@prefix vivo: . @prefix edm: . @prefix ns0: . @prefix dcterms: . @prefix dc: . @prefix skos: . vivo:departmentOrSchool "Medicine, Faculty of"@en ; edm:dataProvider "DSpace"@en ; ns0:degreeCampus "UBCV"@en ; dcterms:creator "Hamm, Lisa"@en ; dcterms:issued "2007-12-20T23:02:12Z"@en, "2007"@en ; vivo:relatedDegree "Master of Science - MSc"@en ; ns0:degreeGrantor "University of British Columbia"@en ; dcterms:description """Retinal degeneration is the progressive loss of neurons lining the posterior surface of the eye. Loss of a certain group of neurons called rod photoreceptors can occur as the result of genetic mutation. In humans, and in mammalian models of retinal degeneration, the death of these cells is permanent, and often followed by cone photoreceptor death, which leads to blindness. As a step towards understanding the implications of rod cell death in the retina, we generated transgenic X. laevis that expressed a novel form of caspase-9, with binding domains specific to the compound AP20187. We treated these transgenic animals with AP20187 and caused rod cell death by apoptosis in tadpoles and post metamorphic animals. Peak rod apoptosis occurred two days after drug exposure. We adapted an electroretinography apparatus, and protocols designed for mammals to measure functional changes in X. laevis rod and cone derived responses. We observed delayed secondary cone cell dysfunction after induced rod cell apoptosis, which was subsequently restored. These animals provide a simple and clinically relevant model of diseases like Retinitis pigmentosa, in which we will be able to probe in detail the mechanisms that govern cone cell dysfunction as a consequence of rod apoptosis. The unique ability of this species to recover from this insult will provide clues towards initiating similar recovery in humans."""@en ; edm:aggregatedCHO "https://circle.library.ubc.ca/rest/handle/2429/236?expand=metadata"@en ; dcterms:extent "4391531 bytes"@en ; dc:format "application/pdf"@en ; skos:note " CONTROLLED ABLATION OF ROD PHOTORECEPTORS IN TRANSGENIC XENOPUS LAEVIS by LISA HAMM B.Sc., Brock University, 2004 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in The Faculty of Graduate Studies (Neuroscience) THE UNIVERSITY OF BRITISH COLUMBIA December 2007 © Lisa Hamm, 2007 ii ABSTRACT Retinal degeneration is the progressive loss of neurons lining the posterior surface of the eye. Loss of a certain group of neurons called rod photoreceptors can occur as the result of genetic mutation. In humans, and in mammalian models of retinal degeneration, the death of these cells is permanent, and often followed by cone photoreceptor death, which leads to blindness. As a step towards understanding the implications of rod cell death in the retina, we generated transgenic X. laevis that expressed a novel form of caspase-9, with binding domains specific to the compound AP20187. We treated these transgenic animals with AP20187 and caused rod cell death by apoptosis in tadpoles and post metamorphic animals. Peak rod apoptosis occurred two days after drug exposure. We adapted an electroretinography apparatus, and protocols designed for mammals to measure functional changes in X. laevis rod and cone derived responses. We observed delayed secondary cone cell dysfunction after induced rod cell apoptosis, which was subsequently restored. These animals provide a simple and clinically relevant model of diseases like Retinitis pigmentosa, in which we will be able to probe in detail the mechanisms that govern cone cell dysfunction as a consequence of rod apoptosis. The unique ability of this species to recover from this insult will provide clues towards initiating similar recovery in humans. iii TABLE OF CONTENTS ABSTRACT............................................................................................................................. ii TABLE OF CONTENTS ...................................................................................................... iii LIST OF FIGURES ............................................................................................................... vi LIST OF ABBREVIATIONS ............................................................................................. viii ACKNOWLEDGEMENTS ....................................................................................................x DEDICATION........................................................................................................................ xi CO-AUTHORSHIP STATEMENT .................................................................................... xii CHAPTER 1. Introduction .....................................................................................................1 1.1 BACKGROUND ............................................................................................................1 1.1.1 Visual processing.....................................................................................................1 1.1.2 Photoreceptors.........................................................................................................2 1.1.3 Inherited photoreceptor diseases...........................................................................6 1.1.4 Inducible rod cell apoptosis....................................................................................7 1.2 OBJECTIVES ..............................................................................................................11 1.3 FIGURES......................................................................................................................12 1.4 REFERENCES.............................................................................................................19 CHAPTER 2. Controlled rod cell ablation in transgenic Xenopus laevis .........................24 2.1 INTRODUCTION........................................................................................................24 2.2 METHODS ...................................................................................................................25 2.2.1 Molecular biology..................................................................................................25 2.2.2 Generation, rearing and AP20187 treatment of transgenic X. laevis...............25 2.2.3 Western blots .........................................................................................................26 2.2.4 Microscopy.............................................................................................................26 2.2.5 Electroretinography..............................................................................................26 2.3 RESULTS .....................................................................................................................27 2.3.1 AP20187-induced rod apoptosis in primary transgenic tadpoles.....................27 2.3.2 ICasp9 did not cause rod death in the absence of AP20187..............................28 2.3.3 AP20187 severely compromised retinal function in primary transgenic post-metamorphic frogs.................................................................................................28 iv 2.3.4 Consistent iCasp9 expression in F1 generation tadpoles yields more reproducible induction of apoptosis .............................................................................29 2.3.5 Longitudinal studies on post-metamorphic frogs confirm rod death and suggest cone cell changes ...............................................................................................30 2.4 DISCUSSION ...............................................................................................................32 2.5 FIGURES......................................................................................................................35 2.6 REFERENCES.............................................................................................................42 CHAPTER 3. Electroretinography of Xenopus laevis ........................................................46 3.1 INTRODUCTION........................................................................................................46 3.2 METHODS ...................................................................................................................47 3.2.1 Generation and rearing ........................................................................................47 3.2.2 Animal preparation ..............................................................................................47 3.2.3 Attenuation and amplification .............................................................................48 3.2.4 Stimuli ....................................................................................................................48 3.2.5 Temperature regulation .......................................................................................48 3.3 RESULTS .....................................................................................................................49 3.3.1 Summary of the X. laevis ERG ............................................................................49 3.3.2 Isolating rod and cone responses .........................................................................50 3.3.3 Average amplitude and variance .........................................................................51 3.4 DISCUSSION ...............................................................................................................52 3.5 FIGURES......................................................................................................................54 3.6 REFERENCES.............................................................................................................59 CHAPTER 4. Discussion .......................................................................................................61 4.1 OBJECTIVES ..............................................................................................................61 4.1.1 Goal 1) Induce rod cell death by apoptosis in primary tadpoles ......................61 4.1.2 Goal 2) Establish an F1 generation with consistent rod apoptosis ...................61 4.1.3 Goal 3) Induce rod cell death in transgenic post-metamorphic frogs..............62 4.1.4 Goal 4) Refine methods with which to assay retinal function longitudinally..62 4.1.5 Goal 5) Gain insight into the implications of rod cell death on the retina.......63 4.2 CHALLENGES............................................................................................................64 v 4.2.1 Difficulties inducing rod cell death in post-metamorphic X. laevis. .................64 4.2.2 Measuring small changes in retinal function with electroretinography ..........67 4.3 FUTURE WORK .........................................................................................................68 APPENDICIES ......................................................................................................................74 5.1 APPENDIX I. Optimizing inducible apoptosis system.............................................74 5.2 APPENDIX II. Progeny from male one and male two .............................................77 5.3 APPENDIX III. University of British Columbia Animal Care Certificates ..........81 vi LIST OF FIGURES CHAPTER 1 Figure 1.1. Retinal circuitry ……………………………………………........................12 Figure 1.2. Phototransduction cascade activation and inactivation. ……………………13 Figure 1.3. Developmental stages of Xenopus laevis…………………………………....14 Figure 1.4. Xenopus laevis photoreceptor types……………………………………...…15 Figure 1.5. Comparison of iCasp9 and iCasp3………………………………………….16 Figure 1.6. Structure of AP20187…….…………………….………………………….. 17 Figure 1.7. Subunits of procaspase………………………..…………………………….18 CHAPTER 2 Figure 2.1. Induced rod cell ablation in primary transgenic tadpoles………………..….35 Figure 2.2. AP20187 injection caused compromised retinal function in primary transgenic post-metamorphic frogs……….………………………….........36 Figure 2.3. Inducible apoptosis was more consistent in the F1 generation……………..37 Figure 2.4. Rod cell apoptosis is complete within four days of AP20187 administration………………………………………………………………38 Figure 2.5. Photopic responses were compromised after AP20187 injection, but showed subsequent improvement……….……………………………….....39 Figure 2.6. Five months after initial AP20187 injection, rod cells were ablated, the INL was thicker, and cone cells were intact and more abundant in eGFP-positive animal 3 than in eGFP-negative animal 4…….………....40 CHAPTER 3 Figure 3.1. ERG setup…………………………………………………………………...54 Figure 3.2. Components of Xenopus laevis electroretinogram………………………….55 Figure 3.3. Defining cone evoking stimuli with paired flash paradigm..……………….56 Figure 3.4. Critical flicker frequency was measured at 6.5Hz ……………………..…...57 Figure 3.5. Average amplitude and variability of responses…………………………….58 vii APPENDIX I Figure 5.1.1. AP20187 has no effect in wild type tadpoles……………………………..74 Figure 5.1.2. A dose response analysis of AP20187 in primary transgenic tadpoles shows optimal results at 10nM………………...…………………….