 607-655. 38. Marc, R. E.; Jones, B. W.; Watt, C. B.; Strettoi, E., Neural remodeling in retinal degeneration. Prog Retin Eye Res 2003, 22 (5), 607-655. 39. Jones, B. W.; Pfeiffer, R. L.; Ferrell, W. D.; Watt, C. B.; Marmor, M.; Marc, R. E., Retinal Remodeling in Human Retinitis Pigmentosa. Exp Eye Res 2016, 150, 149-165. 40. Marc, R. E., Injury and Repair: Retinal Remodeling. In Encyclopedia of the Eye, Dartt, D., Ed. Elsevier: 2010; Vol. 2, pp 414-420. 41. Marc, R.; Pfeiffer, R.; Jones, B., Retinal Prosthetics, Optogenetics, and Chemical Photoswitches. ACS Chem Neurosci 2014, 5 (10), 895-901. 30 42. Strettoi, E.; Porciatti, V.; Falsini, B.; Pignatelli, V.; Rossi, C., Morphological and Functional Abnormalities in the Inner Retina of the rd/rd Mouse. J Neurosci 2002, 22 (13), 5492-5504. 43. Strettoi, E.; Pignatelli, V.; Rossi, C.; Porciatti, V.; Falsini, B., Remodeling of Second-Order Neurons in the Retina of rd/rd Mutant Mice. Vision Res 2003, 43 (8), 867877. 44. Lewis, G. P.; Linberg, K. A.; Fisher, S. K., Neurite Outgrowth from Bipolar and Horizontal Cells After Experimental Retinal Detachment. Invest Ophthalmol Vis Sci 1998, 39 (2), 424-434. 45. Jones, B. W.; Kondo, M.; Terasaki, H.; Watt, C. B.; Rapp, K.; Anderson, J.; Lin, Y.; Shaw, M. V.; Yang, J. H.; Marc, R. E., Retinal Remodeling in the Tg P347L Rabbit, a Large-Eye Model of Retinal Degeneration. J Comp Neurol 2011, 519 (14), 2713-27133. 46. Jones, B. W.; Kondo, M.; Terasaki, H.; Lin, Y.; McCall, M.; Marc, R. E., Retinal remodeling. Jpn J Ophthalmol 2012, 56 (4), 289-306. 47. Marc, R. E.; Jones, B. W.; Anderson, J. R.; Kinard, K.; Marshak, D. W.; Wilson, J. H.; Wensel, T.; Lucas, R. J., Neural Reprogramming in Retinal Degeneration. Invest Ophthalmol Vis Sci 2007, 48 (7), 3364-3371. 48. Fariss, R. N.; Li, Z. Y.; Milam, A. H., Abnormalities in Rod Photoreceptors, Amacrine Cells, and Horizontal Cells in Human Retinas with Retinitis Pigmentosa. Am J Ophthalmol 2000, 129 (2), 215-223. 49. Marc, R. E.; Jones, B. W., Retinal Remodeling in Inherited Photoreceptor Degenerations. Mol Neurobiol 2003, 28 (2), 139-147. 50. Ahuja, A. K.; Dorn, J. D.; Caspi, A.; McMahon, M. J.; Dagnelie, G.; Dacruz, L.; Stanga, P.; Humayun, M. S.; Greenberg, R. J.; Argus, I. I. S. G., Blind Subjects Implanted with the Argus II Retinal Prosthesis are able to Improve Performance in a Spatial-Motor Task. Br J Ophthalmol 2011, 95 (4), 539-543. 51. Lamba, D. A.; Karl, M. O.; Reh, T. A., Strategies for Retinal Repair: Cell Replacement and Regeneration. Prog Brain Res 2009, 175, 23-31. 52. Nagel, G.; Szellas, T.; Huhn, W.; Kateriya, S.; Adeishvili, N.; Berthold, P.; Ollig, D.; Hegemann, P.; Bamberg, E., Channelrhodopsin-2, a Directly Light-Gated CationSelective Membrane Channel. Proc Natl Acad Sci U S A 2003, 100 (24), 13940-13945. 53. Schobert, B.; Lanyi, J. K., Halorhodopsin is a Light-Driven Chloride Pump. J Biol Chem 1982, 257 (17), 10306-10313. 54. Han, X.; Boyden, E. S., Multiple-Color Optical Activation, Silencing, and Desynchronization of Neural Activity, with Single-Spike Temporal Resolution. PLoS One 2007, 2 (3), e299. 31 55. Zhang, F.; Wang, L. P.; Brauner, M.; Liewald, J. F.; Kay, K.; Watzke, N.; Wood, P. G.; Bamberg, E.; Nagel, G.; Gottschalk, A.; Deisseroth, K., Multimodal Fast Optical Interrogation of Neural Circuitry. Nature 2007, 446 (7136), 633-639. 56. Boyden, E. S.; Zhang, F.; Bamberg, E.; Nagel, G.; Deisseroth, K., MillisecondTimescale, Genetically Targeted Optical Control of Neural Activity. Nat Neurosci 2005, 8 (9), 1263-1268. 57. Bi, A.; Cui, J.; Ma, Y. P.; Olshevskaya, E.; Pu, M.; Dizhoor, A. M.; Pan, Z. H., Ectopic Expression of a Microbial-Type Rhodopsin Restores Visual Responses in Mice with Photoreceptor Degeneration. Neuron 2006, 50 (1), 23-33. 58. Tomita, H.; Sugano, E.; Fukazawa, Y.; Isago, H.; Sugiyama, Y.; Hiroi, T.; Ishizuka, T.; Mushiake, H.; Kato, M.; Hirabayashi, M.; Shigemoto, R.; Yawo, H.; Tamai, M., Visual Properties of Transgenic Rats Harboring the Channelrhodopsin-2 Gene Regulated by the thy-1.2 Promoter. PLoS One 2009, 4 (11), e7679. 59. Tomita, H.; Sugano, E.; Isago, H.; Hiroi, T.; Wang, Z.; Ohta, E.; Tamai, M., Channelrhodopsin-2 Gene Transduced into Retinal Ganglion Cells Restores Functional Vision in Genetically Blind Rats. Exp Eye Res 2010, 90 (3), 429-436. 60. Zhang, Y.; Ivanova, E.; Bi, A.; Pan, Z. H., Ectopic Expression of Multiple Microbial Rhodopsins Restores ON and OFF Light Responses in Retinas with Photoreceptor Degeneration. J Neurosci 2009, 29 (29), 9186-9196. 61. Greenberg, K. P.; Pham, A.; Werblin, F. S., Differential targeting of optical neuromodulators to ganglion cell soma and dendrites allows dynamic control of centersurround antagonism. Neuron 2011, 69 (4), 713-720. 62. Gaub, B. M.; Berry, M. H.; Holt, A. E.; Isacoff, E. Y.; Flannery, J. G., Optogenetic Vision Restoration Using Rhodopsin for Enhanced Sensitivity. Mol Ther 2015, 23 (10), 1562-1571. 63. Lin, B.; Koizumi, A.; Tanaka, N.; Panda, S.; Masland, R. H., Restoration of Visual Function in Retinal Degeneration Mice by Ectopic Expression of Melanopsin. Proc Natl Acad Sci U S A 2008, 105 (41), 16009-16014. 64. Polosukhina, A.; Litt, J.; Tochitsky, I.; Nemargut, J.; Sychev, Y.; De Kouchkovsky, I.; Huang, T.; Borges, K.; Trauner, D.; Van Gelder, R. N.; Kramer, R. H., Photochemical Restoration of Visual Responses in Blind Mice. Neuron 2012, 75 (2), 271-282. 65. Tochitsky, I.; Polosukhina, A.; Degtyar, V. E.; Gallerani, N.; Smith, C. M.; Friedman, A.; Van Gelder, R. N.; Trauner, D.; Kaufer, D.; Kramer, R. H., Restoring Visual Function to Blind Mice with a Photoswitch that Exploits Electrophysiological Remodeling of Retinal Ganglion Cells. Neuron 2014, 81 (4), 800-813. 32 66. Stawski, P.; Sumser, M.; Trauner, D., A Photochromic Agonist of AMPA Receptors. Angew Chem Int Ed Engl 2012, 51 (23), 5748-5751. 67. Laprell, L.; Hull, K.; Stawski, P.; Schon, C.; Michalakis, S.; Biel, M.; Sumser, M. P.; Trauner, D., Restoring Light Sensitivity in Blind Retinae Using a Photochromic AMPA Receptor Agonist. ACS Chem Neurosci 2016, 7 (1), 15-20. 68. Lin, Y.; Jones, B. W.; Liu, A.; Vazquez-Chona, F. R.; Lauritzen, J. S.; Ferrell, W. D.; Marc, R. E., Rapid Glutamate Receptor 2 Trafficking During Retinal Degeneration. Mol Neurodegener 2012, 7, 7. 69. Santos-Ferreira, T.; Volkner, M.; Borsch, O.; Haas, J.; Cimalla, P.; Vasudevan, P.; Carmeliet, P.; Corbeil, D.; Michalakis, S.; Koch, E.; Karl, M. O.; Ader, M., Stem Cell-Derived Photoreceptor Transplants Differentially Integrate Into Mouse Models of Cone-Rod Dystrophy. Invest Ophthalmol Vis Sci 2016, 57 (7), 3509-3520. 70. MacLaren, R. E.; Pearson, R. A.; MacNeil, A.; Douglas, R. H.; Salt, T. E.; Akimoto, M.; Swaroop, A.; Sowden, J. C.; Ali, R. R., Retinal Repair by Transplantation of Photoreceptor Precursors. Nature 2006, 444 (7116), 203-207. 71. Singh, M. S.; Charbel Issa, P.; Butler, R.; Martin, C.; Lipinski, D. M.; Sekaran, S.; Barnard, A. R.; MacLaren, R. E., Reversal of End-Stage Retinal Degeneration and Restoration of Visual Function by Photoreceptor Transplantation. Proc Natl Acad Sci U S A 2013, 110 (3), 1101-1106. 72. Seiler, M. J.; Jones, B. W.; Aramant, R. B.; Yang, P. B.; Keirstead, H. S.; Marc, R. E., Computational Molecular Phenotyping of Retinal Sheet Transplants to Rats with Retinal Degeneration. Eur J Neurosci 2012, 35 (11), 1692-1704. 73. Dorn, J. D.; Ahuja, A. K.; Caspi, A.; da Cruz, L.; Dagnelie, G.; Sahel, J. A.; Greenberg, R. J.; McMahon, M. J.; gus, I. I. S. G., The Detection of Motion by Blind Subjects With the Epiretinal 60-Electrode (Argus II) Retinal Prosthesis. JAMA Ophthalmol 2013, 131 (2), 183-189. 74. da Cruz, L.; Coley, B. F.; Dorn, J.; Merlini, F.; Filley, E.; Christopher, P.; Chen, F. K.; Wuyyuru, V.; Sahel, J.; Stanga, P.; Humayun, M.; Greenberg, R. J.; Dagnelie, G.; Argus, I. I. S. G., The Argus II Epiretinal Prosthesis System Allows Letter and Word Reading and Long-Term Function in Patients with Profound Vision Loss. Br J Ophthalmol 2013, 97 (5), 632-636. 75. Ho, A. C.; Humayun, M. S.; Dorn, J. D.; da Cruz, L.; Dagnelie, G.; Handa, J.; Barale, P. O.; Sahel, J. A.; Stanga, P. E.; Hafezi, F.; Safran, A. B.; Salzmann, J.; Santos, A.; Birch, D.; Spencer, R.; Cideciyan, A. V.; de Juan, E.; Duncan, J. L.; Eliott, D.; Fawzi, A.; Olmos de Koo, L. C.; Brown, G. C.; Haller, J. A.; Regillo, C. D.; Del Priore, L. V.; Arditi, A.; Geruschat, D. R.; Greenberg, R. J.; Argus, I. I. S. G., Long-Term Results from an Epiretinal Prosthesis to Restore Sight to the Blind. Ophthalmology 2015, 122 (8), 1547-1554. 33 76. Le Meur, G.; Stieger, K.; Smith, A. J.; Weber, M.; Deschamps, J. Y.; Nivard, D.; Mendes-Madeira, A.; Provost, N.; Pereon, Y.; Cherel, Y.; Ali, R. R.; Hamel, C.; Moullier, P.; Rolling, F., Restoration of Vision in RPE65-Deficient Briard Dogs using an AAV Serotype 4 Vector That Specifically Targets the Retinal Pigmented Epithelium. Gene Ther 2007, 14 (4), 292-303. 77. Sahel, J. A.; Roska, B., Gene Therapy for Blindness. Annu Rev Neurosci 2013, 36, 467-488. 78. Jacobson, S. G.; Cideciyan, A. V.; Roman, A. J.; Sumaroka, A.; Schwartz, S. B.; Heon, E.; Hauswirth, W. W., Improvement and Decline in Vision with Gene Therapy in Childhood Blindness. N Engl J Med 2015, 372 (20), 1920-1926. 79. Acland, G. M.; Aguirre, G. D.; Ray, J.; Zhang, Q.; Aleman, T. S.; Cideciyan, A. V.; Pearce-Kelling, S. E.; Anand, V.; Zeng, Y.; Maguire, A. M.; Jacobson, S. G.; Hauswirth, W. W.; Bennett, J., Gene Therapy Restores Vision in a Canine Model of Childhood Blindness. Nat Genet 2001, 28 (1), 92-95. 80. Cideciyan, A. V.; Hauswirth, W. W.; Aleman, T. S.; Kaushal, S.; Schwartz, S. B.; Boye, S. L.; Windsor, E. A.; Conlon, T. J.; Sumaroka, A.; Pang, J. J.; Roman, A. J.; Byrne, B. J.; Jacobson, S. G., Human RPE65 Gene Therapy for Leber Congenital Amaurosis: Persistence of Early Visual Improvements and Safety at 1 Year. Hum Gene Ther 2009, 20 (9), 999-1004. 81. Cideciyan, A. V.; Hauswirth, W. W.; Aleman, T. S.; Kaushal, S.; Schwartz, S. B.; Boye, S. L.; Windsor, E. A.; Conlon, T. J.; Sumaroka, A.; Roman, A. J.; Byrne, B. J.; Jacobson, S. G., Vision 1 Year After Gene Therapy for Leber's Congenital Amaurosis. N Engl J Med 2009, 361 (7), 725-727. 82. Beltran, W. A.; Cideciyan, A. V.; Iwabe, S.; Swider, M.; Kosyk, M. S.; McDaid, K.; Martynyuk, I.; Ying, G. S.; Shaffer, J.; Deng, W. T.; Boye, S. L.; Lewin, A. S.; Hauswirth, W. W.; Jacobson, S. G.; Aguirre, G. D., Successful Arrest of Photoreceptor and Vision Loss Expands the Therapeutic Window of Retinal Gene Therapy to Later Stages of Disease. Proc Natl Acad Sci U S A 2015, 112 (43), 5844-5853. 83. Jacobson, S. G.; Cideciyan, A. V.; Aguirre, G. D.; Roman, A. J.; Sumaroka, A.; Hauswirth, W. W.; Palczewski, K., Improvement in Vision: a New Goal for Treatment of Hereditary Retinal Degenerations. Expert Opin Orphan Drugs 2015, 3 (5), 563-575. 84. Euler, T.; Schubert, T., Multiple Independent Oscillatory Networks in the Degenerating Retina. Front Cell Neurosci 2015, 9, 444. 85. Margolis, D. J.; Newkirk, G.; Euler, T.; Detwiler, P. B., Functional Stability of Retinal Ganglion Cells After Degeneration-Induced Changes in Synaptic Input. J Neurosci 2008, 28 (25), 6526-6536. 34 86. Stasheff, S. F., Emergence of Sustained Spontaneous Hyperactivity and Temporary Preservation of OFF Responses in Ganglion Cells of the Retinal Degeneration (rd1) Mouse. J Neurophysiol 2008, 99 (3), 1408-1421. 87. Menzler, J.; Zeck, G., Network Oscillations in Rod-Degenerated Mouse Retinas. J Neurosci 2011, 31 (6), 2280-2291. 88. Reichenbach, A.; Bringmann, A., Muller Cells in the Healthy Retina. In Muller Cells in the Healthy and Diseased Retina, Springer Science+Business Meddia, LLC: New York, 2010; pp 35-299. 89. Bringmann, A.; Grosche, A.; Pannicke, T.; Reichenbach, A., GABA and Glutamate Uptake and Metabolism in Retinal Glial (Muller) Cells. Front Endocrinol 2013, 4, 1-14. CHAPTER 2 MATERIALS AND METHODS The main body of the work presented in this dissertation relies on computational molecular phenotyping (CMP). CMP is a robust method for quantitative small molecule immunohistochemistry combined with cluster analysis to identify cell classes found in normal tissue, and identify metabolic variation in degenerative tissue. This chapter will introduce the animal model predominately used throughout this study. The background associated with the methods and model organism is extensive and presented in greater detail here to aid the reader in understanding the recurrent concepts underlying the body of work in the following chapters. 2.1 Computational Molecular Phenotyping (CMP) Accurate classification of cells within biological systems is an important basic component of understanding how systems function. For decades the role of metabolism in cellular systems has been predominantly assessed using homogenates, which yields information about total concentrations of small molecules and enzymes across an entire tissue, but cannot divulge information about single cells, let alone interactions between cells. The nervous system is an exceedingly complex heterocellular system, in which small molecules act within cells and on neighboring cells through tightly regulated, cell-specific 36 mechanisms. Based on early studies in the late 1970s and early 1980s that demonstrated cells had combinations of neurotransmitters and neuropeptides at different levels within single cells,1-2 it stands to reason that the concentration of small molecules contained within the various cellular compartments might be unique to each morphologically and functionally distinct type of cell. This concept was formalized as the signature hypothesis in the 1980s, which posited that each morphologically and functionally unique type of neuron would also possess distinct neurochemical compositions.3-6 Computational molecular phenotyping is a robust method for harnessing the quantitative aspects of electron microscopy compatible, postembedding immunocytochemistry,7 and uses pattern recognition to classify cells into different cellular classes based on their signatures, which are defined by quantitative differences in small molecule concentration mixtures.8-10 2.1.1 Molecular Trapping and IgG Labeling One of the predominant issues with respect to immunohistochemical detection of small molecules is the efficient trapping of small molecules in tissue. Formaldehyde (HCHO) is a 1-carbon small molecule that rapidly penetrates tissue and stops molecular interaction through aldehyde (-CHO) binding to terminal amines on specific proteins and small molecules. Formaldehyde is the traditional aldehyde for immunohistochemistry. However, its ability to cross-link protein amines via the formation of methylene bridges occurs relatively slowly; full formaldehyde cross-linking can take several weeks and even that is very sparse. Glutaraldehyde (HCO-(CH2)3-CHO), on the other hand, is a 5-carbon molecule, with a chain of three methylene groups and two aldehyde groups which rapidly bind with any available amine or imine via nucleophilic attack by N resulting in a Schiff's 37 base formation. The increased number of carbons greatly increases the potential for crosslinking, which generates a tightly cross-linked protein matrix.11 The generated matrix of glutaraldehyde conjugated amines can in turn bind to other free aldehydes via imine condensation, efficiently trapping small molecules quickly in cells. A potential complication of using solely glutaraldehyde in fixation is the two aldehyde groups. Assuming each glutaraldehyde molecule has a single aldehyde group binding an amine associated with protein, there will be an equal number of aldehyde groups available for small molecule binding. However, the number of free amines found in a cell is approximately 10 fold less than protein matrix amines, leaving a large number of unbound aldehyde groups. These unbound aldehyde groups will cause an increase in background due to nonspecific binding by antibodies.11 If solely formaldehyde fixation is used, the slow, and inefficient, cross-linking will cause a large percentage of small molecules to be lost during processing. Therefore, our solution is the use of a mixed aldehyde fixation protocol. By using 1% formaldehyde (333 mM, with 333 mM free aldehyde groups) and 2.5% glutaraldehyde (250 mM, with 500 mM free aldehyde groups), small molecules are still effectively trapped while the total number of free aldehydes is reduced. Using a standard primary amine-secondary amine model12 of trapping where each amine N can capture two aldehyde carbons leads to a minimum of ≈ 84% capture by glutaraldehyde cross-linking, assuming no glutaraldehyde-formaldehyde exchange. But exchange is quite likely during Schiff's base formation as the nucleophilic attack is strongly affected by water molecules at the condensation site and highly reversible under even neutral conditions. Assuming one switch event during condensation leads to 94% capture and two switches yields 97.5% amino acid capture, this is consistent with radiochemistry