……75 Figure 5.1.3. ICasp9 subcloned into a plasmid with double insulators improved the expression of iCasp9 in primary transgenics………………………………76 APPENDIX II Figure 5.2.1. Male 1. Consistent iCasp9 expression at 14 days in eGFP positive tadpoles…………………………………………………………………….77 Figure 5.2.2. Male 1. After 14 days, the expression of iCasp9 decreases such that it interfered with AP20187-induced apoptosis……………………………….78 Figure 5.2.3. Male 2. Expression of iCasp9 differed with the type of eGFP expression, but in the cases where iCasp9 expression was apparent, AP20187 induced apoptosis was dramatic…………………………………………………….79 Figure 5.2.4. Male 2. The progeny from male 2 show high iCasp9 expression, and AP20187 induced retinal apoptosis. ………………………………………80 viii LIST OF ABBREVIATIONS AP Ariad Pharmaceuticals Arr arrestin ARVO association for research in vision and ophthalmology ATP adenosine triphosphate cGMP cyclic guanosine monophosphate CID chemical inducer of dimerization CMV cytomegalovirus CMZ ciliary marginal zone DED death effecter domain DNA deoxynucleic acid dpf days post fertilization eGFP enhanced green florescent protein ERG electroretinography F1 first filial G transducin GCL ganglion cell layer GDP guanosine diphosphate GFAP glial fibrillary acidic protein GMP guanosine monophosphate GTP guanosine triphosphate HA hemagglutinin HEK cells human embryonic kidney cells ICAD inactive caspase activated DNase iCasp9 inducible caspase 9 INL inner nuclear layer IPL inner plexiform layer IS inner segment ISCEV International society for clinical electrophysiology of vision ISI interstimulus interval LGN lateral geniculate nucleus ix ONL outer nuclear layer OPL outer plexiform layer OS outer segment PDE phosphodiesterase R* isomerized rhodopsin R rhodopsin rdCVF rod derived cone viability factor RK rhodopsin kinase RP retinitis pigmentosa RPE retinal pigment epithelium WGA wheat germ agglutinin X. laevis Xenopus laevis x ACKNOWLEDGEMENTS I would like to thank Orson Moritz for his supervision, guidance, and valuable discussions, as well as Beatrice Tam for her teaching and assistance with experiments. I would also like to thank Gerald Li, Alfie Chung and Daniel Yanko for their help with electroretinography, Jenny Wong and Christine Lai for their help with molecular biology, and Anthony Chiam and Chris Wong for assistance with lab maintenance. I would also like to thank Felix Vazquez and David Spencer for their collaboration. xi DEDICATION I would like to dedicate this project to Steven Dirksen and Daniel Yanko. Steve was a close friend, and my room mate through the majority of this degree. He helped me through the beginning by encouraging me, and listening to countless practice presentations. He died in the fall of 2006 in a hiking accident. He wanted me to do the things I enjoyed. Daniel Yanko started as a lab member, and became my best friend. He supported me in so many ways, and helped me deal with things in my personal life, which in turn allowed me to complete this degree. Thanks for showing me what is important, boys. xii CO-AUTHORSHIP STATEMENT Chapters prepared for publication were initiated by ideas from Dr. Orson Moritz. Experiments were carried out by me with the following exceptions: Dr. David Spencer designed the iCasp9 plasmid, Beatrice Tam subcloned iCasp9 initially into the pXOP0.8- eGFP-N1 vector, Gerald Li designed the initial electroretinograph amplification system. Daniel Yanko and Alfie Chung assisted with electroretinography experiments, and Beatrice Tam, Christine Li, and Jenny Wong assisted with molecular biology. Text was edited primarily by Dr. Orson Moritz, and subsequently by my supervisor committee members. CHAPTER 1. Introduction 1.1 BACKGROUND 1.1.1 Visual processing Perception In surprising contrast to a camera, our visual systems are not simply replicating an image of the world for our brains to see. Rather, each level of neural networking yields a representation that is in itself a perception of the visual world. Neurons work together to extract the important information, omit what is extraneous, and fill in what is missing. Elements of a scene that take on visual importance are variable, particularly among different species, and therefore so are the associated neural networks and resulting perceptions. In the case of adult X. laevis, it has been proposed that perception of visual stimuli is indistinguishable from the tactile sensations1. Although X. laevis and human visual processing and perceptions are likely quite different, the retinal cells and circuitry within the eye are strikingly similar2, and as such, we can study this first representation of visual information in many species. The retina The retina is responsible for transducing light information into chemical signals, and then comparing the signals, and presenting the lateral geniculate nucleus (LGN), and to a lesser extent, the superior colliculus, with partially interpreted information about the visual field. In order to accomplish this, there is an intricate circuitry from photoreceptors to ganglion cells. This neural network is one of the most accessible in the brain, and is also on a much smaller scale in terms of the number of neurons (millions rather than billions), and has therefore been studied extensively. Anatomical work was pioneered by Cajal3, the physiology by Adrian and Matthews4 and Hartline5, and countless others have built from this foundational work. Despite the vast knowledge we have gained, large portions of retinal circuitry remain elusive. Here I will give a brief overview of what we know: 2 Light passes through the ganglion, amacrine, bipolar, horizontal and Muller glia to reach the photoreceptors, possibly through Muller cells acting as optical fibers6. Once the light reaches the receptors it is converted to chemical energy and amplified through a process called phototransduction. This results in inhibition of the continual glutamate release at the outer plexiform layer (OPL), called the dark cycle. Changes in glutamate release are detected by receptors on bipolar and horizontal cells within the spherules and pedicles of rods and cones respectively. Bipolar cells are primarily responsible for conveying the graded potentials to ganglion cells. Horizontal cells modulate their responses to specify the relative inhibition of photoreceptor dark cycles across the retina. The projections from bipolar cells terminate in one of ten strata in the inner plexiform layer (IPL), the position of which marks the nature of ganglion cell to which it will be connected. At the synaptic juncture between bipolar and ganglion cells the signal is modified by over thirty types of interneurons analogous to horizontal cells, named amacrine cells, and for this reason the IPL contains more complex synapses than the OPL2 (See Figure 1.1, as well as Wassle, 20047, and Baccus, 20072 for a more in-depth review). 1.1.2 Photoreceptors Photoreceptors are generally categorized into two types, rods and cones. Within these two cell types, photoreceptors are further categorized based on the opsin they contain, allowing isomerization at different wavelengths of light. Humans have one type of rod and three types of cones, long, medium, and short wavelength sensitive (L, M and S). Phototransduction Phototransduction is initiated by isomerization of the visual pigment 11-cis-retinal to all- trans-retinal within a seven-transmembrane domain G-protein coupled receptor called an opsin. The resulting conformational change in the opsin (in this state called metarhodopsin II) promotes interaction with the G-protein transducin, causing exchange of GDP for GTP, and subsequent dissociation of the transducin alpha subunit from the beta and gamma subunits. The alpha subunit interacts with phosphodiesterase (PDE), allowing the hydrolysis of the second messenger cGMP to GMP. When the concentration of cGMP decreases, cGMP gated channels close (for review see8, 9) preventing the flow 3 of sodium and calcium into the cell and hyperpolarizing the membrane potential. This causes a reduction in glutamate release which is detected by bipolar cells as a graded potential, and the previously described retinal circuitry takes effect. The mechanism by which metarhodopsin II is inactivated and subsequently converted back to the rhodopsin form is initiated by phosphorylation by rhodopsin kinase, and subsequent binding of arrestin10, 11. This also elicits a conformational change that allows the release of all-trans-retinal, and leaves the receptor in an apo-opsin form. In some cases, this form of opsin can also activate transducin, although with a much lower probability. The chromophore is recycled (from all-trans-retinal to 11-cis-retinal) by the retinal pigment epithelium (RPE) and Muller cells. Upon binding of 11-cis-retinal to apo- opsin, the light sensitive opsin is reinstated. (See Figure 1.2 for a summary of phototransduction activation and inactivation, and Burns and Arshavsky for review12). Properties of rod and cones Cones provide visual information only in mesopic and photopic condition, as they are 102-103 times less sensitive than rods 13. This reduced sensitivity occurs for two reasons. First, the thermal or spontaneous isomerization of 11-cis-retinal is much higher in cones than in rods, equivalent to 200 photons every second. Second, since the bound chromophore dissociates at a higher rate in cones than rods, more cone opsins are in the apo-opsin state at any given time. Cone apo-opsins can activate transducin independent of photon interaction, as mentioned earlier. Together, this increase in noise could account for the reduced sensitivity (decreased d’14) in cones compared to rods15, although events further downstream are likely to contribute as well13. Rods also have inherent noise, but the spontaneous activation occurs at a much lower rate, so noise, and therefore the detection threshold, is much lower in rods than in cones. The mechanism for increased temporal resolution in cones has also been studied. The proposition is that increased noise leads to a higher level of phototansduction cascade activation, and therefore lower levels of intracellular calcium. Calcium has two roles in the phototransduction cascade, the first is to inhibit rhodopsin kinase, (which normally 4 causes opsin deactivation), and the second is to inhibit guanylate cyclase via binding to guanylate cyclase binding protein. Lower calcium levels therefore increase the activity of two enzymes responsible for recovery, allowing faster response times13. Although the same genes encode many of the phototransduction enzymes in rods and cones, some differ, and therefore contribute to the differences between cell types. Opsins and alpha transducins are cases in which rods and cones express distinct forms of the same phototransduction components. The different opsins account for different spectral sensitivities and different levels of spontaneous activation, while alpha transducins account for discrepancies in light adaptation. The cone alpha transducin subunit is bound to the disk membrane more tightly than the rod version. Upon bleaching, the rod alpha transducin is redistributed throughout the inner and outer segments, leading to a transient lack of function16, while the cone alpha transducin remains bound within the outer segment, and is readily available for subsequent activation. More is known about rod phototransduction cascades than cones, as rod derived proteins are more easily isolated. However, new methods for cone enzyme isolation allows the mechanisms of these differences to be more firmly established13. X. laevis rods and cones X. laevis are anuran amphibians, from the family Pipidae. They were first employed by Nieuwkoop and Faber in 195617 to articulate the steps involved in early development (Figure 1.3). Given the large diameter of their photoreceptor cells (seven µm), they were also used as a source to harvest visual proteins, and record electrophysiological data from single cells (used in the 1950s by Dartnall and Wald, and still used today15). More recently, it became clear that this species is a valuable resource for genetic research. The combination of large photoreceptors, and potential for genetic manipulation18, 19 opened the doors for transgenic X. laevis to play a significant role in retinal research. Both humans and X. laevis are diurnal and have a high concentration of cones in comparison to nocturnal species like mice. Humans have an estimated 4.5 million cone 5 cells and 90 million rod cells, with cone cells dominating the macular region, and rods the periphery. X. laevis have an approximately one-to-one rod to cone ratio20, and these cell types are evenly distributed throughout the retina. Much is known about X. laevis photoreceptor types, mostly due to the work of Witkiovsky, Rohlich and Szel (for review see20, 21). Photoreceptors synapse with horizontal and bipolar cells in X. laevis as they do in humans (similar to that pictured in a mammalian retina in Figure 1.1B and C). At developmental stage 40 these synaptic densities form, and shortly after this stage it is possible to record light-induced ganglion cell firing (Figure 1.3). There are also gap junctions between bipolar cells, as well as between photoreceptor cells, as seen in human retinas. X. laevis rods are connected by gap junctions extensively, as are cones, but the number of gap junctions between the two groups of photoreceptors is much smaller than in mammals21. X. laevis rods are broken into two sub types. Large red transmitting, green absorbing rods, containing opsins maximally sensitive to 520nm light22, make up 97 to 98 percent of rods (known as “red rods”). Thinner, green transmitting, blue sensitive (437nm) rods (“green rods”) make up the remaining two to three percent. These opsins couple with Vitamin A2, the chromophore common to most fully aquatic animals. There are more types of cones in the X. laevis retina than rods, as is the case in humans. In 1985, the availability of monoclonal antibodies expedited the classification of these cells. There are three known cone cell types. Large, long wavelength (611nm) cones make up 86 percent of cone cells, include single and double subtypes, and are recognized by the COS-1 antibody. Large, short wavelength (415-426nm depending on the chromophore23) cones make up ten percent, and are recognized by the OS-2 antibody. Miniature, ultraviolet wavelength cones are thought to make up the remaining four percent of cones, although some controversy exists about this cell type (For review see 24, and Figure 1.4). Anti-calbindin labels each type of cone cell in the X. laevis retina. 6 1.1.3 Inherited photoreceptor diseases There are over 180 genes that, when mutated, are associated with a reduction of photoreceptor function. A summary of these genes is maintained by Stephen P. Daiger at www.sph.uth.tmc.edu/Retnet/. These genes encode proteins that are heterogeneous in expression, biochemical pathways, and secondary implications and therefore beget diverse phenotypes21, 25 26. Rod to cone dystrophies Within this expansive group, there is a subset of mutations that are expressed exclusively in rod cells. In the case that phototransduction or downstream cellular communication is inhibited, but viability is maintained, patients are only mildly inconvenienced. Rod cells are merely dysfunctional, and limit night and peripheral vision. This pathology is known as congenital stationary night blindness (although it is more commonly caused by a mutation in rod bipolar cells than in rods). However, if expression of the mutation causes a toxic gain-of-function, apoptosis is initiated, resulting not only in loss of rod cell function, but also loss of cone cells by an uncharacterized secondary pathway27-29. This secondary cell loss is debilitating for patients, as it eventually results in blindness. Theories on secondary cone death This secondary pathway has been the focus of many studies28, 30, 31. Early research suggested cone cell death could be the result of toxic factors being released by rod cells during death. This hypothesis became less attractive when it was established rod cells almost invariably die by the typically non-inflammatory process of apoptosis32. However, this theory is still possible, as Ripps33 speculated that this toxicity could be in the form of excess calcium and depleted ATP transported by diffusion through gap junctions between rod and cone cells. (Recall that this would occur to a lesser extent in X. laevis, which have fewer rod to cone gap junctions.) Ripps further postulated that not only toxicity, but protective factors could be transported via these channels from rod to cone cells, and the loss of this protection could result in cone cell death. Others have proposed that rods provide necessary components to maintain cone health, but most have focused on 7 secreted factors34-36. Perhaps the most famous was described by Leveillard et al.37, which was named the rod derived cone viability factor (rdCVF). Many other diffusible growth factors have also been shown to slow secondary cone death, originating from various cell types, suggesting rods could have a protective effect through the RPE, interphotoreceptor matrix, or Muller cells34, 36, 38. Banin et al.39, and later Jones and Marc40 focused on the effects of ectopic synaptogenesis between bipolar, horizontal and photoreceptors cells, and suggested these effects could occur prior to cone cell death, potentially acting as the source of cone dysfunction and ultimately death. This hypothesis is supported by inner nuclear layer synaptic reorganization in the absence of cones41. According to this hypothesis, remodeling occurs rapidly when input is lost from a subset of cells, and this reorganization of circuitry is generally not sustainable. These alterations lead to neuroma formation and Muller cell obstruction between photoreceptors and bipolar cells, together causing malfunction and death of retinal neurons. Recently, researchers such as Hackam29, and Shen42 have turned back to the idea of toxicity. This new hypothesis postulates that excess extracellular oxygen accumulates when rod cells are no longer using the vast supply needed for maintenance of the dark cycle. This increase in oxygen causes oxidative damage to adjacent cone cells43. As progress continues in this area, it is important to have models in which to test and develop novel theories about how this interaction occurs. Animals that have a variety of rod to cone ratios, photoreceptor sizes, and interaction properties provide useful clues to elucidate this pathway, and manipulate it for therapeutic purposes. 1.1.4 Inducible rod cell apoptosis For this purpose, in vivo retinal degeneration models are quite valuable, since the interaction between photoreceptors and their environment can be analyzed (organotypic cultures44 have similar benefits). Baehr and Frederick25 and Chadler45 reviewed vertebrate animal models of inherited photoreceptor disease, both naturally occurring, 8 and transgenic. Inducible transgenic models are less common, and typically accomplished using loxP/cre recombinase conditional technology. It is also possible to induce damage physically, by retinectomy, scratches, bright lights, or with various chemicals. Each of these types of models have benefits for certain applications. Since we are interested in the implications of rod death for investigation of human disease processes, we chose to induce rod death by apoptosis, as it is the common path of rod death due to genetic mutation32. A homodimerization system designed by Spencer et al.47 with Schreiber and Crabtree laboratories 46, 47 , subsequently modified by Ariad Pharmaceuticals, allowed modified FK506 binding domains (Fv domains) to be fused to any protein that is activated by dimerization (such as extracellular growth factors, integrins, caspases and many others). When fused proteins are expressed, they sit dormant until contact with a compound that binds two domains together, causing dimerization in the correct orientation, and therefore mutual activation48. We contacted David Spencer, who has worked extensively with controlled dimerization 47-50, about directing this system towards apoptosis in rod cells. He sent us his plasmids containing the inducible caspases, with minor variations, and we subcloned these caspases with associated binding domains into plasmids for use in our lab. We tested these constructs driven by the CMV promoter in HEK 293 cells and determined caspase 9 was the most effective agent of inducible apoptosis both in HEK cells, and in a preliminary transgenic X. laevis experiment (Figure 1.5). Caspase 9 is an initiator caspase associated with mitochondrial activation. As an upstream component of the apoptotic cascade, the probability of amplification is higher. However, the disadvantage to an upstream protein is that there is more opportunity for the cascade to be subject to internal checks and anti-apoptotic resistance through pathways including Bcl-2 and Bcl-xL49. AP20187 is the current dimerization drug used in the homodimerization kit available through ARIAD Pharmaceuticals (Figure1.6), and the one we used in this project. It was 9 initially based on a dimer of FK506 motifs (FK1012) that was able to bind two FK506 binding domains (FKBPs). This dimer was modified to eliminate the innate ability of the complex to inactivate calcineurin, and therefore initiate immunosuppressive activity (the function this complex is best known for47). Ariad Pharmaceuiticals (AP) then synthesized a similar drug called AP1510, and subsequently AP20187, which were designed in conjunction with a FKBP12 containing a single Phe36Val substitution (Fv), allowing minimal endogenous FKBP12 binding, and maximizing dimerization with the Fv fused proteins 49, 51. This substitution allowed the acetyl group of the AP drugs to establish a deeper, and more specific interaction with the binding domain than is possible in the endogenous interaction48. In these studies carried out by Spencer and colleagues47-49, minor variations on the drug, binding domains, associated caspases and localization were found to be best suited for different applications. The optimal balance between low basal activity and high expression is tenuous and changes depending on the application. The caspase 9 derivative tended to be more sensitive, but less robust, while the caspase 3 derivative yielded more apoptosis, but, had higher propensity for auto-activation in the absence of AP20187. Studies articulating the interaction between a chemical inducer of dimerization (CID, or FK1012 based drug) and FKBP12 or appropriate binding domain, had predominantly been carried out in tissue culture. However, in 2000 Jin and colleagues52 fused Fv domains to growth factors (erythropoietin receptors, Kit and Mpl) in order to induce proliferation of transplanted stem cells expressing a therapeutic protein. In 2002, Mallet and colleagues50 fused the two myristoylated Fv domains to caspase 3 for targeted cell death, with objectives similar to ours, but directed towards hepatocytes in mice. The results were promising, and the potential for similar studies in the X. laevis retina seemed feasible. It was possible to induce varied levels of hepatocyte ablation depending on the expression levels of the inducible caspase and concentration of AP20187. The effectiveness of this system in retinal tissue is unknown, and could be different for two reasons. First, the ability of AP20187 to cross the blood retinal barrier is unknown, although FK506 penetrates the blood brain53 and retinal 54barrier. Second, there are ten to 10 50 times more immunophilins in the brain than in the immune system, and the binding levels are up to 50 times higher (see review by Snyder, Lai and Burnett53). If the X. laevis FKBP12 differs in such a way that endogenous FKBP12 activity is elicited by AP20187, despite alterations to minimize activity, the endogenous effects may compete with induced cell death, especially in the case of caspase 954, as will be discussed. Apoptosis Apoptosis is the process of programmed cell death. It is of interest in many clinical contexts, and has been reviewed extensively55-61. It can be initiated intrinsically or extrinsically for an individual cell, and is dependent on cysteine aspartic-specific proteases (CASPases, or caspases). The extrinsic pathway relies on a ligand such as CD95 or TNF-R1 binding to the Fas receptor, which recruits an adaptor protein and subsequently activates procaspase 8 or 10 via their death effector domains (DEDs). This binding results in multiple procaspase zymogens in close proximity, which leads to auto-activation by mutual cleavage on the carboxyl side of an aspartic acid residue. Upon activation, the prodomain is degraded, and two cleaved caspase subunits (p20 and p10, Figure 1.7) from two caspases form a heterotetramer, which is the active form of the protein. This protease activates downstream caspases by the same mechanism. The intrinsic apoptotic pathway involves mitochondrial release of cytochrome c. The adaptor protein Apaf-1 binds cytochrome c and the complex subsequently binds procaspase 9 through recruitment domains (CARDs), to form an apoptosome. Again, aggregation of procaspase 9 is also thought to be associated with auto-activation, and the common effector caspases are activated. The effector, or downstream caspases (3, 6 and 7) are activated by caspase 8, 9 or 10. Effector caspases activate other enzymes that lead to the demise of the cell, such as the inactive caspase activated DNase (ICAD), as well as nuclear lamins, fodrin, gelsolin, and PAK2, together resulting in DNA fragmentation, the changing shape of the cell known as 11 ‘blebbing’, and the condensation and breaking apart of the nucleus often referred to as ‘pyknotic’. Since the final mechanisms of cell death via apoptosis are predicable, it is possible to identify cells dying by this means. Methods of identifying apoptotic cells include morphology, labeling DNA fragmentation, labeling cell surface phosphatidyl serine, or measuring activated caspases. Based on some preliminary experiments identifying caspase 9 as the most effective option, the pathway of interest for this project is intrinsic apoptosis. This mitochondrial, or cytochrome c-mediated apoptosis has been summarized extensively62-64. Relevant points about this pathway in the context of induced dimerization include 1) although proteolytic activation is sufficient for the activity of most caspases, it has been speculated that apoptosome formation is necessary for the downstream effects associated with caspase 957, 2) FK506 (the endogenous ligand for the FKBP12 domain, and precursor to Ariad’s AP20187) inhibits calcineurin which inhibits Bad dephosphorylation, and therefore release of cytochrome-c54, meaning it has anti-apoptotic effects upstream of caspase 9 in the intrinsic pathway, and 3) if the cytochrom c, Apaf-1, Caspase 9 pathway is inhibited by knockout of any component, a common developmental brain defect is seen, suggesting this particular mechanism of apoptosis is important in neural tissue63. 1.2 OBJECTIVES Our goal was to generate transgenic X. laevis in which we could test the effects of rod death in the retina. We broke this goal down into five objectives: 1) Induce rod death by apoptosis in primary tadpoles. 2) Establish an F1 generation with consistent rod apoptosis. 3) Induce rod death in transgenic post-metamorphic frogs. 4) Refine methods to study retinal function in post-metamorphic X. laevis. 5) Gain insight into implications of rod death on the retina. 12 1.3 FIGURES Figure 1.1. Retinal circuitry. There are six groups of neurons within the retina. These include rods (1), cones (2) (each containing an outer segment OS and inner segment IS), horizontal cells (3), bipolar cells (4), amacrine cells (5), and ganglion cells (6). These cells are organized into three layers of cell bodies, the outer nuclear layer (ONL), inner nuclear layer (INL), and ganglion cell layer (GCL). The synapses are organized into two regions, the outer plexiform layer (IPL) and the inner plexiform layer (IPL). These synapses are complex, and pictured in cone pedicles (B), rod spherules (C), and bipolar terminal buttons (D and E). Modified from Wassle, 20047. 13 Figure 1.2. Phototransduction cascade activation and inactivation. The upper disk shows an inactive opsin (R), all subunits of transducin (G), and phosphosdiesterase (PDE) separate in the inactive state. In this situation, the cyclic nucleotide gated channel is open allowing transport of sodium and calcium into the outer segment. In the middle disk, rhodopsin is in the activated form (R*), and is bound to transducin. This causes dissociation of alpha transducin from the gamma and beta subunits, and activation of PDE. The decrease in cyclic GMP (cGMP) resulting from PDE activation causes closure of the CNG channel. The lower disk shows rhodopsin kinase (RK) phosphorylation of R*, and subsequent arrestin (Arr) binding, causing R* to return to R. A similar inactivation occurs for tansducin and PDE by the RGS9, Gβ5, R9AP complex. Modified from Burns and Arshavsky, 200512. 14 Figure 1.3. Developmental stages of Xenopus laevis. The drawings are given with corresponding stage and day post fertalization (dpf). Beyond stage 40, action potentials are measurable upon light stimulation in ganglion cells. Modified from Nieuwkoop and Faber, 195617. 