estimates 38 suggesting glutaraldehyde-formaldehyde mixtures lead to >86-95% retention of [3H]amino acids.13 Immunoglobulins (IgGs) have been developed for detection of small molecules for almost a century. In the 1920s and 30s, Karl Landsteiner and colleagues performed pioneering research introducing the concept of haptens.14 Haptens, whether naturally occurring or synthetic, are small molecules that alone generate no immunological response, but when attached to a carrier molecule form a hapten-carrier adduct that is capable of initiating an immunological response. It is this concept that facilitates the generation of antibodies to small molecules such as amino acids. We harness this principle by using IgGs that have a high specificity to individual amino acid-glutaraldehyde adducts.15 The protein matrix generated by glutaraldehyde cross-linking in cells has a mean pore diameter of 2-5nm as calculated simply from the size of glutaraldehyde. IgGs, on the other hand, are approximately 10x7x2nm, which precludes penetration of IgGs into glutaraldehyde-fixed tissue. This limits all epitope detection to the surface and makes small molecule detection proportional to the density presented on the section surface following surface etching to increase epitope mobility and IgG access. IgG binding is largely hydrogen bonding and requires hydrophilic regimes, so it is necessary to remove the plastic epoxide bond matrix. This feature of molecular trapping combined with surface labeling lends itself to direct quantification of small molecules found within a cell independent of section thickness. Tissue penetration by IgGs or ISH probes in frozen section or whole mounts, though helpful in detection of sparse proteins, results in uneven penetration of the antibodies into the various tissue compartments, making quantification impractical. The evaluation of small molecule concentration from these labeled sections is simple because 39 the available epitope density presenting on the surface of the section scales as a function of concentration (independent calculations‡).7, 15 From the accurate small molecule trapping and IgG labeling described above,7, 9, 15 we have one-dimensional information about the concentrations of amino acids contained within single cells in a tissue sample. The standard method for analysis of multiple antibody labeled cells in immunocytochemistry is by using multiple secondary antibodies conjugated to fluorophores reactive to different wavelengths. A key problem with fluorophores is bleaching, or the progressive decrease in brightness due singlet quenching, largely by intermolecular collision in the aqueous state. This makes fluorescent identification of small molecule concentrations in aqueous mounts nonarchival. Goldconjugated secondary antibodies or proteins binding to IgGs used in transmission electron microscopy, are archival16, but have the problem of being too small for easy detection by light microscopy at low density. Reverse-engineering the gold-toning method of photography, chemical reduction of silver nitrate to elemental silver particles plated on to gold particles was first applied to Western blotting techniques17 and then developed in the Marc lab for immunocytochemistry.15 In the presence of colloidal gold, the induction phase of hydroquinone reduced silver nitrate is drastically reduced, and the silver particles nucleate around the gold particle in a logarithmic fashion.18 This feature lends silverintensification to quantitative immunohistochemistry,15 allowing the concentration of silver particles to function as a direct logarithmic readout of intercellular small molecule concentrations. Marc et al., directly calibrated the method against HPLC determinations of ‡ http://marclab.org/technologies/cmp/cmp-substrates/ 40 small molecules in retina.9 2.1.2 CMP Analysis: Pattern Recognition The above section describes how we can obtain an archival quantitative measure of small molecules in a single section. Quantitative immunocytochemistry on thin sections allowed a direct test the signature hypothesis.9, 19-21 To apply these data to test the signature hypothesis, multiple small molecule signatures need to be evaluated simultaneously in single cells. The fact that IgG labeling for CMP is only surface binding, making section thickness irrelevant, lends it to using multiple thin sections through tissue and aligning them computationally. The smallest neuronal soma of the retina is the rod soma, which measures approximately 5µm in diameter (independent measurements from EM data). If the tissue were sectioned at 500nm, as in Ottersen's experiments,7 10 IgG labeled sections could fit within a single cell; however, that would assume the first section was perfectly aligned with the edge of the cell, which is improbable. As serial sections are cut thinner, a single cell being able to be stained for up to 12 small molecules becomes increasingly more probable. For this reason, we cut samples (except where otherwise noted) at 100nm throughout this dissertation. Because these are individually cut sections, computational alignment of sections is necessary for multidimensional cell classification algorithms to be effective. Images of all molecular signals are acquired under constant gain and flux to allow for direct comparable quantification between metabolites. Images were aligned into a registered volume using nonlinear thin-plate-spline warping22 of selected chosen control 41 points. The primary tool was ir-tweak, available in ir-tools from the NCR§ toolset.9-10, 23-26 Once in a registered volume, the multichannel dataset lends itself to pattern recognition via various clustering methods. By assigning each aligned grayscale image to a red, green, or blue color channel (rgb), we can visually identify the most dramatic differences between the metabolic signatures of major cell types and visualize the retina with more depth than traditional histology. However, this is not a statistical measure, and even with extensive rgb visualization of all possible combinations of all aligned images of IgG concentrations generated, we cannot accurately segment all of the metabolically distinct cells of the retina and one cannot see variance. Computationally, clustering (e.g. K-means or isodata) functions by similar principles: features are extracted from data and used by a classifier to subdivide the large mass of data into smaller data groups that share similar statistical features. These groups are termed classes. Rather than handling each IgG labeled registered image in isolation, duplicate, or triplicate, cluster analysis allows us to evaluate all N generated images simultaneously. Of clustering algorithms a popular unsupervised (i.e. no user directed training or sample classes) learning algorithm is k-means. K-means is a rapid method of separating input data into k clusters (groupings of similar points or data vectors) and has been extensively used in multispectral satellite imaging. We view each N-dimensional pixel representation as a data point in Euclidean space (N = individual image channels composed of an IgG labeled grayscale image) (Figure 2.1b). In its most common implementation, the § http://www.sci.utah.edu/download/ncrtoolset 42 k-means algorithm randomly assigns a set of k N-dimensional center points (centroids) in the data space, and decision boundaries are drawn between the centroids k (Figure 2.1c). All of the data points within each decision boundary are grouped, and the new mean of the points is computed, which becomes and new centroid. New boundary lines are computed from the new centroids (Figure 2.1d). The process of computing new centroids and boundaries are iterated until the movement of the centroid falls below a preset threshold, at which point the algorithm is said to converge (Figure 2.1e). The decision boundaries computed when a k-means converges determines the hypervolume containing the data points for each of the k regions, each denoting a metabolic class. We then remap the points back into the x,y conformation of a 2D image with colors assigned to its cluster membership, generating a theme map (Figure 2.1f). The theme map is then used as a mask for extracting individual histogram values of each small molecule, revealing its modal concentration in the class. In symmetrical unimodal histograms, the mode, median, and mean are the same. In general we summarize the data as spatial means, but the histogram always exists as the underlying dataset. The list of N small molecule means that define a class is termed its signature. Defining k can be a difficult task in the case of biological systems and, in fact, there is no formal method to extract the right k, nor is there an ideal clustering algorithm for all problems.27 We do not know a priori how many metabolically distinct cell types are possible, particularily in cases of pathology where the normal metabolic signatures of cells may have disappeared or diverged, but it is simple to compare k-means results using the same initial k, eliminating empty classes and comparing different classes. Visualization of cell-classes by representative theme-maps permits the identification of any inappropriate 43 classes arising from excessively stringent k values (e.g., differentiation of nucleus from cytoplasm within a single cell), and the emergence of over-clustering enables us to determine when the appropriate k for the sample has been reached. The classes produced are then tested for significant separability from one another using Bhattacharyya separability, which is a weighted measure of the degree of overlap between two statistical populations in N-dimensions. Over the course of this project, we found use of two other unsupervised clustering variants, similar to k-means, to be useful. One is an iteratively masking method, using custom software (CMP, Scientific Computing Institute, Salt Lake City, UT). Though the computational algorithm being run is k-means, it has the option to mask components not relevent to identification of metabolic classes (background, kerf, etc.). All k-means algorithms allow this, but CMP is particularly easy to use. This classification is run iteratively with k being incremented by 1 until it converges on statistics consistent with biological classes,9 that is, no subcellular fractionation. The second is the use of the isodata algorithm first used in neural classification by Marc and Jones, 2002. 10 Isodata starts out with a defined k and then merges or splits them based their mutual variances. Isodata is known as a heuristic or self-adjusting algorithm. In practice, data analysis starts CMP as a first screen, then switches to formal k-means (which has more parameter flexibility), and then switches to isodata if necessary. 2.2 P347L Transgenic Rabbit Model of Retinitis Pigmentosa The Tg P347L rabbit is a rabbit model of initially cone-sparing autosomal dominant retinitis pigmentosa (adRP) originally developed in the Kondo laboratory.28 Although there are a large number of animal models of retinal degeneration in drosophila,29-30 zebrafish,31 44 mouse,32-35 rat,36 chicken,37 cat,38-39 dog,40-41 and pig.42 medium sized animals with large eyes are extremely important for the development and testing of therapeutic interventions for retinal degenerative diseases. Testing surgical interventions, such as epi- or subretinal implants, particularly require a large eye model, and the costs of maintaining cats, dogs, or pigs can be prohibitively expensive. Rabbits provide a potential model organism of medium size, with an eye only ~25% smaller than that of a human.43-44 Harnessing the technology of bacterial artificial chromosomes (BACs), Kondo et al., generated a rabbit with a proline to leucine mutation caused by a C-to-T transition at proline 347 in exon 5 of the rabbit rhodopsin gene.28 This Tg P347L rabbit presents with a rod-dominant progressive photoreceptor degeneration, where rod photoreceptors display signs of degeneration and die prior to the degeneration and loss of cones. Like most transgenic models, the Tg P347L rabbit is an overexpression model with ≈ 6 copies of the defective gene.28 By 12 postnatal weeks, ERGs demonstrate a 72% reduction in scotopic a-wave (rod-response) amplitudes, and 95% reduction by 48 weeks. Under photopic conditions, awave (cone-response) amplitudes persisted, though they did show a 65% reduction in amplitude.28 CMP combined with TEM demonstrated, in randomly selected regions, the P347L Tg rabbit has a profound loss of rod photoreceptors by 16 weeks, and a subsequent loss of cones by 40 weeks.23 At 40 weeks TEM reveals regional fluctuations in Müller cell polyribosome and protein levels, while CMP demonstrates glial seal formation, neurite outgrowth, glutamate receptor phenotype switching, and metabolic alterations.23 All of these changes are found in human RP.23-24 Based on these previous studies detailing the early changes in the retina following rod loss and cone outer segment degeneration, the rabbit model is a likely candidate for 45 determining long-term consequences of photoreceptor degeneration and analyzing the potential for retinas to sustain therapeutic interventions long-term. 46 Figure 2.1 K-means generation of theme maps. (b) N-dimensional pixel representation as a data point in Euclidean space. (c) K center points (centroids) are randomly placed in data and boundary lines are drawn. (d) All vectors within boundaries are averaged, and the centroids are moved to the average means. (e) The process is repeated until the centroids no longer move beyond a threshold in the averaging means step. (f) The clusters are then overlaid on top of the image stack and the clusters are used as masks to extract the underlying histograms. 47 48 2.3 References 1. Watt, C. B.; Su, Y. Y.; Lam, D. M., Interactions Between Enkephalin and GABA in Avian Retina. Nature 1984, 311 (5988), 761-763. 2. Burnstock, G., Do Some Nerve Cells Release More than one Transmitter? Neuroscience 1976, 1 (4), 239-248. 3. Lam, D. M.; Li, H. B.; Su, Y. Y.; Watt, C. B., The Signature Hypothesis: CoLocalizations of Neuroactive Substances as Anatomical Probes for Circuitry Analyses. Vision Res 1985, 25 (10), 1353-1364. 4. Lam, D. M.; Hollyfield, J. G., Localization of Putative Amino Acid Neurotransmitters in the Human Retina. Exp Eye Res 1980, 31 (6), 729-732. 5. Watt, C. B.; Li, H. B.; Lam, D. M., The Presence of Three Neuroactive Peptides in Putative Glycinergic Amacrine Cells of an Avian Retina. Brain Res 1985, 348 (1), 187-191. 6. Hokfelt, T.; Millhorn, D.; Seroogy, K.; Tsuruo, Y.; Ceccatelli, S.; Lindh, B.; Meister, B.; Melander, T.; Schalling, M.; Bartfai, T.; et al., Coexistence of Peptides with Classical Neurotransmitters. Experientia 1987, 43 (7), 768-780. 7. Ottersen, O. P., Postembedding Immunogold Labelling of Fixed Glutamate: An Electron Microscopic Analysis of the Relationship Between Gold Particle Density and Antigen Concentration. J Chem Neuroanat 1989, 2 (1), 57-66. 8. Kalloniatis, M.; Fletcher, E. L., Immunocytochemical Localization of the Amino Acid Neurotransmitters in the Chicken Retina. J Comp Neurol 1993, 336 (2), 174-193. 9. Marc, R. E.; Murry, R. F.; Basinger, S. F., Pattern Recognition of Amino Acid Signatures in Retinal Neurons. J Neurosci 1995, 15 (7 Pt 2), 5106-5129. 10. Marc, R. E.; Jones, B. W., Molecular Phenotyping of Retinal Ganglion Cells. J Neurosci 2002, 22 (2), 413-427. 11. Kiernan, J. A., Formaldehyde, Formalin, Paraformaldehyde and Glutaraldehyde: What They Are and What They Do. Microsc Today 2000, 08 (01), 8-12. 12. Carey, F. A.; Giuliano, R. M., Organic Chemistry. Tenth ed.; McGraw Hill Education: New York, NY, 2017; Vol. 10. 13. Peters, T., Jr.; Ashley, C. A., An Artefact in Radioautography due to Binding of Free Amino Acids to Tissues by Fixatives. J Cell Biol 1967, 33 (1), 53-60. 14. Landsteiner, K., The Specificity of Serological Reactions. Dover Publications: New York, NY, 1936. 49 15. Marc, R. E.; Wei-Ley, S.; Kalloniatis, M.; Raiguel, S. F.; Van Haesendonck, E., Patterns of Glutamate Immunoreactivity in the Goldfish Retina. J Neurosci 1990, 10 (12), 4006-4034. 16. Evans, N. R.; Webb, H. E., Comparison of Protein A-Gold and Ferritin Immunoelectron Microscopy of Semliki Forest Virus in Mouse Brain Using a Rapid Processing Technique. J Histochem Cytochem 1984, 32 (4), 372-378. 17. Daneels, G.; Moeremans, M.; De Raeymaeker, M.; De Mey, J., Sequential Immunostaining (Gold/Silver) and Complete Protein Staining (AuroDye) on Western Blots. J Immunol Methods 1986, 89 (1), 89-91. 18. James, T., The Reduction of Silver Ions by Hydroquinone. Am Chem Soc 1939, 61 (3), 648-652. 19. Marc, R. E.; Jones, B. W., Retinal Remodeling in Inherited Photoreceptor Degenerations. Mol Neurobiol 2003, 28 (2), 139-147. 20. Marc, R. E.; Cameron, D., A Molecular Phenotype Atlas of the Zebrafish Retina. J Neurocytol 2001, 30 (7), 593-654. 21. Kalloniatis, M.; Marc, R. E.; Murry, R. F., Amino Acid Signatures in the Primate Retina. J Neurosci 1996, 16 (21), 6807-6829. 22. Bookstein, F. L., Principal Warps - Thin-Plate Splines and the Decomposition of Deformations. IEEE Trans Pattern Anal and Mach Intell 1989, 11 (6), 567-585. 23. Jones, B. W.; Kondo, M.; Terasaki, H.; Watt, C. B.; Rapp, K.; Anderson, J.; Lin, Y.; Shaw, M. V.; Yang, J. H.; Marc, R. E., Retinal Remodeling in the Tg P347L Rabbit, a Large-Eye Model of Retinal Degeneration. J Comp Neurol 2011, 519 (14), 2713-27133. 24. Jones, B. W.; Pfeiffer, R. L.; Ferrell, W. D.; Watt, C. B.; Marmor, M.; Marc, R. E., Retinal Remodeling in Human Retinitis Pigmentosa. Exp Eye Res 2016, 150, 149-165. 25. Jones, B. W.; Pfeiffer, R. L.; Ferrell, W. D.; Watt, C. B.; Tucker, J. F.; Marc, R. E., Retinal Remodeling And Metabolic Alterations in Human AMD. Front Cell Neurosci 2016, 10, 103. 26. Marc, R. E., Mapping Glutamatergic Drive in the Vertebrate Retina with a Channel-Permeant Organic Cation. J Comp Neurol 1999, 407 (1), 47-64. 27. Meyer, D.; Leisch, F.; Hornik, K., The Support Vector Machine Under Test. Neurocomputing 2003, 55 (1-2), 169-186. 28. Kondo, M.; Sakai, T.; Komeima, K.; Kurimoto, Y.; Ueno, S.; Nishizawa, Y.; Usukura, J.; Fujikado, T.; Tano, Y.; Terasaki, H., Generation of a Transgenic Rabbit Model of Retinal Degeneration. Invest Ophthalmol Vis Sci 2009, 50 (3), 1371-1377. 50 29. Griciuc, A.; Aron, L.; Ueffing, M., Looking Into Eyes: Rhodopsin Pathologies in Drosophila. Adv Exp Med Biol 2012, 723, 415-423. 30. Chow, C. Y.; Kelsey, K. J.; Wolfner, M. F.; Clark, A. G., Candidate Genetic Modifiers of Retinitis Pigmentosa Identified by Exploiting Natural Variation in Drosophila. Hum Mol Genet 2015, 25 (4), 651-659. 31. Brockerhoff, S. E.; Hurley, J. B.; Janssen-Bienhold, U.; Neuhauss, S. C.; Driever, W.; Dowling, J. E., A Behavioral Screen for Isolating Zebrafish Mutants with Visual System Defects. Proc Natl Acad Sci U S A 1995, 92 (23), 10545-10549. 32. Keeler, C. E., The Inheritance of a Retinal Abnormality in White Mice. Proc Natl Acad Sci U S A 1924, 10 (7), 329-333. 33. Pittler, S. J.; Keeler, C. E.; Sidman, R. L.; Baehr, W., PCR Analysis of DNA from 70-Year-Old sections of Rodless Retina Demonstrates Identity with the Mouse rd Defect. Proc Natl Acad Sci U S A 1993, 90 (20), 9616-9619. 34. Chang, B.; Hawes, N. L.; Hurd, R. E.; Davisson, M. T.; Nusinowitz, S.; Heckenlively, J. R., Retinal Degeneration Mutants in the Mouse. Vision Res 2002, 42 (4), 517-525. 35. Han, J.; Dinculescu, A.; Dai, X.; Du, W.; Smith, W. C.; Pang, J., Review: The History and Role of Naturally Occurring Mouse Models with Pde6b Mutations. Mol Vis 2013, 19, 2579-2589. 36. Bourne, M. C.; Campbell, D. A.; Tansley, K., Hereditary Degeneration of the Rat Retina. Br J Ophthalmol 1938, 22 (10), 613-623. 37. Semple-Rowland, S. L.; Lee, N. R., Avian Models of Inherited Retinal Disease. Methods Enzymol 2000, 316, 526-536. 38. Narfstrom, K., Hereditary Progressive Retinal Atrophy in the Abyssinian Cat. J Hered 1983, 74 (4), 273-276. 39. Barnett, K. C.; Curtis, R., Autosomal Dominant Progressive Retinal Atrophy in Abyssinian Cats. J Hered 1985, 76 (3), 168-170. 40. Aguirre, G., Retinal Degenerations in the Dog. I. Rod Dysplasia. Exp Eye Res 1978, 26 (3), 233-253. 41. Suber, M. L.; Pittler, S. J.; Qin, N.; Wright, G. C.; Holcombe, V.; Lee, R. H.; Craft, C. M.; Lolley, R. N.; Baehr, W.; Hurwitz, R. L., Irish Setter Dogs Affected with Rod/Cone Dysplasia Contain a Nonsense Mutation in the Rod cGMP Phosphodiesterase Beta-Subunit Gene. Proc Natl Acad Sci U S A 1993, 90 (9), 3968-3972. 42. Petters, R. M.; Alexander, C. A.; Wells, K. D.; Collins, E. B.; Sommer, J. R.; Blanton, M. R.; Rojas, G.; Hao, Y.; Flowers, W. L.; Banin, E.; Cideciyan, A. V.; 51 Jacobson, S. G.; Wong, F., Genetically Engineered Large Animal Model for Studying Cone Photoreceptor Survival and Degeneration in Retinitis Pigmentosa. Nat Biotechnol 1997, 15 (10), 965-970. 43. Silver, D.; Csutak, A., Human Eye Dimensions for Pressure-Volume Relations. Invest Ophthalmol Vis Sci 2010, 51 (13), 5019. 44. Park, J.; Bungay, P. M.; Lutz, R. J.; Augsburger, J. J.; Millard, R. W.; Sinha Roy, A.; Banerjee, R. K., Evaluation of Coupled Convective-Diffusive Transport of Drugs Administered by Intravitreal Injection and Controlled Release Implant. J Control Release 2005, 105 (3), 279-295. CHAPTER 3 MÜLLER CELL METABOLIC CHAOS DURING RETINAL DEGENERATION Reprinted with Permission from Elsevier Ltd Pfeiffer, R.L.; Marc, R.E.; Kondo, M.; Terasaki, H.; Jones, B.W., Müller Cell Metabolic Chaos During Retinal Degeneration. Exp Eye Res 2016, 150, 62-70. 53 54 55 56 57 58 59 60 61 CHAPTER 4 EVALUATION OF EXCITOTOXICITY AS A MECHANISM OF RETINAL DEGENERATION 4.1 Introduction Müller cells are well known to play important roles in structural and metabolic processes of the retina. They provide central metabolic support to the surrounding cells, particularly in the glutamate cycle.1-2 Müller cells are the only cells of the retina which possess glutamine synthetase.2 In the primary glutamate cycle, glutamate is released from excitatory neurons and perisynaptic Müller cell processes transport the extra cellular glutamate into Müller cells, primarily via EAAT1 (excitatory amino acid transporter 1), also referred to as GLAST (glutamate/aspartate transporter).3 Upon entering the Müller cell, glutamine synthetase rapidly converts glutamate to glutamine by amidation. Following the conversion to glutamine within the Müller cell, glutamine is exported into the extracellular space by glial SN1.4-5 In all neurons, but especially in GABAergic and glutamatergic neurons, glutamine can be converted back to glutamate by glutaminase (EC 3.5.1.2). Glutamate can now be loaded into vesicles via vesicular glutamate transporters (VGLUT) and released upon depolarization or converted to GABA by glutamate decarboxylase (EC 4.1.1.15) and loaded into vesicles via VGAT (which transports both GABA and glycine). The role of Müller glia is integral to both the extracellular glutamate 63 clearance and the synthesis of glutamate and GABA within the retina. Given this central role of Müller cells and glutamine synthetase in retinal glutamate metabolism, it is reasonable to expect that any disruption of this system could be detrimental to the maintenance of glutamate homeostasis in the retina. However, the role of Müller glia during retinal degeneration and whether the changes observed in the glutamate cycle are a cause or result of the degeneration are unclear and current topics of investigation. One popular hypothesis is the possibility of glutamate excitotoxicity in the retina contributing to neuronal degeneration of distressed retina.6-8 4.1.1 Excitotoxicity Hypothesis One proposed mechanism of retinal degeneration is excitotoxicity-mediated degeneration. This concept was originally introduced by Lucas and Newhouse in 1957, where they demonstrated high levels of glutamate can be toxic to neural systems.9 Excitotoxicity as a mechanism of not only brain injury, but as a factor contributing to neural death during retinal degeneration, has been supported by other studies.6-7, 10-11 A study that proposed a possible mechanism of excitotoxic injury was by Szatkowski, Barbour, and Attwell in 1990.12 They demonstrated that under conditions of high extracellular potassium, and high intracellular sodium and glutamate, a nonvesicular mechanism of glutamate release could be evoked from retinal Müller cells, indicating that Müller cells could potentially lose their capacity to adequately sequester glutamate away from neurons, which would be a likely component contributing to the development of an excitotoxic environment. Another study investigating the potential role of excitotoxicity in retinal 64 degeneration was conducted on the rd1 mouse.13 The rd1 mouse is a model of retinitis pigmentosa in which the mouse has early onset retinal degeneration, beginning with the loss (actually a failure to completely develop followed by cell death) of rod photoreceptors.14 Delyfer et al. used a chemiluminescence detection system to analyze the ratio of given amino acids to the total amino acid concentration.13 They found that the levels of glutamate, glutamine, and glycine were all significantly higher in the rd1 mouse than in the control. These data indicate a deficiency in the metabolism of core amino acids as degeneration progresses. Of importance to the question of glutamate-meditated excitotoxicity: Could the significant increase in the levels of glutamate within the degenerate retina could be contributing to the phase 2 photoreceptor degeneration or phase 3 neuronal seen in the rd1 mouse? In addition to the aforementioned papers, which are primarily focused on the Müller glia's ability to properly sequester glutamate, much of the excitotoxicity argument is based on ionotropic glutamate receptor associated toxicity. Ionotropic glutamate receptors (iGluRs) are postsynaptic ligand-gated ion channels, divided into three classes: N-methylD-aspartate (NMDA), α-amino-3-hydroxy-5-methyl-4-isoxazolepionic acid (AMPA), and kainate (KA) receptors. AMPA and NMDA receptors are both calcium permeable when bound by glutamate (KA receptors are less Ca-permeant), and excessive calcium following sustained glutamate activation is thought to be the primary mechanism causing neural death in excitotoxicity.15 AMPA and KA receptors are often lumped together as ionotropic nonNMDA receptors, as NMDA is an agonist at neither while KA and AMPA are agonists both, as is quisqualate. The effects of persistent AMPAR activation have been evaluated via KA administration. KA is a selective ionotropic glutamate receptor agonist, which 65 binds to KA receptors with high affinity and activates currents with fast desensitization, and binds with lower affinity to more prevalent receptors, but does not desensitize, producing a sustained cation current. Previous studies demonstrate that when KA is applied to the retina it greatly outcompetes glutamate for the AMPARs available on a large number of cell types within the retina, including but not limited to: starburst amacrine cells, AII amacrine cells, bipolar cells, and ganglion cells; sustained AMPAR sensitization can trigger apoptotic mechanisms.16 In addition to the studies into AMPARs potential role in mediating excitotoxicity, NMDARs have been implicated in contributing to neural death.17 The sustained sensitization of NMDARs or AMPARs can lead to a prolonged increase in intracellular calcium within the cell, leading to cell death. 4.1.2 Evidence Disputing Excitotoxicity The role of glutamine synthetase in the glutamate cycle of the neural retina was clearly shown by Pow and Robinson18 in which they used the suicide inhibitor D,Lmethionine D,L-sulphoximine (MSO) to inhibit glutamine synthetase (GS) activity in a normal rabbit retina. This demonstrated that following the inhibition of glutamine synthetase, bipolar and ganglion cells of the neural retina no longer exhibit any glutamate immunoreactivity (although the Marc lab has somewhat conflicting unpublished results). In contrast, Müller cells are extremely immunoreactive for glutamate as they no longer possess active GS to convert glutamate to glutamine. Of special interest to the question of excitotoxicity, they also tested the effect of increasing the extracellular levels of glutamate 5 fold (35µM), and this also failed to elicit an increase in glutamate levels in bipolar or ganglion cells. 66 Marc et al. examined the role of the glutamate cycle in injury models of retinal degeneration using amino acid signatures of detached cat retina.19 Following detachment, Müller cells quickly lose their ability to convert glutamate to glutamine as seen by a decrease in Müller cell levels of glutamine and increases in glutamate. Consistent with the results of Pow and Robinson, this loss of apparent glutamine synthetase activity leads to a large increase in levels of intracellular glutamate in Müller cells (~5mM). Although glutamate concentration does later recover in neurons between days 6-30 of detachment, it is still substantially less glutamate than is found in neurons of the healthy retina and never regains normal levels. In addition to the alterations in glutamate-glutamine cycle in an injury model of retinal degeneration and remodeling, many immunocytochemical studies of metabolic changes have been performed on a variety of gene defect models of retinal degeneration.2022 In all of these studies there is a stereotypic pattern to the degeneration of the retina and the responses of the Müller cells, which are extremely similar to those seen in the injury induced model. Pfeiffer et al., showed a rise in the glutamine and glutamine synthetase levels in Müller cells.23 This rise in glutamine synthetase and glutamine subsequently dissipated as the retina continued to degenerate and remodel.24 Glutamate levels have been reported to either increase20 in Müller cells, or decrease.22 This inconsistency between studies is probably the result of variation in time points in addition to the large variability in amino acid signatures that can be observed between individual Müller cells in degenerated retina.20, 23, 25 The overall trend in the degenerate retina is a decline in neuronal glutamate and glutamine as the retina degenerates, however, it appears that following the initial decline there is some recovery, but it never again reaches normal levels. This 67 variability between Müller cells of glutamine, glutamate, and glutamine synthetase and its metabolic impact on surrounding retina warrants further investigation, however, none of these studies indicate excitotoxicity as a component of the degeneration. In this paper we examine the response of Müller cells to retinal degeneration over time, and whether the Müller cell metabolism alterations found in RP are likely to contribute to an excitotoxic mechanism of degeneration. Primarily identifying how the levels of glutamate and glutamine respond to the loss of glutamine synthetase during phase 3 remodeling and degeneration, and how glutamate transport capabilities of Müller cells are affected. 4.2 Materials and Methods 4.2.1 Model Systems The transgenic rabbit model of autosomal dominant retinitis pigmentosa (Tg P347L), described in Chapters 2 and 3,24 was used for the assessment of metabolic variability using quantitative glutamine synthetase immunocytochemistry, quantitative CMP for glutamine, glutamate comparison analysis, and quantitative D-Aspartate loading. Six Tg rabbits spanning in ages from 1.5 years to 6+ years were used, with 3 wild-type (wt) littermates spanning 1.5 to 8 years used as controls. All animal experiments were conducted according to the ARVO Statement for the use of animals in ophthalmic and vision research with the supervision and approval of the Institutional Animal Care and Use Committee (IACUC) at the University of Utah. In addition, two human specimens, aged 21 and 77, diagnosed with RP were collected within 5 hours post mortem. RP tissues were obtained from two sources: The Foundation Fighting Blindness Retina Donor Program at Stanford 68 University and the University of Utah Lions Eye Bank. Genotyping of donors was not available and the mutations involved are unknown. Institutional approval for use of human eyes was obtained from the University of Utah and Stanford University and followed the tenets of the Declaration of Helsinki. All retinal tissues and data were de-identified in accordance with HIPPA Privacy Rules. These specimens were processed for CMP and used for analysis of variability. 4.2.2 Tissue Processing Rabbit and human tissues were processed identically to those in Chapter 3, utilizing mixed aldehyde fixation for small molecules and proteins. In addition to the immediate fixation of many retinal samples, some retinas were processed for D-Aspartate (D-Asp) transport capabilities. In these cases, retinal chips ≈ 10-30 mm2 were transferred to cellulose acetate filter discs and bathed in 35°C Ames medium26 equilibrated with 95% O2/5% CO2 for 2-5 min. Following equilibration, some chips were then transferred to Ames media containing D-Aspartate (D-Asp) at 5 mM (a saturating level for EAATs), to evaluate the transport functions of Müller cells in the intact retina27 prior to fixation. All other chips were immediately transferred to mixed aldehyde fixation buffers. 4.2.3 CMP Analysis CMP analysis of retinas was conducted as described in Chapter 3, which yields quantitative immunohistochemistry allowing us to directly compare levels of metabolites between samples. The previous chapters have described using k-means and isodata unsupervised classification to track the variability of Müller cells in representative fields. 