15 Figure 1.4. Xenopus laevis photoreceptor types. There are two types of rod cell, large red transmitting, green absorbing rods, and thinner, green transmitting, blue sensitive rods. There are three known cone cell types, including large, long wavelength cones (of which there are double and single subtypes), short wavelength cones, and miniature, ultraviolet wavelength cones. Modified from Witkovsky, 200020. 16 Figure 1.5. Comparison of iCasp9 and iCasp3. In HEK 293 cells expression of intact iCasp3 was slightly higher than iCasp9, however it was not broken down by AP20187 as well as in iCasp9. In cells expressing iCasp9 and treated with AP20187, we observed depleted intact iCasp9 (green signal in trasnfected HEK 293 cells), as well as pyknotic nuclei. In transgenic tadpole retinas treated at seven dpf, and sectioned at 14 dpf, rod cells were healthier in untreated iCasp9 expressing retinas (healthy outer segments visible by red WGA staining, as well as normal rhodopsin levels) than in untreated iCasp3 (suboptimal outer segment structure in microscopy, and a 48% reduction in rhodopsin levels). Upon AP20187 exposure, rod cell death was apparent in iCasp9, but morphology and rhodopsin levels in iCasp3 animals were comparable with and without AP20187 exposure. Since the problem with iCasp3 was likely basal activity, the myristolated form was not attempted in transgenic animals, as higher basal activity was suspected given the greater probability of interaction in a two-dimensional membrane. (iCasp9 n=20 per group, iCasp3 n=12 per group) 17 Figure 1.6. Structure of AP20187. The compound contains two FK506 type-motifs. The original molecule developed for this application was comprised of two FK506 motifs in with allyl constituents removed to prevent calcineurin binding, termed FK1012. Further modifications were introduced into FK1012 to increase the specificity of the drug, as described in the text, resulting in the illustrated compound AP20187. Modified from the ARIAD Pharmaceuticals manual, 2005. 18 Figure 1.7. Subunits of procaspase. Procaspases include a prodomain at the n-terminus, and large and a small subunits on the c-terminus. When autoactivated by a cysteine residue on an adjacent caspase, the protein dissociates into two subunits and a prodomain. The large and small subunits form a heterotetramer with the two subunits of the adjacent caspase, and the prodomain is degraded. The binding domain (Fv) and HA epitope tag are located on the N-terminous in our constructs, and are therefore present in inactive, or intact caspases, and are degraded with the prodomain upon activation. 19 1.4 REFERENCES 1. Zeeya M. How African frog hunts with eyes wide shut. New Scientist 2005;187:13-13. 2. Baccus SA. Timing and computation in inner retinal circuitry. Annual Review of Physiology 2007;69:271-290. 3. Cajal SR. La retine des vertebres. Springfield, Illinois.: Thomas; 1892. 4. Adrian ED, Matthews R. The action of light on the eye. Part III. The interaction of retinal neurones. Journal of Physiology 1928;65:273-298. 5. Hartline HK. Visual receptors and retinal interaction. Science 1969;164:270-278. 6. Franze K, Grosche J, Skatchkov SN, et al. Muller cells are living optical fibers in the vertebrate retina. Proceedings of the National Academy of Science 2007;104:8287- 8292. 7. Wassle H. Parallel processing in the mammalian retina. Nature Reviews 2004;5:1- 11. 8. Travis GH. DISCO! Dissociation of cone opsins:The fast and noisy life of cones explained. Neuron 2005;46:840-842. 9. Fernald RD. Casting a genetic light on the evolution of eyes. Science 2006;313:1914-1918. 10. Krispel CM, Chen C-K, Simon MI, Burns ME. Novel form of adaptation in mouse retinal rods speeds recovery of phototransduction. The Journal of General Physiology 2003;122:703-712. 11. Birch DG, Hood DC, Nusinowitzet S, Pepperberg DR. Abnormal activation and inactivation mechanisms of rod transduction in patients with autosomal dominant retinitis pigmentosa and the pro-23-his mutation. Investigative Ophthalmology and Vision Science 1995;36:1603-1614. 12. Burns ME, Arshavsky VY. Beyond counting photons: Trials and trends in vertebrate visual transduction. Neuron 2005;48:387-401. 13. Tachibanaki S, Shimauchi-Matsukawa Y, Arinobu D, Kawamura S. Molecular mechanisms characterizing cone photoresponses. Photochemistry and Photobiology 2007;83:19-26. 14. Stanislaw H. Calculation of signal detection theory measures. Behavior Research Methods, Instruments and Computers 1999;31:137-149. 20 15. Kefalov V, Estevez ME, Kono M, Goletz PW, Yau K-W. Breaking the covalent bond - A pigment property that contributes to desensitization. Neuron 2005;46:879-890. 16. Chen J, Wu M, Sezate SA, McGinnis J, F. Light threshold-Controlled cone alpha- transducin translocation. Investigative Ophthalmology and Vision Science 2007;48:3350- 3355. 17. Nieuwkoop PD, Faber J. Normal table of Xenopus laevis. Amsterdam: North Holland Publishing Company; 1956. 18. Kroll KL, Amaya E. Transgenic Xenopus embryos from sperm nuclear transplantations reveal FGF signaling requirements during gastrulation. Development 1996;122:3173-3183. 19. Amaya E, Kroll KL. A method for generating transgenic frog embryos. Methods in Molecular Biology 1999;97:393-414. 20. Rohlich P, Szel A. Photoreceptor cells in the Xenopus Retina. Microscopy Research and Technique 2000;50:327-337. 21. Witkovsky P. Photoreceptor classes and transmission at the photoreceptor synapse in the retina of the clawed frog, Xenopus laevis. Microscopy Research and Technique 2000;50:338-346. 22. Kefalov V, Fu Y, Marsh-Armstrong N, Yau K-W. Role of visual pigment properties in rod and cone phototransduction. Nature 2003;425:526-531. 23. Vought BW, Dukkipatti A, Max M, Knox BE, Birge RR. Photochemistry of the primary event in short-wavelength visual opsins at low temperature. American Chemical Society 1999;38:11287-11297. 24. Gabriel R, Wilhelm M. Structure and function of photoreceptor and second-order cell mosaics in the retina of Xenopus. International Review of Cytology 2001;210:77- 120. 25. Baehr W, Frederick J. Inherited retinal diseases: Vertebrate animal models. Nature 2001;1-9. 26. Weleber RG. Inherited and orphan retinal diseases: Phenotypes, genotypes, and probable treatment groups. Retina, The Journal of Retinal and Vitreous Diseases 2005;24:S4-S7. 27. Sahel J-A. Saving cone cells in hereditary rod diseases: A possible role for rod- derived cone viability factor (RdCVF) therapy. Retina, The Journal of Retinal and Vitreous Diseases 2005;25:S38-S39. 21 28. Cepko CL. Effect of gene expression on cone survival in retinitis pigmentosa. Retina, The Journal of Retinal and Vitreous Diseases 2005;25:S21-S24. 29. Hackam AS, Strom R, Liu D, et al. Identification of gene expression changes associated with the progression of retinal degeneration in the rd1 mouse. Investigative Ophthalmology and Visual Science 2004;45:2929-2942. 30. Punzo C, Cepko C. Cellular responses to photoreceptor death in the rd1 mouse model of retinal degeneration. Investigative 2007. 31. Tanito M, Kaidzu S, Anderson RE. Delayed loss of cone and remaining rod photoreceptor cells due to impairment of choroidal circulation after acute light exposure in rats. Investigative Ophthalmology and Visual Science 2007;48:1864-1872. 32. Chang GQ, Hao Y, Wong F. Apoptosis: final common pathway of photoreceptor death in rd, rds, and rhodopsin mutant mice. Neuron 1993;11:595-605. 33. Ripps H. Cell death in retinitis pigmentosa: Gap junctions and the 'bystander' effect. Experimental Eye Research 2002;74:327-336. 34. Mohand-Said S, Deudon-Combe A, Hicks D, et al. Normal retina releases a diffusible factor stimulating cone survival in the retinal degeneration mouse. Proceedings of the National Academy of Science, USA 1998;95:8357-8362. 35. Frasson M, Picaud S, Leveillard T, et al. Glial cell line-derived neurotrophic factor induces histologic and functional protection of rod photoreceptors in the rd/rd mouse. Investigative Ophthalmology and Visual Science 1999;40:2724-2734. 36. Lenzi L, Coassin M, Lambiase A, Bonini S, Amendola T, Aloe L. Effect of exogenous administration of nerve growth factor in the retina of rats with inherited retinitis pigmentosa. Vision Research 2005;45:1491-1500. 37. Leveillard T, Mohand-Said S, Lorentz O, et al. Identification and characterization of rod-derived cone viability factor. Nature Genetics 2004;36:744-759. 38. Fintz A-C, Audo I, Hicks D, Mohand-Said S, Leveillard T, Sahel J. Partial characterization of retina-derived cone neuroprotection in two culture models of photoreceptor degeneration. Investigative Ophthalmology and Visual Science 2003;44:818-825. 39. Banin E, Cideciyan AV, Alema TS, et al. Retinal Rod Photoreceptor–Specific Gene Mutation Perturbs Cone Pathway Development. Neuron 1999;23:549–557. 22 40. Jones BW, Marc RE. Retinal remodeling during retinal degeneration. Experimental Eye Research 2005;1-15. 41. Oh ECT, Khan N, Novelli E, Khanna H, Strettoi E, Swaroop A. Transformation of cone precursors to functional rod photoreceptors by bZIP transcription factor NRL. Proceedings of the National Academy of Science 2007;104:1679-1684. 42. Shen J, Yang X, Dong A, et al. Oxidative damage is a potential cause of cone cell death in retinitis pigmentosa. Journal of Cellular Physiology 2005;203:457-464. 43. Wu J, Seregard S, Algvere PV. Photochemical damage of the retina. Survey of Ophthalmology 2006;51:461-482. 44. Reidel B, Orisme W, Goldmann T, Smith WC, Wolfrum U. Photoreceptor vitality in organotypic cultures of mature vertebrate retinas validated by light-dependent molecular movements. Vision Research 2006;46:4464-4471. 45. Chadler GJ. Animal models in research on retinal degenerations: Past progress and future hope. Vision Research 2002;42:393-399. 46. Crabtree GR, Schreiber SL. Three-part inventions: intracellular signalling and induced proximity. Trends Biochem Sci 1996;21:418-422. 47. Spencer DM, Wandless TJ, Schreiber SL, Crabtree GR. Controlling signal transduction with synthetic ligands. Science 1993;262:1019-1024. 48. Fan L, Freeman KW, Khan T, Pham E, Spencer DM. Improved death switches based on caspases and FADD. Human Genetic Therapy 1999;10:2273-2285. 49. MacCorkle RA, Freeman KW, Spencer DM. Synthetic activation of caspases: Artificial death switches. Proceedings of the National Academy of Science, USA 1998;95:3655-3660. 50. Mallet VO, Mitchell C, Guidotti J-E, et al. Conditional cell ablation by tight control of caspase-3 dimerization in transgenic mice. Nature Biotechnology 2002;20:1234-1239. 51. Clackson T, Yang W, Rozamus LW, et al. Redesigning an FKBP-ligand interface to generate chemical dimerizers with novel specificity. Proceedings of the National Academy of Science 1998;95:10437-10442. 52. Jin L, Zeng H, Chien S, et al. In vivo selection using a cell-growth switch. Nature Genetics 2000;26:64-66. 53. Snyder SH, Lai MM, Burnett PE. Immunophilins in the nervous system. Neuron 1998;21:283-294. 23 54. Grosskreutz CL, Hanninen V, A., Pantcheva M, B., Huang W, R. PN, Dobberfuhl A, P. FK506 blocks activation of the intrinsic caspase cascade after optic nerve crush. Experimental Eye Research 2005;80:681-686. 55. Yuan J, Horvitz HR. A first insight into the molecular mechanisms of apoptosis. Cell 2004;S116:S53-S56. 56. Strasser A, O'Connor L, Dixit VM. Apoptosis signalling. Annual Reviews in Biochemistry 2000;69:217-245. 57. Hengartner M, O. The biochemistry of apoptosis. Nature 2000;407:770-776. 58. Kam PCA, Ferch NI. Apoptosis: Mechanisms and clinical implications. Anaesthesia 2000;55:1081-1093. 59. Kaufmann SH, Gores GJ. Apoptosis in cancer: causes and cures. Biosessay 2000;22:1007-1017. 60. Santini MT, Rainaldi G, Indovina PL. Apoptosis, cell adhesion and the extracellular matrix in the three-dimentional growth of multicellular tumor spheroids. Critical Reviews in Oncology/Hematology 2000;36:75-87. 61. Nicholson DW. From bench to clinic with apoptosis-based therapeutic agents. Nature 2000;407:810-816. 62. Li P, Nijhawan D, Budihardjo I, et al. Cytochrome c and dATP-dependent formation of Apaf-1/caspase-9 complex initiates an apoptotic protease cascade. Cell 1997;91:479-489. 63. Jiang X, Wang X. Cytochrome c-mediated apoptosis. Annual Review of Biochemistry 2004;73:87-106. 64. Li P, Nijhawan D, Wang X. Mitochondrial activation of apoptosis. Cell 2004;S116:S57-S59. 24 CHAPTER 2. Controlled rod cell ablation in transgenic Xenopus laevis1 2.1 INTRODUCTION In vertebrate retina, response to neuronal loss varies from complete regeneration to severe secondary degeneration, depending on the species and the anatomical and temporal properties of the damage. Known mechanisms of regeneration include proliferation of cells in the ciliary marginal zone (CMZ), differentiation of inner and outer nuclear layer (INL and ONL) progenitors, as well as transdifferentiation of Muller glia, and retinal pigmented epithelium (RPE)65-67, complementing evidence as early as 1880 68 of profound retinal regeneration in fish and urodele amphibians (i.e. newts). Larval anuran amphibians (i.e. frog tadpoles) can regenerate retinal neurons from the CMZ, from RPE transdifferentiation 65, 69, and from proliferating INL cells 70 after retinal damage. Adult urodele amphibians retain these capacities, while anuran regeneration becomes dependant on the CMZ 69 and RPE 71. Mammalian retinas have limited regenerative capacity 72, most of which is lost early in life, rendering patients with retinal degeneration irretrievably vision-deprived. The reasons for these species differences are the subjects of ongoing research. Rod death is an intriguing and clinically significant subset of retinal damage, since in mammals it is not only irreversible, but followed by cone death that causes blindness, a pathology associated with the disorder retinitis pigmentosa (RP). Secondary cone degeneration is an area of intensive research 28-31, 34, 36, 37, 39, 40, 73-77, but modeling this process is difficult because common lab rodents have rod-dominated retinas. Conversely, zebrafish have a cone-dense retina; however, their regenerative ability masks secondary cone death 78. X. laevis have equal numbers of rods and cones, and some evidence suggests that in X. laevis, cone death occurs in the absence of rods 79, while the recently reported capacity for post-metamorphic regeneration suggests this effect may be masked as in zebrafish71. Understanding the responses of the X. laevis retina to rod death would 1 A version of this chapter will be submitted for publication. Hamm, Tam, B.M, Moritz, O.L. 25 improve our understanding of the mechanistic differences that favor regeneration in lower vertebrates versus progressive degeneration in higher vertebrates. To induce rod ablation via a mechanism consistent with RP pathology 32, we modified a system described by Spencer and colleagues 48, 49 to induce rod apoptosis by activation of caspase 9. (A similar system has been applied to hepatocytes 50.) Caspase 9 is an initiator caspase that cleaves like and downstream, effecter caspases. Dimerization initiates cleavage of the inactive form at D315, causing separation into large and small subunits. The active subunits set into motion an apoptotic cascade, leading to cleavage of caspases 3 and 7 and indirectly 6 (for reviews see55-58). In this study, we utilized a transgene encoding a modified, inducible caspase 9 (iCasp9) incorporating binding domains specific for the compound AP20187 (Ariad Pharmaceuticals, Inc., Cambridge, MA)51, which induces its dimerization and auto-activation. We expressed iCasp9 in transgenic X. laevis rods under control of the rod opsin promoter. We subsequently investigated the effects of AP20187 on the retinas of transgenic and non-transgenic animals using histological, western blot and electrophysiological techniques. 