69 In this study we sought to better understand the extent of the metabolic variability in degenerate retina. To fully characterize the metabolic changes occurring in Müller cells during degeneration, we argue it is necessary to quantify the number of Müller cell metabolic states in a selected sample. To perform this type clustering, a fusion version of unsupervised and supervised classification methods called intraclass separability analysis was used. First, the clusters of distinct Müller cell metabolic states generated from ISODATA classification were used to gain an idea of the total number of metabolic subclasses within the Müller cell class. Using this map, supervised classification using PCI Geomatica (Richmond Hill, Ontario, Canada) was performed as follows: Between 20-25 seed points were set in the training component of supervised classification using raster seeding in which 6 input theme map channels were used (Taurine (τ), Glutamine Synthetase (GS), Glutamine (Q), Glutamate (E), Glutathione (J), and retinaldehydebinding protein 1 (CRALBP)). The pixel tolerance was set at 12 with an 8-connected neighborhood, allowing enough brightness variability to account for pixelation. From these seed points a Bhattacharyya Distance statistic was run on all seed points and any points with a separability of >1.9 were merged into a single metabolic class. A Bhattacharyya Distance of greater than 1.9 is equivalent to a p-value of <0.01. From this training channel, a maximum likelihood with a NULL class classification was run on the entire sample generating a theme map of the Müller cells and their separable metabolic variability. 4.3 Results Previous studies have shown Müller cells diverge from one another in an increasingly chaotic fashion,20, 24-25 however, the impact of this chaotic divergence and its 70 impact on the glutamate cycle is currently unknown. First, we demonstrate that the chaotic divergence of Müller cell metabolic signatures of rabbit retina is also present in human retinal degenerations (Figure 4.1). Levels of taurine, glutamine, and glutamate (red, green, and blue respectively), vary dramatically between Müller cells, including neighboring Müller cells as seen indicated by the red circle (Figure 4.1A). This variability continues throughout degeneration in rabbits (Figure 4.1B). It is also observed relatively early in humans (Figure 4.1C, 21 year old affected by recessive RP), and continues throughout RP progression (Figure 4.1D, 77 year old affected by unknown but likely dominant RP). This demonstrates that the variability found in rabbits is not unique. The observed variety of metabolic signatures in Müller cells led us to further classify the variability between Müller cell signatures within small regions of degenerate retina. Using intraclass separability analysis (see Section 4.2.3), Müller cells within a representative vertical region of intact retina classify into 6 metabolic states (Figure 4.2). This metabolic state separability is further distinguishable by Bhattacharyya distance (>1.9). The observed subclass separability is never found across the superclass of Müller cells in a normal retina, where only one class is known. It is of interest to note that these variable states do not occur as a continuum, which is expected if microenvironmental factors cause the variation. Rather all of the observed Müller cell subclasses appear independently of the metabolic states of their neighbors. As noted in previous descriptions of the metabolic alterations occurring during degeneration.23-25 GS levels are drastically lower throughout the degenerate retina (albeit to varying extents). This paradoxically occurs simultaneously with the observed global increase in glutamine and stable global concentration of Müller cell glutamate.24 This is in 71 stark contrast to the classic model that glutamate and glutamine levels in the retina are stoichiometrically related to the available GS activity. To determine whether this fundamental relationship is maintained despite the degeneration induced heterogeneity observed in Müller cells, we sampled n =5 Müller cells across both healthy and degenerate retinas and evaluated the relative amounts of glutamate, glutamine, and GS. In the healthy retinas, the levels of all three (and thus their ratios) remain constant. In the degenerate retinas, all three metabolites vary widely and independently of one another, conflicting with the standard model of glutamate-glutamine metabolism (Figure 4.3). It has been previously hypothesized that the decrease in GS is related to a loss in EAAT1 (the main E transporter found on Müller cells) found by the Reichenbach laboratory.28-29 To test this hypothesis we performed D-Aspartate loading both WT and TgP347L rabbit retinas. Until extremely late stages of degeneration, D-Asp loading, presumably by EAAT1, was fully functional in the degenerate retinas and capable of transporting any extracellular glutamate into Müller cells (Figure 4.4). In aged and neurodegenerate rabbit retinas (Figure 4.4D), some Müller cells appear to have lost glutamate transport capabilities (white arrows). We cannot yet explain the mechanisms allowing for the continued metabolism of glutamate and the increase in Müller cell glutamine, though it is probable that Müller cells with depressed GS activity unmask less commonly observed metabolic pathways. One potential unmasked mechanism is amidation of glutamate into glutamine through transaminase activity. 72 4.4 Conclusions and Discussion Three main conclusions may be drawn from this study. First, over the course of retinal degeneration, the chaotic phenotypes that appear in Müller cells not only vary with respect to one another, but diverge into multiple statistically separable metabolic classes. The impact of these states on the surrounding neurons warrants further study. Second, although GS is variably reduced in Müller cells across the retina in degeneration, its level appears to have minor impact on the levels of glutamate or glutamine within the same Müller cells. This implies the existence of a network-based metabolic homeostasis. Third, we see no evidence that Müller cell glutamate transport is drastically altered in response to the metabolic variability observed throughout degeneration, until the degeneration has advanced to a point in which it is unlikely there is a very high quantity of glutamate being released to need sequestration. Previous suggestions that Müller cell glutamate transport fails early in retinal degenerations6-7 are unsupported by our data. Importantly, our analyses sample hundreds to thousands of adjacent Müller cells and there is no variation in glutamate transport until very late degeneration. Thus the conflicting results are not likely due to sampling differences. These observations have several implications for retinal degeneration, particularly that arising from retinitis pigmentosa. As levels of GS have no correlated impact in glutamate or glutamine, concentrations within the Müller cells indicates GS is likely not the primary enzyme responsible for glutamate metabolism in the degenerated retina. Based on previous studies in the brain, this is not particularly surprising. This also conflicts with the acute studies of Pow,18 but it is important to recall that retinal degenerations are chronic steady states. Glutamate is a primary carbon skeleton that may be generated from multiple 73 enzymatic pathways and contributes to multiple other pathways. We argue that the glutamate-glutamine cycle should be considered less of a stoichiometric biochemical pathway and more of an intersection with many other metabolic pathways.30 The amount glutamate contributes to one pathway over another is likely to change depending on the metabolic demands of a cell, and in the case of glia, the requirements of the surrounding tissue. There is the caveat of the highly variable metabolic signatures found between Müller cells in close proximity to one another, which indicates a lack of microenvironmental influence over Müller cell metabolism. We do not know what drives the Müller cells to adopt such variability in their metabolic signatures, however to this end, we don't know what causes the metabolism of Müller cells to be so homogeneous in the first place. This question will require further investigation of both healthy retinas and degenerate retinas. A primary question this study sought to address is whether glutamate excitotoxicity is a likely mechanism contributing to the neurodegeneration observed in the inner retina following the loss of photoreceptors. In a more prosaic sense we ask whether the deaths of photoreceptors and any associated bursts of extracellular glutamate cells pose a challenge to cells that express iGluRs, that is, horizontal, OFF cone bipolar, amacrine and ganglion cells. Based on quantitative immunocytochemical methods of determining levels of glutamate within the intact retina, combined with the D-Asp transport results reported in this chapter, we do not find evidence that levels of extracellular glutamate are likely to rise for a sustained period of time. Although the traditional view of excitotoxicity has been largely regarded as a component of retinal degeneration, more recent studies,22 combined with ours, have largely called this view into question. For excitotoxicity to cause cell stress and death, sustained activity by high levels of glutamate are required, as well as expression 74 of iGluRs. Our results indicate that Müller cells continue to transport glutamate with no apparent change in Vmax long after neurodegeneration has greatly reduced the neural population of the retina. Although it is possible that there are local spikes in glutamate due to photoreceptor death, any downstream neuronal death would have to occur in minutes (rather than the reported days in the fastest models31) to potentially overwhelm functional Müller cell transport.32 Therefore we conclude glutamate-mediated excitotoxicity is not likely to significantly contribute to the neuronal death observed in retinal degenerative diseases. Figure 4.1 Müller cell metabolic variation across retinal regions and species. All are RGB registered projections of adjacent images of amino acid immunoreactivity; red = T, green = Q, and blue = E. Color variance in an individual image indicates varying ratios of TQE, and variance in concentration. A) Vertical section through central portion of rabbit retina. Arrows indicate select Müller cells demonstrating high variability. Ellipse highlights 5 Müller cells directly adjacent to one another all with varying metabolic profiles in comparison to their direct neighbor. B) Vertical section through peripheral region of TgP347L degenerate rabbit retina that has undergone massive remodeling and cell death. Arrows indicate Müller cells with varying metabolic profiles. C) Vertical section from a deceased 21yo male affected by early onset retinitis pigmentosa. Arrows indicate Müller cells, with varying levels of metabolites. D) Horizontal section of retina from 77 yo male affected by retinitis pigmentosa. Black arrows indicate blood vessels. White arrows indicate select Müller cells with drastically varying levels of metabolites 75 Figure 4.2: Müller cell metabolic variation within Müller cells. Base image is RGB registered projections of amino acid immunoreactivity with red = T, green = Q, and blue = E. Overlaid colors outlined in white are a representative mask based on the 6 Müller cell metabolic states a vertical section of Tg P347L retina. The mask was formed as described in the Methods section by intraclass separability analysis. Masks of similar color indicate metabolic states that are not highly separable from one another. Masks of differing color indicate that the metabolic states of the underlying Müller cell are significantly separable from one another via Bhattacharyya distance analysis 76 77 Figure 4.3 GS/Q/E variability in WT compared to degenerate retina. Images are RGB registered images of molecular immunoreactivity with red = GS, green = Q, and blue = E. Images are inverted to display intensity, similar to that of fluorescence. Colors in histogram plots are indicative of the same channel as in the RGB image. A) WT retina with 5 selected Müller cell regions. Pixel values for the sampled regions vary slightly, but the location of the individual peaks remains relatively constant indicating consistent concentrations of GS, Q, and E. B) Degenerate retina with 5 selected Müller cell regions. GS (in red) varies only slightly because the quantity of GS in this region of degenerate retina is extremely low. Despite the overall low concentrations of GS, the levels of E (blue) and Q (green) vary dramatically. Both may be higher or lower than is found in WT Müller cells, but variance does not appear to correlate with each other in any consistent way. 78 Figure 4.4 D-Aspartate loading in retinal chips. Amino acid immunoreactivity from 100nm sections of retina loaded in vitro by incubating immediately post-mortem with 5mM D-Asp and subsequently fixing with 2.5% glutaraldehyde. Images are density scaled, meaning darker areas are indicative of greater D-Asp immunoreactivity and thereby greater D-Asp transport prior to fixation. White arrows indicate Müller cells with deficient D-Asp loading, indicating deficient glutamate transport. 79 4.5 References 1. Reichenbach, A.; Bringmann, A., Muller Cells in the Healthy Retina. In Muller Cells in the Healthy and Diseased Retina, Springer Science+Business Meddia, LLC: New York, 2010; pp 35-299. 2. Riepe, R. E.; Norenburg, M. D., Muller Cell Localisation of Glutamine Synthetase in Rat Retina. Nature 1977, 268 (5621), 654-655. 3. Rauen, T.; Rothstein, J. D.; Wassle, H., Differential Expression of Three Glutamate Transporter Subtypes in the Rat Retina. Cell Tissue Res 1996, 286 (3), 325336. 4. Chaudhry, F. A.; Reimer, R. J.; Krizaj, D.; Barber, D.; Storm-Mathisen, J.; Copenhagen, D. R.; Edwards, R. H., Molecular Analysis of System N Suggests Novel Physiological Roles in Nitrogen Metabolism and Synaptic Transmission. Cell 1999, 99 (7), 769-780. 5. Umapathy, N. S.; Li, W.; Mysona, B. A.; Smith, S. B.; Ganapathy, V., Expression and Function of Glutamine Transporters SN1 (SNAT3) and SN2 (SNAT5) in Retinal Muller Cells. Invest Ophthalmol Vis Sci 2005, 46 (11), 3980-3907. 6. Bringmann, A.; Grosche, A.; Pannicke, T.; Reichenbach, A., GABA and Glutamate Uptake and Metabolism in Retinal Glial (Muller) Cells. Front Endocrinol 2013, 4, 1-14. 7. Reichenbach, A.; Bringmann, A., New Functions of Muller Cells. Glia 2013, 61 (5), 651-678. 8. Izumi, Y.; Kirby, C. O.; Benz, A. M.; Olney, J. W.; Zorumski, C. F., Muller Cell Swelling, Glutamate Uptake, and Excitotoxic Neurodegeneration in the Isolated Rat Retina. Glia 1999, 25 (4), 379-389. 9. Lucas, D. R.; Newhouse, J. P., The Toxic Effect of Sodium L-Glutamate on the Inner Layers of the Retina. AMA Arch Ophthalmol 1957, 58 (2), 193-201. 10. Olney, J. W., Glutamate-Induced Retinal Degeneration in Neonatal Mice. Electron Microscopy of the Acutely Evolving Lesion. J Neuropathol Exp Neurol 1969, 28 (3), 455-474. 11. Olney, J. W., The Toxic Effects of Glutamate and Related Compounds in the Retina and the Brain. Retina 1982, 2 (4), 341-359. 12. Szatkowski, M.; Barbour, B.; Attwell, D., Non-Vesicular Release of Glutamate from Glial Cells by Reversed Electrogenic Glutamate Uptake. Nature 1990, 348 (6300), 443-446. 80 13. Delyfer, M. N.; Forster, V.; Neveux, N.; Picaud, S.; Leveillard, T.; Sahel, J. A., Evidence for Glutamate-Mediated Excitotoxic Mechanisms During Photoreceptor Degeneration in the rd1 Mouse Retina. Mol Vis 2005, 11, 688-696. 14. Chang, B.; Hawes, N. L.; Hurd, R. E.; Davisson, M. T.; Nusinowitz, S.; Heckenlively, J. R., Retinal Degeneration Mutants in the Mouse. Vision Res 2002, 42 (4), 517-525. 15. Choi, D. W., Glutamate Neurotoxicity in Cortical Cell Culture is Calcium Dependent. Neurosci Lett 1985, 58 (3), 293-297. 16. Park, Y. H.; Mueller, B. H., 2nd; McGrady, N. R.; Ma, H. Y.; Yorio, T., AMPA Receptor Desensitization is the Determinant of AMPA Receptor Mediated Excitotoxicity in Purified Retinal Ganglion Cells. Exp Eye Res 2015, 132, 136-150. 17. Choi, D. W.; Koh, J. Y.; Peters, S., Pharmacology of Glutamate Neurotoxicity in Cortical Cell Culture: Attenuation by NMDA Antagonists. J Neurosci 1988, 8 (1), 185196. 18. Pow, D. V.; Robinson, S. R., Glutamate in Some Retinal Neurons is Derived Solely from Glia. Neuroscience 1994, 60 (2), 355-366. 19. Marc, R. E.; Murry, R. F.; Fisher, S. K.; Linberg, K. A.; Lewis, G. P., Amino Acid Signatures in the Detached Cat Retina. Invest Ophthalmol Vis Sci 1998, 39 (9), 1694-1702. 20. Jones, B. W.; Kondo, M.; Terasaki, H.; Watt, C. B.; Rapp, K.; Anderson, J.; Lin, Y.; Shaw, M. V.; Yang, J. H.; Marc, R. E., Retinal Remodeling in the Tg P347L Rabbit, a Large-Eye Model of Retinal Degeneration. J Comp Neurol 2011, 519 (14), 2713-27133. 21. Ross, J. W.; Fernandez de Castro, J. P.; Zhao, J.; Samuel, M.; Walters, E.; Rios, C.; Bray-Ward, P.; Jones, B. W.; Marc, R. E.; Wang, W.; Zhou, L.; Noel, J. M.; McCall, M. A.; DeMarco, P. J.; Prather, R. S.; Kaplan, H. J., Generation of an Inbred Miniature Pig Model of Retinitis Pigmentosa. Invest Ophthalmol Vis Sci 2012, 53 (1), 501-507. 