2.2 METHODS 2.2.1 Molecular biology The iCasp9 cDNA from pSH1/S-Fvls-p30casp9-E48 given to us by Spencer was sub- cloned into pXOP0.8-eGFP-N1, and the plasmid (XOP0.8-iCasp9-N1) was prepared for transgenesis as previously described 80. 2.2.2 Generation, rearing and AP20187 treatment of transgenic X. laevis We generated transgenic X. laevis using the method of Kroll and Amaya 81, as previously detailed 82. We co-injected XOP0.8-iCasp9-N1 and XOP0.8-eGFP-N1 plasmids (2:1 ratio), and identified transgenic animals by screening for eGFP expression five days post fertilization (dpf). Tadpoles and frogs were reared as described80. AP20187 was added to tadpole medium at a final concentration of 10nM. Adult frogs received 1ul of 10mM 26 AP20187 (in ethanol) per gram body weight injected into the dorsal sac. All procedures were carried out in adherence with The ARVO Statement for the Use of Animals in Ophthalmic and Vision Research, and met UBC animal care committee approval. 2.2.3 Western blots Solubilized eye extracts and western blots were prepared and imaged as previously described 80. Antibody dilutions were 1:1000 for anti-HA (Cedarlane Laboratories Limited, Hornby, ON), 1:10 for anti-Rhodopsin B630N (gift of Dr. P. Hargrave, University of Florida), and 1:10,000 for IRDye800-CW conjugated anti-mouse secondary antibody (Rockland, Gilbertsville, PA). 2.2.4 Microscopy Eyes were fixed and processed for microscopy as previously described80. In addition to the antibodies listed above, we used anti-calbindin (Calbiochem) at 1:250 dilution and Cy3 or Cy5-conjugated anti-mouse and -rabbit secondary antibodies (Jackson Immunoresearch, West Grove, PA) at 1:750 dilutions. TUNEL staining was performed using the ApoTag Red Kit (Chemicon International). Sections were imaged using a Zeiss 510 laser scanning confocal microscope. We counted cells manually, and measured distances using Zeiss software. Quantification was performed on consecutive sections in the region of the optic nerve. The most peripheral region was excluded, as it contains the CMZ with developing photoreceptors. TUNEL-positive nuclei were tallied for entire sections, while cone counts were averaged from five consecutive 150µm samples of each section, and measurements of INL width were averaged from ten measurements per section. 2.2.5 Electroretinography Post-metamorphic animals were dark adapted for 12 hours, anesthetized in 0.05% tricaine methanesulfonate (MS-222 – prevents peripheral, but not central action potentials), and placed on an anesthetic-soaked sponge in front of a UTAS-E 2000 LKC Gansfeld stimulator (LKC Technologies, Geithersburg, MD). The corneal electrode was a pulled glass micropipette filled with 1XMMR, while needle electrodes were used for reference 27 and ground. Stimuli included scotopic flashes at five intensities, 0dB flicker flashes at five frequencies and 0dB flashes superimposed on a lit background (photopic stimuli). The intensity series waveforms were an average of four or five responses, while photopic waveforms present the average of ten responses. Equations used to fit a-wave and intensity data were the Shady et al. 83 modification of the Lamb and Pugh 84 equation: R(I,t)={1-exp[-IS(t-td)2]}Rmp3 and the Naka-Rushton equation: V(I)=Vmax{In /[In+Kn]}85 respectively. 2.3 RESULTS 2.3.1 AP20187-induced rod apoptosis in primary transgenic tadpoles Fan and colleagues developed artificial death switches based on membrane-bound or soluble forms of caspases 48. A Caspase 9-based switch was selected over other forms, as apoptosis was more dramatic in AP20187 treated HEK293 cell transfections and preliminary X. laevis transgenesis experiments.We generated double-transgenic X. laevis expressing both iCasp9 and eGFP under the control of the X. laevis rod opsin promoter 86. This dual plasmid method allowed non-invasive identification of iCasp9-positive animals due to the high frequency of transgene co-integration 81, 87. At 14 dpf we exposed eGFP- positive tadpoles to AP20187 for 48 hours, after which the tadpoles were sacrificed. Because the iCasp9 construct incorporates an HA epitope tag that is degraded upon caspase activation, visualization and quantification of inactive iCasp9 was possible. A western blot analysis of untreated eGFP-positive transgenic tadpoles showed that detectable levels of intact iCasp9 were expressed in six of seven eGFP-positive eyes, although expression levels varied (Figure 2.1A). Therefore, eGFP expression was a reliable indicator of iCasp9 expression. No anti-HA signal was detectable in a western blot analysis of primary transgenic tadpoles exposed to AP20187 for two days (p=0.039, Mann-Whitney test, n=7 per group, Figure 2.1B), indicating no intact caspase, as anticipated for drug-induced iCasp9 activation. Rhodopsin (a measure of rod viability as it is abundant and expressed exclusively in rods) was not significantly reduced at this time point (p=0.528, Mann-Whitney test, n=7 per group, not shown). 28 Immunofluorescence microscopy of contralateral eyes confirmed rod-specific expression of intact iCasp9 and eGFP, co-expressed in most cells, as well as normal rod morphology in untreated retinas despite iCasp9 expression (Figure 2.1C). Microscopy also showed ongoing rod death in treated retinas despite high rhodopsin levels, as classic apoptotic morphology was observed including blebbing and pyknotic nuclei (Figure 2.1D). Wild- type tadpole rods appeared normal in both treated and untreated groups (not shown). Together these results demonstrate that iCasp9 was expressed in the rods of eGFP- positive primary transgenic X. laevis tadpoles, and that AP20187 was absorbed from the medium, transported to the retina, and interacted with iCasp9 to initiate apoptosis. 2.3.2 ICasp9 did not cause rod death in the absence of AP20187 Although we did not observe evidence of abnormal apoptosis in the absence of AP20187 in 14 dpf tadpoles, it is possible that high-level iCasp9 expression could eventually cause cell death via spontaneous auto-activation. To examine the long-term effects of iCasp9 expression, three eGFP-positive post-metamorphic primary transgenic frogs were sacrificed at three months post-fertilization. Rods from all three were healthy (similar to wild-type animals, not shown) and were anti-HA reactive. To corroborate this data, we used ERGs to measure function in two eGFP-negative, two eGFP-positive, and ten age- matched wild-type frogs. Retinal function, measured by a scotopic intensity series, was comparable to controls (eGFP-positive, and negative controls shown in Figure 2.2 in black, wildtype controls not shown). In conjunction with histology, the strong ERG response suggested abundant and healthy rods and therefore negligible drug-independent activation of iCasp9. 2.3.3 AP20187 severely compromised retinal function in primary transgenic post- metamorphic frogs After an initial ERG measurement, the two eGFP-positive and two eGFP-negative frogs, as well as four wild-type frogs received subcutaneous injections of AP20187, and were re-tested five days post drug administration. The pre- and post-drug waveforms (Figure 2.2A) demonstrate a dramatic reduction in scotopic ERG response in eGFP-positive animals injected with AP20187. Both log Rmp3, and b-wave amplitude measures were 29 virtually unchanged in eGFP-negative animals, yet dramatically reduced in eGFP- positive transgenics (Figure 2.2B). We fit trough-to-peak amplitudes at increasing stimulus intensity to the Naka-Rushton equation (Figure 2.2C), and calculated 95% confidence intervals from the same equation for ten untreated wild-type animals (superimposed on Figure 2.2C). Before drug administration, all four animals fell within the confidence intervals (again, indicating minimal drug-independent iCasp9 activation), and after drug injection, animals 1, 3 and 4 fell outside. From the Naka-Rushton fit, we derived Vmax (projected maximal amplitude), and Km (intensity required for a half- maximal response). In animals 1 and 3, Vmax dropped dramatically (rendering Km an insignificant parameter), suggesting almost complete destruction of rods, as well as INL abnormalities. From the dramatic changes in Rmp3, b-wave amplitude, and Vmax in animals 1 and 3, we are confident that AP20187 induced rod death, as demonstrated in our initial experiments on tadpoles. In animal 4, Km increased considerably, yet Vmax was only moderately affected, suggesting mild rod dysfunction. This could be caused by expression of the iCasp9 in only a subset of rods, and is also influenced by the fact that animal 4 had the smallest initial ERG amplitude. Animal 2 was not effected by AP20187, as we would expect from an eGFP-negative primary transgenic animal. 2.3.4 Consistent iCasp9 expression in F1 generation tadpoles yields more reproducible induction of apoptosis We analyzed the progeny of the first animal to reach sexual maturity, and found consistent iCasp9 expression in eGFP-positive tadpoles which was depleted after seven days of AP20187 exposure (p= 0.0088, Mann-Whitney, n=7 per group, Figure 2.3A). This was associated with a significant reduction of rhodopsin levels (p=0.0004, Mann- Whitney, n=7 per group, Figure 2.3B), as well as rod clearance from the ONL (not shown). Since expression was more consistent in this F1 generation, more advanced analysis was possible. To generate a time-line of events from drug administration to rod clearance, we analyzed F1 tadpoles exposed to AP20187 at times ranging from one hour to four days. ICasp9 and rhodopsin expression were quantified for all five animals in each group from western and 30 dot blots, and the number of apoptotic nuclei were counted for the first three animals in each group from TUNEL-labeled sections; results are summarized in Figure 2.4A. ICasp9 and rhodopsin expression increased for 12 hours after drug exposure due to tadpole growth. At 12 hours, the first TUNEL-positive nuclei were apparent in the ONL, and iCasp9 and rhodopsin levels began to decline. Two days after drug exposure, HA signal was undetectable, and the number of TUNEL-positive nuclei in the ONL peaked. Rhodopsin levels continued to decrease as the number of TUNEL-positive nuclei declined. In contrast to treated transgenic animals, no TUNEL-positive nuclei were found in the ONL of untreated transgenic animals (Figure 2.4B), or treated control animals (data not shown). This is consistent with the morphological signs of apoptosis seen in Figure 2.1 after two days of AP20187 exposure in primary transgenic animals, the compromised retinal function measured five days after drug injection in Figure 2.2, and the reduction in rhodopsin seen in Figure 2.3 after seven days exposure in F1 tadpoles. TUNEL-positive nuclei were also apparent in other layers of the retina. As apoptosis is a normal part of development, we corrected for the small number of TUNEL-postive nuclei found in each retinal layer in untreated transgenic control animals (an average of 0.6 per section). Two interesting trends emerged: First, TUNEL-positive nuclei were more abundant in the RPE of treated transgenic animals, closely following trends in the ONL, and likely representing phagocytosed rods. Second, the RPE, INL and ganglion cell layer (GCL) of treated control tadpoles appeared to be spared of natural apoptosis one hour after AP20187 exposure. However, these trends were not statistically significant. 