22. Gibson, R.; Fletcher, E. L.; Vingrys, A. J.; Zhu, Y.; Vessey, K. A.; Kalloniatis, M., Functional and Neurochemical Development in the Normal and Degenerating Mouse Retina. J Comp Neurol 2013, 521 (6), 1251-1267. 23. Pfeiffer, R. L.; Marc, R. E.; Kondo, M.; Terasaki, H.; Jones, B. W., Muller Cell Metabolic Chaos During Retinal Degeneration. Exp Eye Res 2016, 150, 62-70. 24. Pfeiffer, R. L.; Marc, R. E.; Jones, B. W., Muller cell metabolic chaos during retinal degeneration. Exp Eye Res 2016. 25. Jones, B. W.; Pfeiffer, R. L.; Ferrell, W. D.; Watt, C. B.; Marmor, M.; Marc, R. E., Retinal Remodeling in Human Retinitis Pigmentosa. Exp Eye Res 2016, 150, 149-165. 81 26. Ames, A., 3rd; Nesbett, F. B., In Vitro Retina as an Experimental Model of the Central Nervous System. J Neurochem 1981, 37 (4), 867-877. 27. Marc, R. E.; Murry, R. F.; Basinger, S. F., Pattern Recognition of Amino Acid Signatures in Retinal Neurons. J Neurosci 1995, 15 (7 Pt 2), 5106-5129. 28. Ishikawa, M.; Yoshitomi, T.; Zorumski, C. F.; Izumi, Y., Downregulation of Glutamine Synthetase via GLAST Suppression Induces Retinal Axonal Swelling in a Rat Ex Vivo Hydrostatic Pressure Model. Invest Ophthalmol Vis Sci 2011, 52 (9), 6604-6616. 29. Rauen, T.; Wiessner, M., Fine Tuning of Glutamate Uptake and Degradation in Glial Cells: Common Transcriptional Regulation of GLAST1 and GS. Neurochem Int 2000, 37 (2-3), 179-189. 30. McKenna, M. C., The Glutamate-Glutamine Cycle is not Stoichiometric: Fates of Glutamate in Brain. J Neurosci Res 2007, 85 (15), 3347-3358. 31. Sahaboglu, A.; Paquet-Durand, O.; Dietter, J.; Dengler, K.; Bernhard-Kurz, S.; Ekstrom, P. A.; Hitzmann, B.; Ueffing, M.; Paquet-Durand, F., Retinitis Pigmentosa: Rapid Neurodegeneration is Governed by Slow Cell Death Mechanisms. Cell Death Dis 2013, 4, e488. 32. Sarthy, V. P.; Pignataro, L.; Pannicke, T.; Weick, M.; Reichenbach, A.; Harada, T.; Tanaka, K.; Marc, R., Glutamate Transport by Retinal Muller Cells in Glutamate/Aspartate Transporter-Knockout Mice. Glia 2005, 49 (2), 184-196. CHAPTER 5 LONG-TERM RETINAL DEGENERATION LEADS TO EXTENSIVE NEURODEGENERATION AND ALPHA SYNUCLEIN AGGREGATION 5.1 Effects of Long-Term Remodeling on Therapeutic Interventions Retinal degenerations encompass many disorders that manifest with the death of photoreceptors leading to a loss of vision due to lack of sensory retinal input. Though some retinal degenerations are age-related, others may be early onset. In either case the affected individual is likely to suffer severe vision impairment for years or decades. Vision rescue of this population is an audacious goal, which many groups are pursuing through a variety of mechanisms. Cell replacement, genetic alteration, bionic implants, optogenetics, and stem cell therapy have all been proposed as possible mechanisms to restore vision following the loss of photoreceptors.1-6 However, the long-term impact and potential of each of these interventions cannot be adequately predicted without a full understanding of retinal remodeling of the inner retina following deafferentation.7 There have been a number of proposed mechanisms contributing to remodeling and neuronal cell death in the inner retina associated with the loss of photoreceptors, including metabolic alterations,8-11 retinoic acid synthesis,12 changes in protein homeostasis,13 oxidative stress, and changes in glucose metabolism.14 Protein aggregates are associated 83 with neurodegenerative diseases in the brain and have also been correlated to some retinal diseases where age is a primary risk factor (age-related macular degeneration and glaucoma).15 Although it is likely that a combination of the above mechanisms ultimately contribute to retinal remodeling, pathologic protein aggregation or proteinopathy is a probable central component of pathogenesis in retinal remodeling. 5.2 Animal Models of Retinal Remodeling and Possible Caveats Retinal remodeling is associated with loss of retinal input in both humans and animal models. A caveat in the current knowledge of retinal remodeling is that primarily short-lived animal models have been employed to evaluate the extent and components of remodeling.16 This causes a distinct disconnect between the experimental models of vision rescue and the ultimate goal of restoring sight in individuals who have severely remodeled inner retinas following years to decades of vision loss. In contrast to mouse or rat models of retinal degeneration (with mean lifespans of 700-800 days), the rhodopsin transgenic (Tg) P347L rabbit permits evaluation of retinal remodeling over many years (7-8y), recapitulating the remodeling observed in humans (see section 5.3.1).17 Using this model we evaluate late stage remodeling and mechanisms active long after the degeneration of photoreceptors is complete using computational molecular phenotyping (CMP),18 electron microscopy (EM), and immunohistochemical (IHC) analysis of proteins found to aggregate in other neuropathies. 84 5.3 Materials and Methods 5.3.1 Model Systems The Tg P347L rabbit is a rabbit model of cone sparing autosomal dominant retinitis pigmentosa (adRP) originally developed in the Kondo laboratory. It contains a rhodopsin proline 347 to leucine mutation that causes primary rod degeneration.17 Control rabbits were the unaffected littermates of Tg P347L rabbits (WT) (n = 7). We have previously characterized the progression of degeneration found in this rabbit, demonstrating the progression of degeneration in this rabbit follows the same trajectory of degeneration and subsequent remodeling seen in humans.9, 19-21 In brief, the P347L rabbit presents initially with rod photoreceptor degeneration and loss, though some rods survive up to 12-16 weeks, followed subsequently by progressive cone photoreceptor loss. By 40 weeks the majority of photoreceptors have degenerated and the inner retina is in varying states of phase 2 and 3 remodeling. We expanded on the previous work by evaluating 10 Tg P347L rabbits spanning ages 3 mo to 6+ years for protein and metabolic changes. 5.3.2 Tissue Processing Retinal samples were collected as previously described.22-23 All animal experiments were conducted according to the ARVO Statement for the use of animals in ophthalmic and vision research with the supervision and approval of the institutional animal care and use committee (IACUC) at the University of Utah. Animals were euthanized with urethane, isoflurane, or pentobarbital (depending on USDA guidelines at the time of experiment), followed by bilateral thoracotomy. Following euthanasia, tissues were cut into smaller pieces and split between 3 fixative solutions: 85 1. 2.5% glutaraldehyde solution (2.5%FGMg) used for quantification: 2.5% glutaraldehyde, 1% formaldehyde, 0.1M phosphate buffer (pH 7.4), 1 mM MgSO4, 3% sucrose. 2. 4% formaldehyde solution (4%F) used for initial antibody testing and protein expression pattern identification: 4% formaldehyde, 0.1M phosphate buffer (pH 7.4), 1 mM MgSO4, 3% sucrose. 3. Cacodylate glutaraldehyde solution (2.5%Cac) for ultrastructural analysis by transmission electron microscopy (TEM). 2.5% glutaraldehyde, 1% formaldehyde, 0.1M cacodylate buffer (pH 7.4), 1 mM MgSO4, 3% sucrose. All tissues for CMP and TEM were processed as previously described.18, 24-25 In addition to the small molecules used for CMP described below in section 5.3.3, we used glutaraldehyde tolerant antibodies to full protein α-synuclein (α-syn), phosphorylated αsynuclein (P-α-syn), Ubiquitin (Ubi), phosphorylated tau protein (tau), and β-amyloid. For a full list of antibodies used see Table 1. 5.3.3 Computational Molecular Phenotyping (CMP) We used CMP for the identification of cell classes and quantified the changes in protein levels of the inner retina.18, 25-26 Each well of the array slide was probed for an IgG for 1 of 6 small molecules: aspartate (D), glutamate (E), glycine (G), glutamine (Q), taurine (τ), or GABA (γ). All primary antibodies were probed with a gold conjugated secondary then silver intensified for 6.5mins on a 30°C plate.18, 27 Slides were then captured as 8-bit images using a QImaging Retina CCD camera (Surrey, BC) under constant gain and offset. The grayscale images were then computationally aligned using ir-tweak,28 generating a 86 registered volume where each grayscale image is an individual metabolite. We used computational clustering (isodata) to define class signatures (PCI Geomatica), which allows cells to be separated into metabolically relevant cell classes according to small molecule signals.27, 29 Cell classes are then represented as a theme map, in which each color represents a single statistical cell class. These theme maps can then be used for masking all but a single cell type and evaluating the associated protein expression, or can be overlaid on traditional TEM images.10, 25 In addition to theme maps, molecular signals are also viewed in triplicate combinations using red, green, and blue (rgb). For example, τQErgb assigns taurine, glutamine, and glutamate to red, green, and blue channels, respectively. 5.3.4 Protein Variation Quantification In addition to the small molecules used for identifying cell classes present in tissue, we used a number of antibodies for the identification, localization, and relative concentration of proteins. Proteins probed for were: glutamine synthetase (GS), GFAP, phosphorylated ser129 alpha-synuclein (Pα-syn), alpha-synuclein (α-syn), ubiquitin (ubi), tau, beta-amyloid (β-amy), rhodopsin (1D4), and tomato-lectin (TL). Relative protein concentrations were calculated by probing the tissue stacks with monoclonal rhodopsin antibody 1D4 in addition to proteins of interest. Based on our experience using CMP with small molecules, we extended the approach to also include proteins, by using the known concentration of rhodopsin (5 mM) and the known efficacy of 1D430 to estimate the levels of each of the proteins over time. In addition, we performed Western blot analysis to further validate our observations of increases in α-syn in Tg retinas. Briefly, dissected retinas were homogenized, and then solubilized in lysis buffer.31-32 After sonication, homogenates were 87 centrifuged for 15min at 12000 rpm, supernatant transferred to new tubes, and total protein content was measured by the Bio-Rad Bradford protein assay. 20µg of protein was incubated at room temperature for at least 30 mins with Laemmli sample buffer [50mM Tris-HCL, 2% SDS, 100mM DTT, and 10% glycerol], separated on a 15% SDS gel and transferred to a PVDF membrane in transfer buffer. Blots were blocked using 5% dried milk, 0.1% Tween diluted in TBS (TBST). Blots were probed using anti-α-syn (1:2000) and anti-actin (1:2000) overnight at 4°C. Membranes were washed 3x in 5% dried milk in TBST (~5 mins each) and incubated in either goat-anti-mouse or goat-anti-chicken IgGs at a concentration of 1:2000 for 1 hr at room temperature. Membranes were washed 2x in 5% dried milk in TBST followed by 3x in TBST (~5min each). The blots were exposed to film after SuperSignal chemiluminescent substrate treatment (Pierce, Rockford IL) and imaged by FluorChem Q with AlphaInnotech AlphaView software (Cell Biosciences, Inc., Santa Clara, CA, USA). 5.4 Characterization of Aged P347L Rabbit By 40 weeks of age the Tg P347L rabbit retina is largely devoid of rod and cone outer segments and has an extremely depressed ERG (80% scotopic a- and b-wave reduction, and 70% photopic a and b wave reduction.17 During years 2-4 the retina manifests a total loss of photoreceptor nuclei in a patchy fashion. Ultimately, with the loss of cones, the retina enters into phase 3 remodeling (Figure 5.1A-C). Although it was previously assumed (with little evidence) that the retina would enter into a state of equilibrium at this level of remodeling, CMP analysis of retinas from 2-, 4-, 5-, and 6-yearold degenerate retinas show pronounced loss of glutamine synthetase (GS) (Figure 5.2 88 A,D,G,J,M), high levels of metabolic heterogeneity between neighboring Müller cells, and increased levels of anatomical remodeling most notably in their endfeet (Figure 5.2 B,E,H,K,N). This continuum of metabolic revision and structural remodeling continues until the Müller cells have no recognizable metabolic signature or form, but are transformed into thick, overlapping blocks of glia or replaced by a debris-like assemblies filling large areas of the remnant retina (Figure 5.2 C,F,I,L,O). In tandem with the Müller cell alterations and ultimate degeneration in aged Tg P347L rabbits, neurons of all types have been lost, exceeding 90% in some areas. Though a large number of neurons were absent by 6 years, some identifiable neurons remained and these were of particular interest. To evaluate their potential to be long term survivors despite the breakdown of the surrounding retina, we combined CMP with TEM imaging which allows us to map the main molecular classes of cell types in TEM images (see materials and methods). In these zones we examined, virtually all of the remnant ganglion cell somas and processes displayed severe cytoplasmic pathologies, apparent proteins aggregates and/or massive internal vesicle accumulations (Figure 5.3). This led us to explore some of the possible molecular markers of similar pathologies found in CNS neurodegenerations. 5.5 Ubiquitin Expression in Retinal Degeneration The severe late-stage remodeling in the Tg P347L rabbit, including pronounced cell death, metabolic variation, and debris accumulation, is reminiscent of protein aggregation associated with neurodegenerative diseases. For initial evaluation we assessed ubiquitin levels and localization as a measure of protein degradation and aggregation. Soon after the onset of outer segment stress and degeneration levels of ubiquitin are increased in nuclei 89 of the ONL, with focal patchy increases within ganglion cells (Figure 5.4B). Following the loss of the outer segments, but prior to the complete loss of the ONL (~2yr), ubiquitin levels appear stable, with continued low expression in remnant cones, some cells of the INL, focal expression in ganglion cells, and the beginnings of expression in the optic fiber (Figure 5.4C). By 3-4 years, along with the complete loss of the ONL, ubiquitin expression is dramatically increased in neurons of the inner retina (Figure 5.4D). At 5 years the ubiquitin is found to be patchy in its expression. Some cells have low levels, but regional aggregates appear, which are difficult to assign to intracellular or extracellular compartments (Figure 5.4E). At 6 years the ubiquitin localization remains patchy, similar to that seen at 5 years (Figure 5.4F). 5.6 Alpha Synuclein Aggregation in Older Transgenic Retina Given these variations in ubiquitin expression and localization, we hypothesized some other proteins were likely involved in remodeling, perhaps those found in slow CNS neurodegenerations. We evaluated the location and expression of α-synuclein (α-syn), βamyloid, tau phosphorylated at T231, and α-synuclein phosphorylated at locus at S129 (Pαsyn), which is associated with oligomerization. Evaluation of transgenic retinas spanning in age from 3 months through 6 years using protein CMP revealed no aggregation or specialized localization of phosphorylated tau or β-amyloid in the P347L rabbit (data not shown). In contrast, α-syn and Pα-syn showed marked changes in expression and localization as remodeling progressed (Figure 5.5). In WT tissues α-syn is found throughout the retina, with higher levels found in rod outer segments33 and the ONL, than in neurons or the IPL of the inner retina. Pα-syn is found at low levels distributed in neurons 90 of the ONL and the INL, along with some signals in fibers and nuclei of the ganglion cell layer. In Tg retinas, the levels of both α-syn and Pα-syn vary with age and level of remodeling present in the retina. Although the signals may vary depending on the level of remodeling in a given region, the basic pattern is as follows: Early in degeneration (3mo1.5yrs) levels of α-syn are roughly equal to that found in WT, with higher levels found in the outer segments of degenerating rods and lower levels throughout neurons of the outer and inner retina. Pα-syn is dramatically increased in horizontal cells, along with a more subtle uniform increase in the INL, IPL, and GCL. By ~2yrs the thinned ONL shows high levels of α-syn, with increased concentrations in the inner retina as well. Of note, at this stage the α-syn in ganglion cells appears to no longer be uniformly distributed, but rather is aggregated along the outer cell membrane. Extracellular aggregation of Pα-syn around and within some ganglion cells emerges. At ~4yrs degeneration the ONL is largely absent and α-syn is found solely distributed in the nuclei of the inner retina, with some punctate staining distributed within the IPL, presumable at large synapses. Pα-syn levels are twofold higher than α-syn throughout neurons of the inner retina. Pα-syn has also begun to aggregate inside cells, particularly ganglion cells. At 5yrs the Pα-syn continues to aggregate within ganglion cells, in addition to displaced neurons found within the IPL, and at high concentrations in many INL soma. α-syn is found highly expressed in all remaining cells and is predominantly colocalized with high levels of Pα-syn, with an exception of punctate levels in the IPL as IPL staining is unique to the full protein. By 6-7yrs the structure of the retina is severely decimated. α-syn is distributed widely throughout remnant cells and processes of the retina. Pα-syn is not only ubiquitous in the remaining neurons but is also found in aggregates in the cellular debris making up much of the 91 remnant retina. Combined, these data indicate α-syn overall increases throughout degeneration while there is a pronounced increase and eventual aggregation of Pα-syn. In support of these data, Western blot analysis reveals a two-fold increase in monomeric αsyn at 4 years of degeneration in P347L rabbit retina, and continual increase to five-fold by 5 years, which is consistent with data from Parkinson's disease suggesting that native cytosolic multimers shift to a monomeric form as a mechanism of disease initiation.34 In contrast, there is no increase in a-syn in wt rabbit even at 8 years of age, indicating this increase in α-syn is not age-dependent (Figure 5.6). In order to confirm the effects seen in our Western Blot and histological analysis, we performed a blocking experiment, which demonstrated that the antibodies are specific for Pα-syn and α-syn (Figure 5.7). 