2.3.5 Longitudinal studies on post-metamorphic frogs confirm rod death and suggest cone cell changes We monitored the AP20187-treated primary transgenic post-metamorphic animals previously described, and four treated wild-type X. laevis frogs, for eight months. To counteract addition of healthy rods at the CMZ due to growth of the eye, or possible rod regeneration, we injected each animal with AP20187 monthly. We recorded the ERG response to 10Hz flicker and photopic stimuli as a measure of cone function. Figure 2.5 summarizes the photopic waveforms generated by transgenic animals compared to 31 averaged non-trasngenic AP20187-treated frogs. Control animals had consistent responses over time, while eGFP-positive transgenic animals had compromised and irregular cone function within three months of the intial AP20198 injection, which subsequently improved despite continual rod ablation. Two days after the five month AP20187 injection, we sectioned one eye from the eGFP- negative animal 4 (Figure 2.6A, also in Figures 2.2 and 2.5) and the eGFP-positive animal 3 (Figure 2.6B, also in Figures 2.2 and 2.5). We observed healthy rod outer segments in animal 4 (Figure 2.6A), and profound rod ablation in animal 3 (Figure 2.6B). We also observed some dysmorphic rods in the central retina in animal 3 (Figure 2.6B, arrowheads), suggesting regeneration followed by apoptosis upon repeated drug administration. This is consistent with the fragmented cell bodies apparent two days after drug administration in the tadpoles shown in Figure 2.1D. Cones appeared healthy in the presence or absence of rods (Figure 2.6C and D); in fact, the density of cones was higher in transgenic animal 3 than in non-transgenic controls (Panel E). We also noticed previously undetected faint eGFP expression in some rods of animal 4, and two small patches of rod ablation (Figure 2.6F). This is consistent with the decreased sensitivity (increased Km) for animal 4 (Figure 2.2C), suggesting limited and mosaic expression of iCasp9. We therefore separated rod-rich (labeled 4) and rod-less (labeled 4*) portions of the retina for this analysis. The width of the INL, and of the total retina, differed between the rod-less animal and controls. Panels A and B in Figure 2.6 demonstrate that the INL was 30% thicker in animal 3 (B) than 4 (A). Although the nuclei were spaced further apart in some rod-ablated areas, INL thickening was still associated with an overall increase in INL somas as the density was not significantly different between animal 3 and control animals. Interestingly, in the rod-less regions of animal 4’s retina, the INL was thickened relative to rod-rich areas (Panels F), and was similar to the INL of rod-ablated animal 3 (Panel G). These rod-less regions did not have increased cone counts as was the case in rod-less animal 3 (Panel E). These data confirm the complete absence of rods in AP20187-treated transgenic frogs, and suggest an associated cone-derived functional irregularity. However, these data also show 32 improvement in cone function after three months of abnormality, and the presence and proliferation of cones five months after the initial AP20187 exposure despite continued rod ablation. Together with changes observed in the INL, theses observations are suggestive of retinal remodeling after rod death. 2.4 DISCUSSION We have successfully co-expressed iCasp9 and eGFP in X. laevis rod photoreceptors, and identified animals expressing iCasp9 by eGFP expression. Furthermore, we were able to induce rapid, controlled apoptosis of rods by administering AP20187 to these animals. In the absence of AP20187, the transgene products were non-toxic and had no significant effect on retinal physiology or morphology. AP20187 was successfully administered via epithelial absorption in tadpoles, and subcutaneous injection in post-metamorphic frogs. The drug crossed the blood-retinal barrier and reached photoreceptors. This was not unexpected, given that the structurally similar compound FK506 crosses the blood brain barrier 88. AP20187 had little effect in the absence of iCasp9, limited to a possible anti-apoptotic effect noted in F1 tadpoles. As a similar anti-apoptotic effect in ganglion cells has been demonstrated for FK506 54, 89, AP20187 suppression of apoptosis is feasible. More surprising is the timing of this effect; although the rapid reduction in apoptosis in the RPE, INL and GCL suggests that the drug reaches the retina quickly, a 12 hour latency exists between drug administration and iCasp9 activation. This may be a concentration-dependent effect, with small quantities available early after exposure to the drug inhibiting normal apoptosis, and larger quantities promoting iCasp9 activation. In the presence of sufficient AP20187, auto-activation of iCasp9 leading to apoptosis was robust, as previously demonstrated in studies carried out by Spencer and colleagues 48-50. We established the efficiency of this system directed towards X. laevis rods by western blots, immunohistochemistry, and eletroretinography. We demonstrated depletion of 33 intact caspase and rhodopsin, and observed fragmenting anti-rhodopsin reactive cell bodies, as well as pyknotic and TUNEL-positive nuclei in the ONL, all hallmarks of rod apoptosis. This was corroborated functionally by severely compromised scotopic ERGs after AP20187 treatment in post metamorphic frogs. ICasp9 activation occurred in less than a day, and rod apoptosis was apparent from two to four days after drug interaction. This was consistent in primary transgenics, F1 generation, and post-metamorphic X. laevis. Therefore we are confident that it is possible to induce controlled rod apoptosis at any point in X. laevis development. We can use this system to model human pathologies where apoptotic rod death is the primary event. RP is the most common cause of inherited blindness 74, and the common path of rod death, despite over 160 different initiating mutations (http://www.sph.uth.tmc.edu/Retnet/), is apoptosis 32. After inducing post-metamorphic rod ablation in this model, we noticed a decrease in amplitude of photopic and flicker ERGs – indicating rod death was affecting cone function and/or viability, as commonly observed in human RP pathology. Understanding this interaction may help us understand the mechanisms underlining secondary cone death causing blindness in RP. Several non- mutually exclusive theories have been proposed to explain this phenomenon, including direct interaction 33, indirect toxicity 29, 42, lack of protection from secreted factors 36, 37, or functional loss through ectopic synaptogenesis and retinal reorganization 39, 77. The limited longitudinal data we have collected is in favor of the last hypothesis, since alterations in INL are the only overt abnormalities two months after reduction in cone function. However, we also noted a reinstatement of photopic responses five months after initial rod ablation, and increased numbers of cones, indicating functional restoration or regeneration. In addition to normal growth from the periphery, we noticed blebbing cell bodies immuno-positive for anti-rhodopsin in the central retina two days after AP20187 administration, after five months of continued rod cell ablation. This suggests that without continued AP20187 adminstration, central rods may regenerate. Currently, proliferation from the CMZ and transdifferentiation of the RPE are generally accepted as 34 the only methods of regeneration in post-metamorphic X laevis 69, 71. Given the INL thickening in conjunction with the presence of new photoreceptors, it seems plausible that regeneration could initiate from progenitor cells residing in this region, just as in adult newts 90 and X. laevis larvae 70 after retinal damage. Thickening of the INL (as a sign of reorganization and/or regeneration in response to rod death), and loss of b-wave amplitude in scotopic and photopic ERGs, could reflect changes in Muller glia. These cells are associated with the ERG b-wave, have somas located in the INL, and can transdifferentiate into progenitor cells capable of replacing retinal cells 91. It will be interesting to further probe the involvement of Muller glia in events following rod death. For the majority of the experiments described, we used animals expressing uniform high levels of transgene products, and measured robust effects; however, mosaic expression patterns commonly occur in our system, and could be used to examine the subtleties of regional versus complete ablation, as demonstrated for animal 4 where rod-rich and rod- less regions can be compared in the same animal. In future studies involving primary transgenic animals, this property will be used to examine implications of partial rod ablation. In summary, we have developed a transgenic X. laevis model of RP, in which rod apoptotsis is initiated by administration of a normally innocuous compound. This experimental model is distinguished from other inducible models (such as light damage) in that the cell death pathway is well characterized, and rod photoreceptors are the only cell affected by the primary insult. Thus, these animals provide an excellent system for examining the effects of rod apoptosis on other retinal cell types at any stage of development. Rod-cone interactions are of particular interest, due to the high rod to cone ratio in the X. laevis retina (1:1) and the involvement of secondary cone degeneration in human RP. Furthermore, because AP20187 can be administered either acutely or chronically, it will be possible to examine the regenerative capacity of the developing and mature X. laevis retina. 35 2.5 FIGURES Figure 2.1. Induced rod cell ablation in primary transgenic tadpoles. Western blot and immunohistochemical assays of primary transgenic tadpoles, either untreated, or treated with AP20187 for two days. A) Western blot analysis shows variable expression of intact iCasp9 (anti-HA signal from epitope on caspase prodomain) in six of seven eGFP-positive untreated animals. B) In tadpoles exposed to AP20187 for two days the anti-HA signal was absent, consistent with degradation of the epitope following auto- activation. C) Microscopy confirms anti-HA labeling (red) co-localized with GFP (green) exclusively in rod cells, and that these cells retain healthy morphology in the absence of AP20187. D) Treated retinas were depleted of intact iCasp9, and display morphological hallmarks of apoptosis, including loss of structural integrity, cell blebbing (arrows) and pyknotic nuclei (arrow heads), indicating that activation of iCasp9 was associated with rod cell death. Scale bar = 20µm. 36 Figure 2.2. AP20187 injection caused compromised retinal function in primary transgenic post-metamorphic frogs. Scotopic ERG recordings from before (black) and after (gray) AP20187 injection in two eGFP-positive (1 and 3) and two eGFP negative (2 and 4) frogs. A) Raw waveforms at increasing stimulus intensity superimposed for each animal show that scotopic retinal function was normal in untreated animals, but was severely compromised in eGFP-positive frogs after drug injection. B) Projected a-wave asymptote (log Rmp3), and b- wave amplitude for the maximal recorded responses confirm photoreceptor and resulting inner nuclear layer (INL) functional abnormalities. C) The amplitude of response to each of five stimulus intensities was fit to the Naka- Rushton equation. The fit from ten wild type frogs was averaged (individual fits in inset), and the 95% confidence intervals derived from these animals were superimposed on the graph. All untreated animals fell within the 95% confidence interval, but when treated with AP20187 animals 1 and 3 did not. 37 Figure 2.3. Inducible apoptosis was more consistent in the F1 generation. Western blot assays of F1 transgenic tadpoles, either untreated, or treated with AP20187 for seven days. A) Untreated eGFP-positive tadpoles showed more consistent iCasp9 expression (upper), and elimination of HA signal seven days after drug administration (lower). B) Rhodopsin levels (B630N signal) were high and consistent in untreated animals (upper), and severely reduced in treated tadpoles (lower). A control sample (untreated primary transgenic retina) was included on each blot (asterisks). 38 Figure 2.4. Rod cell apoptosis was complete within four days of AP20187 administration. Signal intensity from western blots probed for intact iCasp9 (anti-HA signal) and rhodopsin (B630N signal), as well as counts of TUNEL-positive nuclei from retinal sections graphed for eight time points after AP20187 administration in F1 tadpoles. Each time point represents an averaged, normalized value. A) ICasp9 and rhodopsin levels increased up to 12 hours after drug exposure (n=5). At the 12 hour time point the first TUNEL –positive nuclei were apparent in the photoreceptor layer (n=3), and subsequently the HA epitope of iCasp9 was progressively degraded, and rhodopsin levels progressively decreased. The number of TUNEL-positive cells peaked at day two (average of six nuclei per section) and subsequently declined. B) Numerous TUNEL- positive cells were observed in the ONL of treated transgenic animals 36 hours after AP20187 administration, but absent from the ONL of untreated transgenic animals. Scale bar = 20µm 39 Figure 2.5. Photopic responses were compromised after AP20187 injection, but showed subsequent improvement. The graphs summarize the responses to photopic stimuli over an eight month period. Animals 1 and 3 are the same as described in Figure 2, and control waveforms (two eGFP-negative and five wild type frogs) were averaged together for each month. A) Responses to photopic stimuli were irregular three and four months after the initial injection, but subsequently improved. Animal 3 was sacrificed after the five month ERG, and morphological data from this animal is shown in Figure 2.6. 40 Figure caption on next page. 41 Figure 2.6. Five months after initial AP20187 injection, rod cells were ablated, the INL was thicker, and cone cells were intact and more abundant in eGFP-positive animal 3 than in eGFP-negative animal 4. Animals 3 and 4 were sacrificed two days after the five month AP20187 injection. “4*” identifies regions of animal 4’s retina that are rod-deprived. A) Animal 4 had healthy rod outer segments (ROSs) visualized by B630N labeling (red), and a 22µm thick INL (±3.1), measured from boundaries of Hoechst3392 labeled nuclei (blue). B) Animal 3 had ablated rod cells (absence of B630N labeled outer segments), except for centrally located dysmorphic rods (arrow heads), as well as a thickened INL (34.9 ±4.2µm). C) Retinal sections from animal 4 labeled with anti-calbindin displayed normal cone cells. D) Retinal sections from animal 3 labeled with anti-calbindin also displayed healthy cones despite functional irregularity shown in Figure 2.5, and lack of intact rod cells displayed in Panel B. E) Quantification of cone cells in untreated and AP20187 treated controls, as well as animal 3, and the rod-rich and rod-less portions of animal 4, show that the density of cone cells was greater in rod-less animal 3, but not in the small rod-less portions of otherwise normal animal 4. F) Region of retina from animal 4 showing two regions of rod-deprived retina (rod -) flanking a relatively normal rod-rich region (rod +). The panel shows the Hoechst dye channel only, but rod outer segments (ROS) are also visible due to autofluoresence. In the rod- deprived regions there is a local densification of INL nuclei as seen in the retina of animal 3, but the INL is of normal thickness in the intervening rod-rich region. G) The INL was thicker in both the rod-ablated retina of animal 3 and the small rod-ablated portions of animal 4. Rod ablation was associated with thickening of the INL, and proliferation of cone cells. INL thickening appeared to be a local, possibly direct effect of rod ablation. Scale bars = 100µm in A and B, 20µm in C and D, and 50µm in F. Error bars are +/- S.E.M. 42 2.6 REFERENCES 1. Del Rio-Tsonis K, Tsonis PA. Eye regeneration at the molecular age. Developmental Dynamics 2003;226:211-224. 2. Tsonis PA, Del Rio-Tsonis K. Lens and retina regeneration: transdifferentiation, stem cells and clinical applications. Experimental Eye Research 2004;78:161-172. 3. Reh TA, Levine RL. Multipotent stem cells and progenitors in the vertebrate retina. Journal of Neurobiology 1998;36:206-220. 4. Philipeaux JM. Note sur production de l-oil chez la salamandre aquatique. Gaz Med France 1880;51:453-457. 5. Mitashov VI. Retinal regeneration in amphibians. The International Journal of Developmental Biology 1997;41:893-905. 6. Levine RL. La regenerescence de la retine chez Xenopus laevis. Dev Can Biol 1981;40:19-27. 7. Yoshii C, Ueda Y, Okamoto M, Araki M. Neural retinal regeneration in the anuran amphibian Xenopus laevis post-metamorphosis: Transdifferentiation of retinal pigmented epithelium regenerates the neural retina. Developmental Biology 2006;1-12. 8. Kubota R, Hokoc JN, Moshiri A, McGuire C, Reh TA. A comparative study of neurogenesis in the retinal ciliary marginal zone of homeothermic vertebrates. Developmental Brain Research 2002;134:31-41. 9. Punzo C, Cepko C. Cellular responses to photoreceptor death in the rd1 mouse model of retinal degeneration. Investigative 2007. 10. Gargini C, Terzibasi E, Mazzoni F, Strettoi E. Retinal organization in the retinal degeneration 10 (rd10) mutant mouse: A morphological and ERG study. The Journal of Comparative Neurology 2007;500:222-238. 11. Mohand-Said S, Hicks D, Leveillard T, Picaud S, Porto F, Sahel JA. Rod-cone interactions: Developmental and clinical significance. Progress in retinal and eye research 2001;20:451-467. 12. Banin E, Cideciyan AV, Alema TS, et al. Retinal Rod Photoreceptor–Specific Gene Mutation Perturbs Cone Pathway Development. Neuron 1999;23:549–557. 13. Cepko CL. Effect of gene expression on cone survival in retinitis pigmentosa. Retina, The Journal of Retinal and Vitreous Diseases 2005;25:S21-S24. 43 14. Delyfer M-N, Simonutti M, Neveux N, Leveillard T, Sahel J-A. Does GDNF exert its neuroprotective effects on photoreceptors in the rd1 retina through the glial glutamate transporter GLAST? Molecular Vision 2005;11:677-687. 15. Delyfer M-N, Forster V, Neveux N, Picaud S, Leveillard T, Sahal J-A. Evidence for glutamate-mediated excitotoxic mechanisms during photoreceptor degeneration in the rd1 mouse. Molecular Vision 2005;11:688-696. 16. Leveillard T, Mohand-Said S, Lorentz O, et al. Identification and characterization of rod-derived cone viability factor. Nature Genetics 2004;36:744-759. 17. Hackam AS, Strom R, Liu D, et al. Identification of gene expression changes associated with the progression of retinal degeneration in the rd1 mouse. Investigative Ophthalmology and Visual Science 2004;45:2929-2942. 18. Mohand-Said S, Deudon-Combe A, Hicks D, et al. Normal retina releases a diffusible factor stimulating cone survival in the retinal degeneration mouse. Proceedings of the National Academy of Science, USA 1998;95:8357-8362. 19. Lenzi L, Coassin M, Lambiase A, Bonini S, Amendola T, Aloe L. Effect of exogenous administration of nerve growth factor in the retina of rats with inherited retinitis pigmentosa. Vision Research 2005;45:1491-1500. 20. Jones BW, Marc RE. Retinal remodeling during retinal degeneration. Experimental Eye Research 2005;1-15. 21. Jones BW, Watt CB, Marc RE. Retinal remodelling. Clinical and Experimental Optometry 2005;88:282-291. 22. Tanito M, Kaidzu S, Anderson RE. Delayed loss of cone and remaining rod photoreceptor cells due to impairment of choroidal circulation after acute light exposure in rats. Investigative Ophthalmology and Visual Science 2007;48:1864-1872. 23. Morris AC, Schroeter EH, Bilotta J, Wong ROL, Fadool JM. Cone survival despite rod degeneration in XOPS-mCFP transgenic zebrafish. Investigative Ophthalmology and Visual Science 2005;46:4762-4771. 24. Moritz OL, Tam BM, Hurd LL, Peranen J, Deretic D, Papermaster DS. Mutant rab8 impairs docking and fusion of rhodopsin-bearing post-golgi membranes and causes cell death of transgenic Xenopus rods. Molecular Biology of the Cell 2001;12:2341-2351. 25. Chang GQ, Hao Y, Wong F. Apoptosis: final common pathway of photoreceptor death in rd, rds, and rhodopsin mutant mice. Neuron 1993;11:595-605. 44 26. MacCorkle RA, Freeman KW, Spencer DM. Synthetic activation of caspases: Artificial death switches. Proceedings of the National Academy of Science, USA 1998;95:3655-3660. 27. Fan L, Freeman KW, Khan T, Pham E, Spencer DM. Improved death switches based on caspases and FADD. Human Genetic Therapy 1999;10:2273-2285. 28. Mallet VO, Mitchell C, Guidotti J-E, et al. Conditional cell ablation by tight control of caspase-3 dimerization in transgenic mice. Nature Biotechnology 2002;20:1234-1239. 29. Strasser A, O'Connor L, Dixit VM. Apoptosis signalling. Annual Reviews in Biochemistry 2000;69:217-245. 30. Kam PCA, Ferch NI. Apoptosis: Mechanisms and clinical implications. Anaesthesia 2000;55:1081-1093. 31. Hengartner M, O. The biochemistry of apoptosis. Nature 2000;407:770-776. 32. Yuan J, Horvitz HR. A first insight into the molecular mechanisms of apoptosis. Cell 2004;S116:S53-S56. 33. Clackson T, Yang W, Rozamus LW, et al. Redesigning an FKBP-ligand interface to generate chemical dimerizers with novel specificity. Proceedings of the National Academy of Science 1998;95:10437-10442. 34. Tam BM, Xie G, Oprian DD, Moritz OL. Mislocalized rhodopsin does not require activation to cause retinal degeneration and neurite outgrowth in Xenopus laevis. The Journal of Neuroscience 2006;26:203-209. 35. Kroll KL, Amaya E. Transgenic Xenopus embryos from sperm nuclear transplantations reveal FGF signaling requirements during gastrulation. Development 1996;122:3173-3183. 36. Moritz OL, Tam BM, Knox BE, Papermaster DS. Fluorescent photoreceptors of transgenic Xenopus laevis imaged in vivo by two microscopy techniques. Invest Ophthalmol Vis Sci 1999;40:3276-3280. 37. Shady S, Hood DC, Birch DG. Rod phototransduction in retinitis pigmentosa. Distinguishing alternative mechanisms of degeneration. Investigative Ophthalmology and Visual Science 1995;36:1027-1037. 38. Lamb TD, Pugh ENJ. A quantitative account of the activation steps involved in phototransduction in amphibian photoreceptors. Journal of Physiology\\ 1992;499:719- 758. 45 39. Naka K, Rushton WH. The generation and spread of S-potentials in fish (Cyprinidae). Journal of Physiology 1970;192:437-461. 40. Batni S, Scalzetti L, Moody SA, Knox B, E. Characterization of the Xenopus rhodopsin gene. The Journal of Biological Chemistry 1996;271:3179-3186. 41. Hartley KO, Hardcastle Z, Fridy RV, Amaya E, Papalopulu N. Transgenic Xenopus embryos reveal that anterior neural development requires continued suppression of BMP signaling after gastrulation. Developmental Biology 2001;238:168-184. 42. Synder SH, Lai MM, Burnett PE. Immunophilins in the nervous systerm. Neuron 1998;21:283-294. 43. Freeman E, Grosskreutz CL. The effects of FK506 on retinal ganglion cells after optic nerve crush. Investigative Ophthalmology and Visual Science 2000;41:1111-1115. 44. Grosskreutz CL, Hanninen V, A., Pantcheva M, B., Huang W, R. PN, Dobberfuhl A, P. FK506 blocks activation of the intrinsic caspase cascade after optic nerve crush. Experimental Eye Research 2005;80:681-686. 45. Ripps H. Cell death in retinitis pigmentosa: Gap junctions and the 'bystander' effect. Experimental Eye Research 2002;74:327-336. 46. Shen J, Yang X, Dong A, et al. Oxidative damage is a potential cause of cone cell death in retinitis pigmentosa. 2005;203:457-464. 47. Grigorian EN, Ivanova IP, Poplinskaia VA. The discovery of new internal sources of neural retinal regeneration after its detachment in newts. Morphological and quantitative research. Izv Akad Nauk Ser Biol 1996;3:319-332. 48. Fisher AJ, Reh TA. Potential of muller glia to become neurogenic retinal progenitor cells. Glia 2003;43:70-76. 46 CHAPTER 3. Electroretinography of Xenopus laevis2 3.1 INTRODUCTION The first electroretinogram (ERG) was recorded from a frog in 1865 by Holmgren. Five years later, Dewar and McKendrick proved this response originated in the retina. Largely due to advancements in physics and neuroscience92, by the 1940’s ERGs were used in clinics. However, the connection between specific cell types within the retina and parameters of the various waves measured in the ERG continued to be the subject of debate. Since this time, this connection has been analyzed in many species, including frogs 93, 94, using many types of stimuli. In the face of this diversity, the International Society for Clinical Electrophysiology of Vision (ISCEV) was established in 1989 in an effort to standardize clinical ERG procedures. The standards are updated approximately every four years, with the most recent publication in 2004. Today full field flash ERGs are commonly used in clinical, veterinary, and research settings. Within research, publications can be divided into the analysis of ERG data, and application to retinal function. The former has been piloted by Hood, Birch, Shadey, Lamb, and Pugh83, 95-98. This subset of the field seeks to correlate full field recordings to the biochemical events in retinal cells, and find appropriate ways to apply these findings. This is complimented by single cell electrophysiology in- and ex- vivo, as well as tissue culture, and in vitro studies of the biochemical pathways of phototransduction and retinal circuitry. The latter refers to the application of these findings to clinical and veterinary settings, as well as use in animals models of various human diseases. In animal models, ERGs are used to measure the functional damage the given alteration elicits, and to assay the extent 2 A version of this chapter will be submitted for publication. Hamm, L.M., Li, G., Moritz, O.L. 47 to which therapeutic interventions restore function. Mice have dominated genetic research in the last two decades, and retinal diseases have been modeled in these animals extensively. Therefore after the initial studies on amphibians, techniques and analyses have been refined predominantly for mammalian retinas. Although a shift occurred from the original ERGs conducted on frogs toward mammals, frogs are again surfacing as a species useful for retinal research. Since the mid nineties, it has been possible to manipulate the DNA of Xenopus laevis18, and related amphibians. Xenopus laevis also have large-diameter photoreceptor cells, particularly useful for studying retinal disease pathologies, single cell electrophysiological investigations, and developing therapeutic interventions. It is therefore valuable to refine techniques and analyses for X. laevis ERGs to make them a relevant tool for investigation in this species in light of what has been learned from mammalian ERGs in the interim. 3.2 METHODS 3.2.1 Generation and rearing Frogs were generated in our facilities, reared in 12 hour light/dark cycles, and fed three times a week. All procedures were carried out in adherence with The ARVO Statement for the Use of Animals in Ophthalmic and Vision Research, and met UBC animal care committee approval. 3.2.2 Animal preparation The night before each set of tests, animals were fed99, and dark adapted overnight. Frogs ranging from one to ten grams were anaesthetized by submergence in 0.05% tricaine methanesulfonate (MS-222 – prevents peripheral, but not central action potentials) for two to ten minutes depending on weight (heavier frogs swam in 0.1%). When no longer responsive to tactile stimuli, frogs were removed from the solution, and placed on a tricaine-soaked sponge. The head and eyes were dried with paper towel. The animal was place in front of the Ganzfeld stimulating dome, and electrodes positioned on the eye, 48 perpendicular to the surface. As X. laevis do not have eyelids, corneal contact is simplified. 3.2.3 Attenuation and amplification The amount of current generated by the X. laevis eye is quite small and prone to attenuation before detection. In our hands, connecting electrodes directly from the eye to detection device yielded only noise. However, with the use of a headstage amplifier, we were able to match the impedance, by providing the power required for the signal to be relayed with fidelity. The resulting amplification was compensated for with a voltage divider circuit. Signals of known magnitude that either circumvented or traveled through the introduced amplifiers and voltage divider circuit were equal in magnitude, verifying that the calculations used in circuit design were correct. (Figure 3.1) 3.2.4 Stimuli According to the ISCEV, a standard flash is between 1.5 and 3.0cd.s/m2. We set 2.5cd.s/m2 as 0, and measured changes in intensity on a log scale from this point. 3.2.5 Temperature regulation Frogs are poikilotherms, and therefore adapt to the environmental temperature. X. laevis are healthiest at approximately 18 degrees Celcius. Mammals, on the other hand, maintain a 37 degree core temperature. Temperature is known to affect ERGs, with most enzymatic reactions occurring at a higher rate with increased temperature. The same is true for X. laevis. We increased temperature from 12 to 24 degrees Celcius in two degree increments, and measured response to a standard flash. As temperature increased, a-wave amplitude did not change in a consistent way, but the b-wave amplitude increased, and implicit time of both a- and b- waves decreased, indicating faster kinetics. For the remainder of the experiments, temperature was maintained constant at 24 degrees, unless otherwise specified, to maximize the amplitude of response, yet maintain healthy core temperature. 