5.7 Discussion By evaluating structural, ultrastructural, metabolic, and protein alterations in advanced age TgP347L rabbit retinas we characterize late retinal remodeling as a neurodegenerative proteinopathy. Based on these data we conclude that the previously reported plateau of remodeling in experimental animals and its extrapolation to humans is incorrect and due to incomplete data produced by the short lifespan of rodents. This development in our understanding of retinal remodeling presents both new hurdles and opportunities to the vision community and neuroscience as a whole. Progressive neurodegeneration and glial deconstruction after the complete loss of photoreceptors suggests that optogenetic and photoswitch interventions are destined to fail in all patients long term, regardless of their initial mechanism of photoreceptor loss. Similarly, cell regenerative approaches using stem cells or differentiated cellular 92 replacement are also futile because proteinopathy, as in the CNS, likely does not respect cellular origins. A particularly potent example of this the accumulation of pathologic levels of a-syn in some fetal nigral transplants in Parkinson's disease.35 Our discovery of retinal remodeling as a neurodegenerative proteinopathy also affords a number of opportunities to the neurodegeneration community. The retina is unrivaled by any other area of the CNS in its accessibility and our understanding of the associated metabolism and circuitry. These factors combined make the remodeled retina an attractive candidate organ for the discovery of pathways leading to α-synuclein aggregation and ways to prevent or reverse it36 in Parkinson's disease, Lewy-body dementia, multiple-system atrophy, and other α-synuclein associated neurodegenerations. In the short term this has two primary impacts on research: First, groups interested in success in treating retinal degenerations need to devote more resources to generating interventions that prevent or delay the onset of remodeling, such as blocking retinoic acid activation of neuritogenesis.12 Second, it is imperative that at least as many resources are devoted to basic research understanding the development of proteinopathies and neurodegeneration as translational therapeutics. Without basic research understanding all of the primary components leading to diseases like neurodegeneration, the vast majority of the developed therapeutics will have small odds of long-term success. 93 Table 1. Immunocytochemistry reagents Reagent RRID Source Dilution anti-D-aspartate IgG NA Signature Immunologics 1:100 anti-L-aspartate IgG AB_2341093 Signature Immunologics 1:100 anti-L-glutamate IgG AB_2532055 Signature Immunologics 1:100 anti-glycine IgG AB_2532057 Signature Immunologics 1:100 anti-glutathione IgG AB_2532058 Signature Immunologics 1:100 anti-L-glutamine IgG AB_2532059 Signature Immunologics 1:100 anti-taurine IgG AB_2532060 Signature Immunologics 1:100 anti-GABA IgG AB_2532061 Signature Immunologics 1:100 anti-GABA IgY AB_2532062 Signature Immunologics 1:100 anti-GS IgG AB_397879 BD Biosciences 1:50 anti-CRALBP IgG AB_2314227 Gift of Dr. Jack Saari 1:400 anti-GFAP AB_1001338 2 Dako 1:400 Anti-α-syn pS129 NA AbCam ab184674 1:500 Anti-α-syn (pan) NA AbCam ab190376 1:2000 Anti-ubiquitin AB_305802 AbCam ab7254 1:200 94 Figure 5.1 Representative registered images of amino acid immunoreactivity in 100nm sections of WT, Phase 2, and Phase 3 remodeling in rabbit retinas. 1A) γGE rgb of a WT rabbit retina. The lamination of outer segments (OS), outer nuclear layer (ONL), inner nuclear layer (INL), inner plexiform layer (IPL), and ganglion cell layer (GCL) are all clearly laminated. 1B) γGE rgb of a TgP347L degenerate rabbit retina currently in phase 2 remodeling. The yellow circles demonstrate examples of remaining cone cell bodies, although the majority of the OS and ONL is absent. 1C) γGE rgb of a TgP347L degenerate rabbit retina in phase 3 remodeling. White arrows indicate amacrine cells, which have translocated from their appropriate position either into the IPL, or have moved apically towards the remnant RPE. Inlaid density scaled images demonstrate aberrant horizontal direction of some GABAergic processes, indicated by boxes. 95 Figure 5.2 Representative images from WT rabbit (A-C), 2yr (D-F), 4yr (G-I), 5yr (J-L), and 6yr (M-O) Tg P347L rabbit retina. 2A) WT retina stained for glutamine synthetase (GS) showing strong staining throughout Müller cell and their endfeet 2B) WT retina stained for tomato lectin (TL) stains the extra cellular matrix (ECM) allowing the inference of cellular compartments. 2C) τQE  rgb sCMP of a WT rabbit retina. This combination of amino acid facilitates the visualization of most major cell types in the retina. Müller cells appear yellowishorange in this configuration (indicated by the oval). Ganglion cells appear teal (upward white arrow). Amacrine cells are pink or green immediately adjacent to the IPL (examples demonstrated by black upward arrows). Bipolar cells are the cells of varying shades of pink above the amacrine cells (black horizontal arrows). Horizontal cells are the dull green cells at the border between the ONL and INL (downward white arrows). 2D) GS staining of a 2yr P347L rabbit retina. GS staining shows a global decrease in GS levels, but a nonuniform pattern of decrease between neighboring Müller cells. 2E) TL staining of 2yr Tg P347L rabbit retina. The extra cellular matrix stain visualizes the increase in the fingerlike projections of the Müller cell endfeet. Also there is a thinning in the external limiting membrane and the predicted decrease in nuclei in the outer nuclear layer following the loss of outer segments. 2F) τQE  rgb sCMP of a 2yr Tg P347L rabbit retina. The outer segments and outernuclear layer have degenerated, though a few cone cell bodies remain (examples indicated by yellow circles). Müller cells, although similar, have diverged in metabolic profile as indicated by the varying colors of orangeish-yellow (brackets). Müller cell hypertrophy is also visible at this stage (oval). 2G) GS staining of 4yr Tg P347L rabbit retina. At this stage of degeneration GS has dropped to almost undetectable levels. It is likely that using conventional fluorescent staining, this level of GS would be indistinguishable from background fluorescence. 2H) TL staining of a 4yr P347L rabbit retina. The majority loss of the outernuclear layer is now leading to more aberrant architecture of the Müller glial endfeet. In addition to the continual development of finger-like projections the Müller cell endfeet appear to have become intertwined with some large stalks pushing the IPL out of its normal lamination. The outer limiting membrane has also continued to break down. 2I) τQE  rgb sCMP of a 4yr Tg P347L rabbit retina. Müller cells demonstrate a complete loss of metabolic homogeneity, as demonstrated by the range of colors. The IPL is pushed apically in some regions (white arrows) while it is pushed basally in others (black arrows). 2J) GS staining of a 5yr Tg P347L rabbit retina. GS levels remain markedly lower than WT, with more prominent staining along apical region of retina. 2K) TL staining of a 5yr Tg P347L rabbit retina. ECM staining continues to show the elaboration of intertwined Müller cell projections throughout the GCL, where endfeet once resided. Indicated by a circle is a point in which the basement membrane appears to be breached by some components of the inner retina. 2L) τQE  rgb sCMP of a 5yr Tg P347L rabbit retina. The endfoot region of the Müller glia has expanded to double its original size, while the metabolism has not returned to a homogeneous population as evident by the range of orange, brown, and green. Holes have begun to form in some regions, likely due to compromised aquaporin access to the vitreous. The number of neuronal nuclei has decreased, and lamination in the INL is unclear. 2M) GS staining of a 6yr Tg P347L rabbit retina. GS expression is still at an almost undetectable level. 2N) TL staining of a 6yr Tg P347L rabbit retina. The ECM staining demonstrates a lack of identifiable retinal structure beyond a few cell nuclei. 2O) τQE  rgb sCMP of a 6-7yr Tg P347L rabbit retina. The former GCL has now tripled in size, but has only remnants of processes likely to have arisen from ganglion cells. The sole remaining ganglion cell in this region has migrated into the IPL. Some glial-like signatures remain, but they appear 96 drastically altered from those found in WT retina. Intermixed with the glia, there are regions largely devoid of any signature, which are being termed debris. 97 Figure 5.3 Theme map overlay of TEM image of retina. 3A) Theme map generated via k-means clustering of a 6-7yr Tg P347L rabbit retina. Metabolic signatures used for clustering: E, Q, J, D, G, γ, τ, and GS. Purple theme class indicates ganglion cells (GCs), while dark blue in the GCL layer indicates the theme class containing ganglion cell processes. 3B) TEM close-up of large ganglion cell found in the middle of the IPL. Swirl structures are indicative of autophagy. 3C) Magnified TEM image of multiple GC processes in intermediate stage of autophagy. 3D) Magnified TEM image of another GC process almost entirely filled with autophagy-associated swirl structures. 3E) Magnified TEM image of another GC without autophagic swirl structures. Inset magnified mitochondrion demonstrates significant swelling and breakdown of cristae indicating cell stress. 98 99 Figure 5.4: Grayscale ubiquitin labeling in rabbit retinas. 4A) WT P347L littermate control demonstrating light ubiquitin immunoreactivity distributed around Müller cells. 4B) 3mo Tg P347L rabbit retina demonstrating increased expression in ONL and some ganglion cells. 4C) 2yr Tg P347L rabbit retina with moderate levels in nuclei of inner retina. 4D) 4yr Tg P347L rabbit retina. Levels of ubiquitin have risen in soma of the INL and GCL.4E) 5yr Tg P347L rabbit retina, demonstrating variable level of ubiquitin across all cell types. 4F) 6yr Tg P347L rabbit retina. Levels of ubiquitin in probable debris is patchy. The remaining somas have varying levels of ubiquitin, but all above baseline levels. 100 Figure 5.5 Endogenous signaling of phosphorylated α-synuclein (Pα-syn), full protein αsynuclein (α-syn), and GFAP. 5A-C) WT rabbit retina. A) Pα-syn, α-syn, GFAP  RGB mapped B) grayscale Pα-syn and C) grayscale Pα-syn, demonstrating normal expression pattern of Pα-syn and α-syn in the absence of significant stress (indicated by absence of GFAP). 5D-F) 3mo Tg P347L rabbit retina. D) Pα-syn, α-syn, GFAP  RGB mapped E) grayscale Pα-syn and F) grayscale Pα-syn, during early phase 1 retinal remodeling. The largest change is the presence of Pα-syn in the horizontal cells (HCs) indicated by black arrows. 5G-I) 2yr Tg P347L rabbit retina. G) Pα-syn, α-syn, GFAP  RGB mapped H) grayscale Pα-syn and I) grayscale Pα-syn, during early phase 2. The largest change is the emergence of Pα-syn outside of ganglion cells (GCs) indicated by a red circle. 5J-L) 4yr Tg P347L rabbit retina. J) Pα-syn, α-syn, GFAP  RGB mapped K) grayscale Pα-syn and L) grayscale Pα-syn, during late phase 2 retinal remodeling (prior to loss of remnant cones). As the ONL disappears, levels of Pα-syn begin to dramatically increase in the INL and GCL with more diffuse phosphorylated aggregates being found throughout the GCL (downward facing arrows). 5M-O) 5yr Tg P347L rabbit retina. M) Pα-syn, α-syn, GFAP  RGB mapped N) grayscale Pα-syn, during phase 3 retinal remodeling (no remnant photoreceptors). Increased Pα-syn in cells of INL and GCL. In GCs Pα-syn begins to strongly aggregate along outer cell membrane. 5P-R) 6yr Tg P347L rabbit retina. P) Pαsyn, α-syn, GFAP  RGB mapped Q) grayscale Pα-syn and R) grayscale Pα-syn, during the most advanced phase 3 retinal remodeling seen to date. Increased Pα-syn in all remaining cells, and aggregates dispersed throughout previous GCL. 101 102 Figure 5.6 Western Blot: Beta actin demonstrated some increase in Tg retinas when compared to WT. α-synuclein demonstrated 2-fold increase at the 4 year time point and increased to 5-fold by 5 years of degeneration. 103 Figure 5.7 Antibody inhibition experiment. A) Pα-syn, without block. B) α-syn without block. C) Pα-syn inhibited by Pα-syn protein fragment. D) α-syn inhibited by full α-syn protein. 104 5.8 References 1. Busskamp, V.; Picaud, S.; Sahel, J. A.; Roska, B., Optogenetic Therapy for Retinitis Pigmentosa. Gene Ther 2012, 19 (2), 169-175. 2. Petrs-Silva, H.; Linden, R., Advances in Gene Therapy Technologies to Treat Retinitis Pigmentosa. Clin Ophthalmol 2014, 8, 127-136. 3. Seiler, M. J.; Aramant, R. B., Cell Replacement and Visual Restoration by Retinal Sheet Transplants. Prog Retin Eye Res 2012, 31 (6), 661-687. 4. Lamba, D. A.; Karl, M. O.; Reh, T. A., Strategies for Retinal Repair: Cell Replacement and Regeneration. Prog Brain Res 2009, 175, 23-31. 5. Weiland, J. D.; Cho, A. K.; Humayun, M. S., Retinal Prostheses: Current Clinical Results and Future Needs. Ophthalmology 2011, 118 (11), 2227-2237. 6. Weitz, A. C.; Behrend, M. R.; Lee, N. S.; Klein, R. L.; Chiodo, V. A.; Hauswirth, W. W.; Humayun, M. S.; Weiland, J. D.; Chow, R. H., Imaging the Response of the Retina to Electrical Stimulation with Genetically Encoded Calcium Indicators. J Neurophysiol 2013, 109 (7), 1979-1988. 7. Marc, R.; Pfeiffer, R.; Jones, B., Retinal Prosthetics, Optogenetics, and Chemical Photoswitches. ACS Chem Neurosci 2014, 5 (10), 895-901. 8. Fletcher, E. L.; Kalloniatis, M., Neurochemical Architecture of the Normal and Degenerating Rat Retina. J Comp Neurol 1996, 376 (3), 343-360. 9. Jones, B. W.; Kondo, M.; Terasaki, H.; Watt, C. B.; Rapp, K.; Anderson, J.; Lin, Y.; Shaw, M. V.; Yang, J. H.; Marc, R. E., Retinal Remodeling in the Tg P347L Rabbit, a Large-Eye Model of Retinal Degeneration. J Comp Neurol 2011, 519 (14), 2713-27133. 10. Marc, R. E.; Jones, B. W., Retinal Remodeling in Inherited Photoreceptor Degenerations. Mol Neurobiol 2003, 28 (2), 139-147. 11. Pfeiffer, R. L.; Marc, R. E.; Jones, B. W., Muller cell metabolic chaos during retinal degeneration. Exp Eye Res 2016. 12. Lin, Y.; Jones, B. W.; Liu, A.; Tucker, J. F.; Rapp, K.; Luo, L.; Baehr, W.; Bernstein, P. S.; Watt, C. B.; Yang, J. H.; Shaw, M. V.; Marc, R. E., Retinoid Receptors Trigger Neuritogenesis in Retinal Degenerations. FASEB J 2012, 26 (1), 81-92. 13. Lewis, G. P.; Erickson, P. A.; Guerin, C. J.; Anderson, D. H.; Fisher, S. K., Changes in the Expression of Specific Muller Cell Proteins During Long-Term Retinal Detachment. Exp Eye Res 1989, 49 (1), 93-111. 14. Hurley, J. B.; Lindsay, K. J.; Du, J., Glucose, Lactate, and Shuttling of Metabolites in Vertebrate Retinas. J Neurosci Res 2015, 93, 1079-1092. 105 15. Ohno-Matsui, K., Parallel Findings in Age-Related Macular Degeneration and Alzheimer's Disease. Prog Retin Eye Res 2011, 30 (4), 217-238. 16. Chang, B.; Hawes, N. L.; Hurd, R. E.; Davisson, M. T.; Nusinowitz, S.; Heckenlively, J. R., Retinal Degeneration Mutants in the Mouse. Vision Res 2002, 42 (4), 517-525. 17. Kondo, M.; Sakai, T.; Komeima, K.; Kurimoto, Y.; Ueno, S.; Nishizawa, Y.; Usukura, J.; Fujikado, T.; Tano, Y.; Terasaki, H., Generation of a Transgenic Rabbit Model of Retinal Degeneration. Invest Ophthalmol Vis Sci 2009, 50 (3), 1371-1377. 18. Marc, R. E.; Murry, R. F.; Basinger, S. F., Pattern Recognition of Amino Acid Signatures in Retinal Neurons. J Neurosci 1995, 15 (7 Pt 2), 5106-5129. 19. Pfeiffer, R. L.; Marc, R. E.; Kondo, M.; Terasaki, H.; Jones, B. W., Muller Cell Metabolic Chaos During Retinal Degeneration. Exp Eye Res 2016, 150, 62-70. 20. Sakai, T.; Kondo, M.; Ueno, S.; Koyasu, T.; Komeima, K.; Terasaki, H., Supernormal ERG Oscillatory Potentials in Transgenic Rabbit with Rhodopsin P347L Mutation and Retinal Degeneration. Invest Ophthalmol Vis Sci 2009, 50 (9), 4402-4409. 21. Jones, B. W.; Pfeiffer, R. L.; Ferrell, W. D.; Watt, C. B.; Marmor, M.; Marc, R. E., Retinal Remodeling in Human Retinitis Pigmentosa. Exp Eye Res 2016, 150, 149-165. 22. Marc, R. E.; Wei-Ley, S.; Kalloniatis, M.; Raiguel, S. F.; Van Haesendonck, E., Patterns of Glutamate Immunoreactivity in the Goldfish Retina. J Neurosci 1990, 10 (12), 4006-4034. 23. Marc, R. E., Mapping Glutamatergic Drive in the Vertebrate Retina with a Channel-Permeant Organic Cation. J Comp Neurol 1999, 407 (1), 47-64. 24. Anderson, J. R.; Jones, B. W.; Watt, C. B.; Shaw, M. V.; Yang, J. H.; Demill, D.; Lauritzen, J. S.; Lin, Y.; Rapp, K. D.; Mastronarde, D.; Koshevoy, P.; Grimm, B.; Tasdizen, T.; Whitaker, R.; Marc, R. E., Exploring the Retinal Connectome. Mol Vis 2011, 17, 355-379. 25. Marc, R. E.; Jones, B. W.; Watt, C. B.; Strettoi, E., Neural remodeling in retinal degeneration. Prog Retin Eye Res 2003, 22 (5), 607-655. 26. Marc, R. E.; Cameron, D., A Molecular Phenotype Atlas of the Zebrafish Retina. J Neurocytol 2001, 30 (7), 593-654. 27. Marc, R. E.; Jones, B. W., Molecular Phenotyping of Retinal Ganglion Cells. J Neurosci 2002, 22 (2), 413-427. 28. Anderson, J. R.; Jones, B. W.; Yang, J. H.; Shaw, M. V.; Watt, C. B.; Koshevoy, P.; Spaltenstein, J.; Jurrus, E.; U, V. K.; Whitaker, R. T.; Mastronarde, D.; Tasdizen, T.; 106 Marc, R. E., A Computational Framework for Ultrastructural Mapping of Neural Circuitry. PLoS Biol 2009, 7 (3), 493-512. 29. Pfeiffer, R. L.; Marc, R. E.; Kondo, M.; Terasaki, H.; Jones, B. W., Muller cell metabolic chaos during retinal degeneration. Exp Eye Res 2016. 30. MacKenzie, D.; Arendt, A.; Hargrave, P.; McDowell, J. H.; Molday, R. S., Localization of Binding Sites for Carboxyl Terminal Specific Anti-Rhodopsin Monoclonal Antibodies Using Synthetic Peptides. Biochemistry 1984, 23 (26), 65446549. 31. Chakraborty, D.; Ding, X. Q.; Fliesler, S. J.; Naash, M. I., Outer Segment Oligomerization of Rds: Evidence from Mouse Models and Subcellular Fractionation. Biochemistry 2008, 47 (4), 1144-1156. 32. Zou, J.; Zheng, T.; Ren, C.; Askew, C.; Liu, X. P.; Pan, B.; Holt, J. R.; Wang, Y.; Yang, J., Deletion of PDZD7 Disrupts the Usher Syndrome Type 2 Protein Complex in Cochlear Hair Cells and Causes Hearing Loss in Mice. Hum Mol Genet 2014, 23 (9), 2374-2390. 33. Cuenca, N.; Fernandez-Sanchez, L.; Campello, L.; Maneu, V.; De la Villa, P.; Lax, P.; Pinilla, I., Cellular Responses Following Retinal Injuries and Therapeutic Approaches for Neurodegenerative Diseases. Prog Retin Eye Res 2014, 43, 17-75. 34. Dettmer, U.; Newman, A. J.; Soldner, F.; Luth, E. S.; Kim, N. C.; von Saucken, V. E.; Sanderson, J. B.; Jaenisch, R.; Bartels, T.; Selkoe, D., Parkinson-Causing AlphaSynuclein Missense Mutations Shift Native Tetramers to Monomers as a Mechanism for Disease Initiation. Nat Commun 2015, 6, 7314. 35. Brundin, P.; Kordower, J. H., Neuropathology in Transplants in Parkinson's Disease: Implications for Disease Pathogenesis and the Future of Cell Therapy. Prog Brain Res 2012, 200, 221-241. 36. Tran, H. T.; Chung, C. H.; Iba, M.; Zhang, B.; Trojanowski, J. Q.; Luk, K. C.; Lee, V. M., Alpha-Synuclein Immunotherapy Blocks Uptake and Templated Propagation of Misfolded Alpha-Synuclein and Neurodegeneration. Cell Rep 2014, 7 (6), 2054-2065. CHAPTER 6 DISCUSSION AND FUTURE DIRECTIONS The purpose of this project was to understand the long-term effects of retinal remodeling on the neural retina and use this data to gauge likelihood of long-term success of different therapeutic interventions in late-stage degeneration. At the time of the project's inception, the belief in the field (including our laboratory) was that the degenerate neural retina would eventually hit a plateau phase of remodeling, where microneuromas would be prevalent with significant changes in retinal circuitry, some cell loss will have occurred, Müller glia will be altered, and RPE will have infiltrated, but that the neural retina will continue to survive in that remodeled state. The simple conclusion to be drawn from this project is that there is no such plateau in remodeling. Remodeling progresses long after photoreceptors have died but eventually gives way to a full neurodegeneration phenotype apparently similar to CNS neurodegenerations. 6.1 Müller Cell Metabolic Chaos The first key finding of this project was significant alterations in Müller cell metabolism during late phase 3 remodeling. In aged human retinas, it is common to find variable levels of glutathione across Müller cells, which intuitively makes sense given its role in buffering against cellular damage caused by reactive oxygen species, with which 108 we come in regular contact, and given the varied experiences of normal humans (smoking, inflammation, obesity, alcoholism, stress, circadian variations, etc.,) However, in addition to glutathione, the increases in variability of glutamate, glutamine, taurine, and glutamine synthetase in Müller cells, as was found in degenerate retina, was unanticipated. If such changes were driven by the microenvironment, that is, if hypertrophic Müller cell columns changed their metabolic, gene, and protein expression in response to environmental cues generated by regional differences in photoreceptor degeneration, we would expect to find Müller cell metabolic variation along a gradient centered on regions of photoreceptor degeneration. This is not what we observe. Adjacent Müller cells appear to have individual metabolic profiles that vary irrespective of their surrounding environment in retinal degeneration. There are a number of avenues for investigation into the source of these changes. First, we hypothesize a Müller cell synchrony signal that regulates metabolism and synchronizes cells across the normal retina. We expect that it will function based on a number of factors present within the intact retina, perhaps additional photoreceptor signals such as melatonin1 or endothelin, as first described by Rattner and Nathans.2 Many studies attempting to identify potential factors leading to Müller cell regulation use dissociated Müller cells.3-8 If we are correct in our hypothesis that Müller cells are somehow synchronously regulated, dissociated cells are likely to express differing phenotypes than those found in vivo. In fact, gene expression studies have indicated heterogeneity in dissociated mouse Müller cells.7 Even anatomical studies on dissociated cells3 may result in misleading information. Though the level of diversity in Müller cells and the potential for multiple types is still in debate for adult Müller cells, it is likely that they do respond to 109 circadian rhythm. Müller cells have vasopressin receptors, which facilitate their response to the circadian variations in vasopressin secretion by some amacrine cells.9 From this it is possible that there are circadian variations in Müller cell metabolism, but these potential variations would be expected to occur across the retina, not differentially in individual Müller cells. We have evaluated the small molecule signatures of many thousands of Müller cells from hundreds of samples, generated from dozens of species, and all normal retinas display a stereotyped τ+Q+E- signature with exceedingly low variances between Müller cells. This indicates that despite possible genetic and anatomic variations between cells, regulating those metabolites within a narrow range is important to Müller cell function. Second, in healthy retina we observe a clear formal tiling of the endfeet of Müller cells. In healthy retina, the processes originating from a Müller cell column do not intertwine with the processes from adjacent Müller cells, nor do Müller cell endfeet overlap. This motif is lost in degenerating retina (Chapter 5). Müller cell endfeet begin to invaginate to form finger-like projections that encroach upon neighboring Müller cells domains. It is unknown if this process is also occurring more apically as the Müller cells hypertrophy. This phenomenon of glial morphology variation coinciding with metabolic variability during neurodegeneration is not without parallel in the brain.10 McKee et al. found that following repeated head injury astrocytic processes hypertrophied and became intertwined.11 The impact of glial entanglement is currently unclear, however, its coincidence with high levels of metabolic variability raises a question of whether the tangles themselves may affect the metabolic interactions between neurons and glia and contribute to the variability. 110 These hypotheses are not exclusive. Both may play a role in the metabolic disruption seen throughout retinal degeneration, in addition to currently unidentified factors. Determining the functional impact of metabolic and anatomic disruption throughout retinal degeneration is still an ongoing question that will require further research in the future. 6.2 Excitotoxicity in the Retina The concept of excitotoxicity in the brain is extensively cited in the field of neuroscience.12-13 In the retina, it has been proposed that increased glutamate contributes to excitotoxicity in the retina during retinal degeneration.14-15 For this to happen, glutamate near neurites expressing iGluRs must reach a concentration high enough to cause neuronal stress and death. Although glutamate was shown to be toxic and induce lesions in the retina of young mice in 195716 and was replicated through the 1960s,17-18 it is important to consider what concentration of glutamate was used. Lucas and Newhouse report that to produce a lesion with a single subcutaneous injection they needed a near-lethal glutamate level, and to avoid this multiple smaller injections were given.16 Still, these were increasing concentrations from 2mg to 30mg of L-glutamate per day, with a maximum concentration of 150M. For reference, the average retina has a total glutamate concentration of 1.5 mM;19 with some ganglion cells containing the highest level of glutamate, ~10mM, bipolar cells ranging from 1-5 mM, and photoreceptors containing 1 mM or less depending on adaptation state. Glutamate is not transported across the blood retinal barrier or blood brain barrier at physiological levels,20 it is possible the lesions seen in retina may have been caused by a separate mechanism, such as ischemia from particulate21 or RBC22 aggregates 111 following high glutamate exposure, rather than from the added glutamate directly. More recent studies have demonstrated that if the extracellular concentration is raised above 30100µM and maintained there for a couple of hours it is sufficient to induce Müller cell swelling and some neuronal stress,23 though these results are not consistent with previous literature24 or any of our findings. In any case, a large number of glutamate-containing cells would have to die in a single region in an exceedingly short period of time (on the order of minutes), and extracellular glutamate would need to remain at that concentration for hours to induce an excitotoxic effect. To this end, the rate that Müller cells can transport glutamate away from neurons is of interest. Data from our lab have shown that the Müller cells continually buffer up to approximately 5mM extracellular glutamate,25 more than 3 times the total glutamate of the retina.26 Therefore, for glutamate excitotoxicity to potentially contribute to the neurodegeneration seen in the retina, there would have to be significant disruption of glutamate transport by the Müller cells. It has been demonstrated conclusively in Chapter 4 of this dissertation that glutamate transport by the Müller cells is not significantly disrupted until very late stages of retinal degeneration, and most of that is in single Müller cells, not widespread populations. The timepoint at which the Müller cell transport begins to be widely compromised is in our 6-year-old rabbit, which has already undergone massive cell loss. Based on this, it is unlikely that glutamate excitotoxicity plays any significant role in retinal degeneration and future drugs or therapies based on this premise (e.g., neuroprotective glutamate receptor antagonists) are likely to have no effect on the progression of retinal degenerative diseases. 112 6.3 Retinal Neurodegeneration During the course of retinal degeneration it is clear that both remodeling and neurodegeneration play important but differential roles in determining the final state of the neural retina. Previous studies evaluated the neural retina up to a year after the loss of photoreceptors.27 Based on our analysis in Chapter 5, we conclude that retinal remodeling and neurodegeneration are separate overlapping processes occurring over the course of retinal degeneration (illustrated in Figure 6.1). Retinal remodeling begins in tandem with the loss of photoreceptors, and is characterized by progressive receptor alteration, neurite sprouting (followed by microneuroma formation), and early metabolic variation.27-35 Retinal neurodegeneration on the other hand is what leads to the eventual decimation of the retina, and is characterized by metabolic chaos36 and collapse within Müller cells, widespread cell death, debris accumulation, and likely has a proteinopathy component as evident by an increase in α-synuclein. Neurodegeneration begins after photoreceptors are lost and eventually supersedes remodeling years after the photoreceptors were lost, and continuing until the retina is entirely devoid of recognizable cells. Although this development is unfortunate for late-stage therapeutic interventions, this opens up a new potential for evaluating early disease progression in other synucleinopathies. Within the brain, neurodegenerative proteinopathies propagate slowly throughout affected regions following modifications of certain proteins making them prone to misfolding and aggregation.37 One such protein is alpha-synuclein, which is associated with multiple neurodegenerative diseases.38-40 Multiple mechanisms of alpha synuclein toxicity have been proposed, including proteasome impairment and stress, mitochondrial dysfunctions, ER stress, and membrane disruption and pore formation.41 During the course of 113 degeneration, we see ultrastructural evidence for mitochondrial and ER stress. The increases seen in ubiquitin aggregation in the nuclei could also be attributed to the ubiquitin proteasome system being impaired.42 All of these processes are slow progressing in the brain. We see similar signs of neurodegeneration occurring in the retina combined with focal increases in alpha synuclein. We posit that these processes are likely to occur at an accelerated rate in the retina in comparison to the brain in large part due to the close proximity of neurons to one another and relatively short axons, which can facilitate transneuronal propagation in a shorter time frame than that seen in the brain. 6.4 Implications for Therapeutic Interventions Taken together, these findings show that the best time to intervene in retinal degenerative disease is as early as possible. When a specific genetic mutation is known, gene-replacement therapy shows the best promise for preventing widespread photoreceptor degeneration. Immediately following the onset of photoreceptor degeneration, optogenetics or photoswitches may be capable of restoring some, albeit coarse, vision to help provide independence; though complications due to rewiring will have to be addressed (see section 6.5, Future Directions). Later in disease, epi-retinal bionic implants may be a solution to bypass corrupted retinal circuits, though glial separation of ganglion cells from the vitreous and the progressive loss of ganglion cells will be an obstacle. However, Loizos et al. have shown by modeling that it is unlikely that epiretinal implants only stimulate ganglion cells.43 It is unlikely that photoreceptor or whole retina replacement therapy will ever be the best choice. Photoreceptor replacement will always depend on our ability to prevent remodeling or even death of the transplanted cells,44 and if we have that capability 114 the cellular replacements will likely be unnecessary. Whole retina replacement requires the ability to properly rewire the ganglion cell axons correctly to targets in the brain, which is an independently exceedingly difficult and probably insurmountable challenge. From the potential postphotoreceptor degeneration interventions, one fact is clear: preventing neurodegeneration progression will be crucial to long-term vision maintenance. 6.5 Future Directions Based on our assessment of neurodegeneration being a large obstacle to late state retinal degeneration, we propose that the best timepoint for intervention will be as soon as possible after diagnosis. Because it is common for considerable photoreceptor loss to have occurred prior to a patient seeking treatment, remodeling will likely have already altered neural circuits in the inner retina. Prior studies have demonstrated neurite sprouting, with molecular markers indicative of active synapses,33, 45 however, it remains to be seen whether the network topology of this system is correct. The synaptic and gap junction partners of neurons remaining in the retina during remodeling have not been addressed. To fully determine the extent of network alterations in the degenerating retina and their potential impacts on signaling capabilities of the remodeling retina, it will be necessary to construct connectomes for varying phases of remodeling and neurodegeneration. We have generated a preliminary pathoconnectome volume (RPC1) that combines the high resolution ultrastructure of automated TEM with the quantitative aspects of CMP, and is the first pathoconnectome ever generated. RPC1 consists of 948 serial sections spanning from the proximal ONL to the vitreous, with 1 of every 30 sections reserved for CMP data that are intercalated into the volume. In brief, sections were serially sectioned at 90nm and 115 prepared for TEM capture or CMP analysis. The resolution afforded by TEM is vastly superior to serial block face scanning electron microscopy (SBFSEM).46 Our standard TEM acquisitions have a resolution of 2.18 nm/pixel, and because the grids are archival, we have the ability to reimage structures at a resolution of 0.27 nm/pixel. Alternatively, SBFSEM has a published resolution of 8-16 nm, which is too coarse to positively identify gap junctions or small synapses.47 CMP lends itself to positive cell identification through classification enabled by the quantitation of small molecules,19, 25, 48-50 which are stoichiometrically trapped during fixation,19, 48, 51 in addition to glutaraldehyde/osmiumtolerant IgGs. For the identification of all cell structures and genetically labeled cells in the volumes we will use a combination of small molecule IgGs to GABA, glutamate, glutamine, glycine, and taurine, in addition to IgGs targeting GFAP, GFP/YFP, TH, Tomato lectin and GS. RPC1 was assembled using a modified version of the NCR toolkit**. Briefly, the toolkit combines individual images acquired from a JEOL electron microscope with a 16-Mpixel Gatan camera (EM sections) or a Leica light-microscope affixed with 8bit CCD camera (CMP) and registers them into assembled mosaics before using automated registration to align adjacent sections throughout the volume (Figure 6.2).46-47 Following the assembly of the connectome volume, navigation and annotation of the dataset will be done in the software Viking, which facilitates rapid navigation, visualization, and annotation of TEM and CMP data.52 Viking annotations retain information about the size, structure, location, and may be queried for population statistics.47, 53 From a Viking- annotated dataset, we can render 2D and 3D representations of cell including their ** http://www.sci.utah.edu/download/ncrtoolset 116 subcellular components (e.g., pre- and postsynaptic densities, organelles, gap junctions, etc.).54 Preliminary data reveal three distinct phenomena in early degeneration (10mo old P347L rabbit retina). First, dendritic structures of rod bipolar cells are significantly altered indicating the initiation of dendritic remodeling (Figure 6.3A) likely associated with reprogramming described by Marc et al.35 Annotation of these dendrites will allow us to determine what combinations of photoreceptors contribute to the inputs of rod bipolar cells during the loss of rod photoreceptors. Second, by tracking neurite outgrowths from bipolar cells (that Marc et al., argued were supernumerary axons) it is clear that many of the ribbon synapses generated therein are morphologically altered (Figure 6.3B). The evaluation of these ribbons and their synaptic partners may yield insights regarding revised networks of the inner retina. Third, although it has been previously observed that high levels of electron density (presumed to be hyperosmication due to high levels of ribosomes, RNAs, aromatics, or unsaturated fatty acids) in Müller cells of degenerating retina, we had previously described it in the primary trunk. It is now clear that hyperosmication also appears in the finest processes of a Müller cell as it encroaches on a neighboring Müller cell's domain (Figure 6.3C). Continued evaluation of the ultrastructure of the Müller cells, combined with metabolic information afforded by CMP, will allow us to further evaluate the responses of Müller glia to retinal degeneration. 117 Figure 6.1 Retinal degeneration, a combination of remodeling and neurodegenerative processes. Retinal remodeling consists of the processes undergone by the retina in initial response to photoreceptor loss. In phase 1, rods begin to degenerate, and Rod bipolar cells retract their dendrites from rod spherules. Phase 2 continues the loss of photoreceptors with the loss of cone outer segments and complete loss of rods. Phase 2 is also characterized by initial metabolic variability seen in Müller cells. Phase 3 is initiated by the complete loss of photoreceptors. In the remodeling processes, neurite sprouting is abundant and leads to microneuroma formation. While remodeling continues to progress long after photoreceptors are absent, neurodegeneration processes eventually overwhelm the remnant neural retina. Retinal neurodegeneration is characterized by a complete lack of Müller cell coordination and increased α-syn in remaining neurons. As these initial processes persist, cell death occurs over large parts of the retina and coincides with debris accumulation. The timeline demonstrates the increasing impact of retinal neurodegeneration over the lifespan of rabbit and mouse. It is probable that mice do not live long enough for the end stages of retinal degeneration-induced neurodegeneration to be evident. 118 119 Figure 6.2 Workflow for generation of pathoconnectome. a) Tissue from a 10 month Tg P347L rabbit retina was prepared for TEM/CMP via osmication, b) embedded in epon blocs, and c) was serially sectioned at 90nm. Every 30 sections one was reserved for CMP analysis. D) TEM and CMP sections were imaged using a JEOL or Leica light microscope, respectively. Following image capture, sections were assembled, aligned, and imported into a Viking volume by custom software. E) Further annotation will continue in the Viking environment. 120 Figure 6.3 Preliminary RPC1 data. A) 3D reconstruction of rod bipolar cell annotated in Viking, rendered in VikingView. B) Abnormal morphology ribbon synapses from the regions indicated in A. C) Hyperosmication of the protruding tip of a Müller cell invading a neighbor's domain. 121 6.6 References 1. Hiragaki, S.; Baba, K.; Coulson, E.; Kunst, S.; Spessert, R.; Tosini, G., Melatonin Signaling Modulates Clock Genes Expression in the Mouse Retina. PLoS One 2014, 9 (9), e106819. 2. Rattner, A.; Nathans, J., The Genomic Response to Retinal Disease and Injury: Evidence for Endothelin Signaling from Photoreceptors to Glia. J Neurosci 2005, 25 (18), 4540-4549. 3. Anezary, L.; Medina, J. I.; Sanchez-Nogueiro, J.; Lopez-Gallardo, M.; Prada, C., Shape Diversity Among Chick Retina Muller Cells and their Postnatal Differentiation. J Comp Neurol 2001, 438 (1), 32-49. 4. Dai, M.; Xia, X. B.; Xiong, S. Q., BDNF Regulates GLAST and Glutamine Synthetase in Mouse Retinal Muller Cells. J Cell Physiol 2012, 227 (2), 596-603. 5. Poitry, S.; Poitry-Yamate, C.; Ueberfeld, J.; MacLeish, P. R.; Tsacopoulos, M., Mechanisms of Glutamate Metabolic Signaling in Retinal Glial (Muller) Cells. J Neurosci 2000, 20 (5), 1809-1821. 6. Reichelt, W.; Pannicke, T.; Biedermann, B.; Francke, M.; Faude, F., Comparison Between Functional Characteristics of Healthy and Pathological Human Retinal Muller Glial Cells. Surv Ophthalmol 1997, 42 S105-117. 7. Roesch, K.; Jadhav, A. P.; Trimarchi, J. M.; Stadler, M. B.; Roska, B.; Sun, B. B.; Cepko, C. L., The Transcriptome of Retinal Muller Glial Cells. J Comp Neurol 2008, 509 (2), 225-238. 8. Roesch, K.; Stadler, M. B.; Cepko, C. L., Gene Expression Changes Within Müller Glial Cells in Retinitis Pigmentosa. Mol Vis 2012, 18, 1197-1214. 9. Koh, S. W.; Kyritsis, A.; Chader, G. J., Interaction of Neuropeptides and Cultured Glial (Muller) Cells of the Chick Retina: Elevation of Intracellular Cyclic AMP by Vasoactive Intestinal Peptide and Glucagon. J Neurochem 1984, 43 (1), 199-203. 10. Acosta, C.; Anderson, H. D.; Anderson, C. M., Astrocyte Dysfunction in Alzheimer Disease. J Neurosci Res 2017, 95 (12), 2430-2447. 11. McKee, A. C.; Cantu, R. C.; Nowinski, C. J.; Hedley-Whyte, E. T.; Gavett, B. E.; Budson, A. E.; Santini, V. E.; Lee, H. S.; Kubilus, C. A.; Stern, R. A., Chronic Traumatic Encephalopathy in Athletes: Progressive Tauopathy after Repetitive Head Injury. J Neuropathol Exp Neurol 2009, 68 (7), 709-735. 12. Dong, X. X.; Wang, Y.; Qin, Z. H., Molecular Mechanisms of Excitotoxicity and Their Relevance to Pathogenesis of Neurodegenerative Diseases. Acta Pharmacol Sin 2009, 30 (4), 379-387. 122 13. Lai, T. W.; Zhang, S.; Wang, Y. T., Excitotoxicity and Stroke: Identifying Novel Targets for Neuroprotection. Prog Neurobiol 2014, 115, 157-188. 14. Bringmann, A.; Grosche, A.; Pannicke, T.; Reichenbach, A., GABA and Glutamate Uptake and Metabolism in Retinal Glial (Muller) Cells. Front Endocrinol 2013, 4, 1-14. 15. Reichenbach, A.; Bringmann, A., New Functions of Muller Cells. Glia 2013, 61 (5), 651-678. 16. Lucas, D. R.; Newhouse, J. P., The Toxic Effect of Sodium L-Glutamate on the Inner Layers of the Retina. AMA Arch Ophthalmol 1957, 58 (2), 193-201. 17. Olney, J. W., Glutamate-Induced Retinal Degeneration in Neonatal Mice. Electron Microscopy of the Acutely Evolving Lesion. J Neuropathol Exp Neurol 1969, 28 (3), 455-474. 18. Olney, J. W., The Toxic Effects of Glutamate and Related Compounds in the Retina and the Brain. Retina 1982, 2 (4), 341-359. 19. Marc, R. E.; Murry, R. F.; Basinger, S. F., Pattern Recognition of Amino Acid Signatures in Retinal Neurons. J Neurosci 1995, 15 (7 Pt 2), 5106-5129. 20. Yudkoff, M., Interactions in the Metabolism of Glutamate and the BranchedChain Amino Acids and Ketoacids in the CNS. Neurochem Res 2017, 42 (1), 10-18. 21. Tang, N.; Skibsted, L. H., Calcium Binding to Amino Acids and Small Glycine Peptides in Aqueous Solution: Toward Peptide Design for Better Calcium Bioavailability. J Agric Food Chem 2016, 64 (21), 4376-4389. 22. Hossmann, K. A., Reperfusion of the Brain after Global Ischemia: Hemodynamic Disturbances. Shock 1997, 8 (2), 95-101; discussion 102-103. 23. Izumi, Y.; Kirby, C. O.; Benz, A. M.; Olney, J. W.; Zorumski, C. F., Muller Cell Swelling, Glutamate Uptake, and Excitotoxic Neurodegeneration in the Isolated Rat Retina. Glia 1999, 25 (4), 379-389. 24. Pow, D. V.; Robinson, S. R., Glutamate in Some Retinal Neurons is Derived Solely from Glia. Neuroscience 1994, 60 (2), 355-366. 25. Marc, R. E., Mapping Glutamatergic Drive in the Vertebrate Retina with a Channel-Permeant Organic Cation. J Comp Neurol 1999, 407 (1), 47-64. 26. Murry, R. F.; Marc, R. E., Muller's Cells are Not a Major Source of Excitotoxic Glutamate in Hypoxic/Hypoglycemic in vitro Retina. Investigative Opthalmology & Visual Sciences 1997, 34 (4), 2836-B2526. 123 27. Jones, B. W.; Kondo, M.; Terasaki, H.; Watt, C. B.; Rapp, K.; Anderson, J.; Lin, Y.; Shaw, M. V.; Yang, J. H.; Marc, R. E., Retinal Remodeling in the Tg P347L Rabbit, a Large-Eye Model of Retinal Degeneration. J Comp Neurol 2011, 519 (14), 2713-27133. 28. Jones, B. W.; Kondo, M.; Terasaki, H.; Lin, Y.; McCall, M.; Marc, R. E., Retinal remodeling. Jpn J Ophthalmol 2012, 56 (4), 289-306. 29. Jones, B. W.; Marc, R. E., Retinal Remodeling During Retinal Degeneration. Exp Eye Res 2005, 81 (2), 123-137. 30. Jones, B. W.; Marc, R. E.; Pfeiffer, R. L. Retinal Remodeling and Plasticity. http://webvision.med.utah.edu/. 31. Jones, B. W.; Pfeiffer, R. L.; Ferrell, W. D.; Watt, C. B.; Marmor, M.; Marc, R. E., Retinal Remodeling in Human Retinitis Pigmentosa. Exp Eye Res 2016, 150, 149-165. 32. Marc, R. E.; Jones, B. W.; Watt, C. B.; Strettoi, E., Neural remodeling in retinal degeneration. Prog Retin Eye Res 2003, 22 (5), 607-655. 33. Jones, B. W.; Watt, C. B.; Marc, R. E., Retinal Remodeling Clin Exp Optom 2005, 88 (5), 282-291. 34. Marc, R. E.; Jones, B. W., Retinal Remodeling in Inherited Photoreceptor Degenerations. Mol Neurobiol 2003, 28 (2), 139-147. 35. Marc, R. E.; Jones, B. W.; Anderson, J. R.; Kinard, K.; Marshak, D. W.; Wilson, J. H.; Wensel, T.; Lucas, R. J., Neural Reprogramming in Retinal Degeneration. Invest Ophthalmol Vis Sci 2007, 48 (7), 3364-3371. 36. Pfeiffer, R. L.; Marc, R. E.; Jones, B. W., Muller cell metabolic chaos during retinal degeneration. Exp Eye Res 2016. 37. Sweeney, P.; Park, H.; Baumann, M.; Dunlop, J.; Frydman, J.; Kopito, R.; McCampbell, A.; Leblanc, G.; Venkateswaran, A.; Nurmi, A.; Hodgson, R., Protein Misfolding in Neurodegenerative Diseases: Implications and Strategies. Transl Neurodegener 2017, 6 (6), 1-13. 38. Hu, R.; Luo, J.; Wang, W.; Wang, X. F.; Xi, Z. Q., Alpha-Synuclein is a Potential Biomarker in the Serum and CSF of Patients with Intractable Epilepsy. Seizure-Eur J Epilepsy 2015, 27, 6-9. 39. Polymeropoulos, M. H.; Lavedan, C.; Leroy, E.; Ide, S. E.; Dehejia, A.; Dutra, A.; Pike, B.; Root, H.; Rubenstein, J.; Boyer, R.; Stenroos, E. S.; Chandrasekharappa, S.; Athanassiadou, A.; Papapetropoulos, T.; Johnson, W. G.; Lazzarini, A. M.; Duvoisin, R. C.; DiIorio, G.; Golbe, L. I.; Nussbaum, R. L., Mutation in the Alpha-Synuclein Gene Identified in Families with Parkinson's Disease. Science 1997, 276 (5321), 2045-2047. 124 40. Ganguly, G.; Chakrabarti, S.; Chatterjee, U.; Saso, L., Proteinopathy, Oxidative Stress and Mitochondrial Dysfunction: Cross Talk in Alzheimer's Disease and Parkinson's Disease. Drug Des Dev Ther 2017, 11, 797-810. 41. Gallegos, S.; Pacheco, C.; Peters, C.; Opazo, C. M.; Aguayo, L. G., Features of Alpha-Synuclein that Could Explain the Progression and Irreversibility of Parkinson's Disease. Front Neurosci 2015, 9, 59. 42. Zheng, Q.; Huang, T.; Zhang, L.; Zhou, Y.; Luo, H.; Xu, H.; Wang, X., Dysregulation of Ubiquitin-Proteasome System in Neurodegenerative Diseases. Front Aging Neurosci 2016, 8, 303. 43. Loizos, K.; Lazzi, G.; Lauritzen, J. S.; Anderson, J.; Jones, B. W.; Marc, R., A Multi-Scale Computational Model for the Study of Retinal Prosthetic Stimulation. Conf Proc IEEE Eng Med Biol Soc 2014, 2014, 6100-6103. 44. Seiler, M. J.; Aramant, R. B., Cell Replacement and Visual Restoration by Retinal Sheet Transplants. Prog Retin Eye Res 2012, 31 (6), 661-687. 45. Marc, R. E.; Jones, B. W.; Watt, C. B.; Strettoi, E., Neural Remodeling in Retinal Degeneration. Prog Retin Eye Res 2003, 22 (5), 607-655. 46. Anderson, J. R.; Jones, B. W.; Yang, J. H.; Shaw, M. V.; Watt, C. B.; Koshevoy, P.; Spaltenstein, J.; Jurrus, E.; U, V. K.; Whitaker, R. T.; Mastronarde, D.; Tasdizen, T.; Marc, R. E., A Computational Framework for Ultrastructural Mapping of Neural Circuitry. PLoS Biol 2009, 7 (3), 493-512. 47. Marc, R. E.; Jones, B. W.; Lauritzen, J. S.; Watt, C. B.; Anderson, J. R., Building Retinal Connectomes. Curr Opin Neurobiol 2012, 22 (4), 568-574. 48. Marc, R. E.; Cameron, D., A Molecular Phenotype Atlas of the Zebrafish Retina. J Neurocytol 2001, 30 (7), 593-654. 49. Marc, R. E.; Jones, B. W., Molecular Phenotyping of Retinal Ganglion Cells. J Neurosci 2002, 22 (2), 413-427. 50. Marc, R. E.; Murry, R. F.; Fisher, S. K.; Linberg, K. A.; Lewis, G. P.; Kalloniatis, M., Amino Acid Signatures in the Normal Cat Retina. Invest Ophthalmol Vis Sci 1998, 39 (9), 1685-1693. 51. Ottersen, O. P., Postembedding Immunogold Labelling of Fixed Glutamate: An Electron Microscopic Analysis of the Relationship Between Gold Particle Density and Antigen Concentration. J Chem Neuroanat 1989, 2 (1), 57-66. 52. Anderson, J. R.; Jones, B. W.; Watt, C. B.; Shaw, M. V.; Yang, J. H.; Demill, D.; Lauritzen, J. S.; Lin, Y.; Rapp, K. D.; Mastronarde, D.; Koshevoy, P.; Grimm, B.; Tasdizen, T.; Whitaker, R.; Marc, R. E., Exploring the Retinal Connectome. Mol Vis 2011, 17, 355-379. 125 53. Lauritzen, J. S.; Anderson, J. R.; Jones, B. W.; Watt, C. B.; Mohammed, S.; Hoang, J. V.; Marc, R. E., ON Cone Bipolar Cell Axonal Synapses in the OFF Inner Plexiform Layer of the Rabbit Retina. J Comp Neurol 2013, 521 (5), 977-1000. 54. Lauritzen, J. S.; Sigulinsky, C. L.; Anderson, J. R.; Kalloniatis, M.; Nelson, N. T.; Emrich, D. P.; Rapp, C.; McCarthy, N.; Kerzner, E.; Meyer, M.; Jones, B. W.; Marc, R. E., Rod-Cone Crossover Connectome of Mammalian Bipolar Cells. J Comp Neurol 2016.