49 3.3 RESULTS 3.3.1 Summary of the X. laevis ERG The major components of an ERG are the a- wave, b- wave, x- wave and oscillatory potentials. The initial slope of the a- wave mimics the response of an individual rod to light, and can be used as a direct measurement of photoreceptor activity (parameters of this slope can be calculated with the Hood and Birch95 equation R(I,t)={1-exp[-IS(t- td)2]}Rmp383, 96). The b- wave is derived from the inner nuclear layer (INL) cells that amplify and transmit the photoreceptor response to ganglion cells. It is useful to look at b-wave amplitude at increasing stimulus intensity. Naka and Rushton used a Michealis Menton-type equation to plot this relationship, which has become a common analytical tool. Vmax, and Km can be derived from this equation, which tell us about the overall health and sensitivity of the retina. X- waves (positive deflection in the rising phase of the b- wave) are apparent in humans100, 101 and other species with abundant cones. Oscillatory potentials (usually seven in total102) are superimposed on the b- wave in response to medium or high intensity stimuli. These are thought to be derived from INL feedback systems associated with inhibitory amacrine cells. In X. laevis, we have been able to measure each of these components. Overall amplitude is lower than that measured in mammalian eyes, and implicit times are longer. We can measure X. laevis x- waves in dark adapted animals upon photopic stimulation, but not in response to long wavelength light as is the case in human recordings. In mesopic conditions we measure up to seven oscillatory potentials, as is expected from other species. Other than the differences mentioned, in adult frogs we measure similar responses to mammalian scotopic intensity series and photopic ERGs (Figure 3.2). Similar ERGs have been recorded in X. laevis tadpoles, using slightly different methods and stimulation103. We recorded ERGs from twelve animals at various stages between 54 and 6617, six pre-metamorphic, and six post-metamorphic. In our hands, tadpoles generated smaller responses than frogs, which gradually improved throughout metamorphosis. Recordings were measured without cutting the pellucida, as previously 50 suggested to gain access to cornea in animals staged 59-60 17, 103. Animals in all developmental stages recovered, and remained healthy. This is important for use in longitudinal studies, and in cases when animals are genetically unique or limited in number. We sacrificed all twelve animals to examine the health, and length of photoreceptor outer segments, as some previous investigations have suggested the possibility of morphological changes in photoreceptors during metamorphosis. We found healthy outer segments throughout metamorphosis and into adulthood. 3.3.2 Isolating rod and cone responses It is possible to separate rod and cone derived responses by exploiting differences in the kinetics of phototransduction. We replicated methods of separating rods and cones published from human or mice experiments to determine the stimuli eliciting individual photoreceptor type responses in X. laevis. The ‘paired flash paradigm’ is a test in which two flashes (a ‘test’ and a ‘probe’), are separated by an inter-stimulus interval (ISI). This test has been used to measure various photoreceptor recovery properties by changing the intensity and colour of the probe and test flashes, as well as the length of the ISI in humans11, 97, 98 and mice 104, 105. This protocol has helped to uncover photoreceptor activity from the masking effects of INL neurons, and has also enabled the separation of rod and cone function. In the latter case, the test flash elicits a mixed response, while the probe flash generates an all cone response, as the ISI is at the point of maximal cone response prior to recovery of rod cells. The difference between the two waveforms represents the rod derived response. We conducted an experiment to determine the appropriate ISI in X. laevis with similar parameters to that used by Pennesi and colleagues 106, in which we plotted the amplitude of response to the probe flash at increasing ISIs. We measured a subtle break point at 1280ms (Figure 3.3). Flicker tests measure photoreceptor recovery, which is known to be different between rods and cones. Since cones recover faster than rods, responses generated at higher frequencies are attributed to cones, and at lower frequencies, to a combination of rods and 51 cones. In humans, a standard flash flickering at 30Hz is commonly used to isolate cone function. However, 30Hz flicker invoked no response in X. laevis. Therefore we did two experiments to determine the frequency at which we could measure X. laevis cone responses to flicker stimuli. We stimulated wild type X. laevis at nine stimulus intensities, and noted the frequency at which the retina no longer responded, as explained in a similar experiment by Norwak and Green107. When these data were plotted, a plateau became apparent between six and seven Hz (Figure 3.4), very close to the six Hz branch point measured in Bufo marinus107. We also conducted a similar experiment in which we measured the amplitude of response to increasing body temperature at nine frequencies. Upon superimposition of the graphs we noticed a difference in the slope above eight Hz. Combining evidence from these two experiments conservatively, we used ten Hz stimulation as a measure of cone function. When a ten Hz flicker is used to stimulate after dark adaptation, the response is of a magnitude of a reliable range. When we follow this with a two minute period of photopic illumination, the retinal response to a ten Hz flicker increased. This is likely due to rod cell saturation, and therefore lack of suppression, or inhibition of cone signal. When responses from five animals were averaged, the baseline and light adapted responses were significantly different (p=0.0058, t test, n=5)- approximately doubled. The amount of noise in X. laevis cones is comparable to light stimulation, and therefore cone cells do not saturate at any light intensity15. All changes due to light adaptation are therefore caused by interaction with rods. Therefore, tests intending to measure cone function must take rod suppression into account. 3.3.3 Average amplitude and variance We were interested in the average amplitudes, and variation of responses between different wild type animals, and the reproducibility of these responses. We tested six wild type frogs on the same day, and one of these animals on six different days. The average Vmax was 138.81µV±27.15µV for different animals on the same day and 126.29µV±15.83µV for the same animal on different days. The sensitivity measure, or 52 the intensity of light that generated half the maximal response was -2.11log cd.s/m²± 0.61log cd.s/m² for different animals on the same day, and -1.94log cd.s/m² ± 0.36log cd.s/m² for the same animal on different days. Since the average amplitude for the animal we tested multiple times was lower than that obtained on the day we tested all animals, we divided all values from multiple animals by the factor responsible for the difference (1.18µV for amplitude, and 0.82log cd.s/m² for sensitivity). Calculations based upon the described normalization, yielded an average Vmax of 117.49µV ± 22.98µV, and a Km of - 2.71log cd.s/m² ± 0.75 log cd.s/m². The average normalized amplitude recorded in responses to a blue flash (0.025 cd.s/m²) was 66.41µV ± 5.35 µV, and in response to a red flash (2.5 cd.s/m²) an amplitude of 112.47µV ± 11.55µV was obtained. A white ten Hz flicker standard flash elicited a normalized average amplitude of 95.04µV ± 9.56µV, and a white flash (25 cd.s/m²) superimposed on a white background (2.5 cd.s/m²) (photopic) stimulated a response with an amplitude of 112.19µV ± 7.67µV. The variability was higher between animals than between testing days on the same animal for all analyses. (See Figure 3.5 for summary.) 3.4 DISCUSSION In this paper we have presented a method of reliably recording X. laevis ERGs. We show typical responses for protocols similar to that recommended by the ISCEV. These responses appear to gradually grow in amplitude during metamophosis, but do not change dramatically during this time. We optimized methods of isolating rod and cone function, and suggested appropriate parameters for doing so in this species. We measure several wild type responses to standard stimuli, and the response of a single animal on separate days, which allowed us to determine typical responses, and confirm reproducibility. Amphibians are very useful research subjects for single cell recordings94, or for recording from isolated retina. The ERG is essentially a summation of single cell responses, and is often used to develop hypotheses regarding the underlying single cell responses. 53 However, in many cases (particularly human patients) these hypotheses are very difficult to confirm at a single cell level. X. laevis may provide a much more tractable model in which ERG-derived measurements can be directly compared to single-cell measurements. X. laevis have outer nuclear layers containing relatively equal numbers of evenly distributed rods and cones, and are easy to manipulate genetically. Therefore, they are valuable in studying a variety of inherited photoreceptor diseases, particularly the progression of rod-cone diseases. Since retinal function changes with time in these diseases, it is valuable to repeatedly test X. laevis harboring the mutations, and measure rod and cone function separately over time. ERGs allow this type of longitudinal experimental design. Regeneration is also an interesting subject that can be assessed using electroretinography. Upon retinectomy, post-metamorphic X. laevis are able to regenerate the entire retina if appropriate blood supply is available71. Again, the ability to measure the progression functionally is valuable. The use of the ERG as an assay of retinal function in X. laevis is not yet as advanced as in mammals. Continued work in this field will help us understand the relationship between single cell and summed responses to light, measure retinal functional in X. laevis models of human diseases, and work towards applying similarities as well as unique aspects of X. laevis retinas towards therapeutic interventions. 54 3.5 FIGURES Figure 3.1. ERG setup. We placed an anesthetized frog or tadpole on a tricane soaked sponge, and dried both eyes and head. We placed gold corneal electrodes on the center of the eye, and the reference needle between the eyes on the head. A ground electrode was placed in the sponge, touching the body. Electrodes were plugged into two headstage amplifiers, each receiving information from all three electrodes. Current generated by the retina in response to light stimuli was matched in impedance, and amplification was conserved through the headstage amplifier, differential amplifier, and resistors. Signals from each eye, and a combined reference and ground were inserted into channels one and two via the patient montage, and connected to software to measure response and control protocols. (DiagnoSYS product pictures taken from websites). 55 Figure 3.2. Components of Xenopus laevis electroretinogram. A) A scotopic intensity series was recorded in a wild type frog after dark adaptation, with stimuli ranging from -4 to +2 (0 = 2.5cd*s/m2). The amplitude of a- to b- wave increased with light intensity in a sigmoidal fashion. Oscillatory potentials are superimposed on the b-wave from -1 to 2. B) Photopic (25cd*s/m2 flash on a 2.5 cd*s/m2 background) response recorded after dark adaptation display a- and b- waves, as well as an x- wave on the leading edge of the b- wave. The photopic b- wave occurs earlier, and recovers faster than the scotopic b-wave. (Black= right eye, gray= left eye) 56 Figure 3.3. Defining cone evoking stimuli with paired flash paradigm. X. laevis rod cone break estimated via paired flash stimuli with increasing interstimulus intervals. A) Using a protocol and analysis similar to Pennesi et al. 106, we plotted the amplitude of the test flash at increasing interstimulus intervals (ISIs). From this plot we measured a subtle break point at an ISI of 1280ms. B) Two eyes from the same animal superimposed showing characteristic response to stimuli at this ISI. 57 Figure 3.4. Critical flicker frequency was measured at 6.5Hz. We stimulated wild type X. laevis at nine stimulus intensities, and noted the frequency at which the retina no longer responded, as explained in a similar experiment by Norwak and Green107. When these data were plotted, a plateau became apparent at between six and seven Hz. According to Norwak and Green, this is likely due to the switch from rod to cone dominated response. (n=3) 58 Figure 3.5. Average amplitude and variability of responses. We measured average amplitude and variance (standard error of the mean) for six animals on the same day (grey) and the same animal on six different days (black A) Naka Rushton curve fit from an intensity series including eight flashes in logarithmic increments (0=2.5cd.s/m2). B) A- to b- wave amplitude for four stimuli with intensity in labeled in brackets. ). There was greater variance between animals (gray) than between days (black) in both analyses. (Scale bar applies to A and B) 59 3.6 REFERENCES 1. Karpe G. The basis of clinical electroretinography. Acta Ophthalmol (Kbh) 1945;24:1-166. 2. Yonemura D, Hatta M. Studies of the minor components of the frog's electroretinogram. Japan Journal of Physiology 1966;16. 3. Brindley GS. Responses to illumination recorded by microelectrodes from frog's retina. Journal of Physiology 1956;134. 4. Hood DC, Birch DG. The a-wave of the human ERG and rod photoreceptor function. Investigative Ophthalmology and Vision Science 1990;31:2070-2081. 5. Hood DC, Birch DG. Rod phototransduction in retinitis pigmentosa: Estimation and interpretation of parameters derived from the rod a-wave. Investigative Ophthalmology and Vision Science 1994;35:2948-2961. 6. Shady S, Hood DC, Birch DG. Rod phototransduction in retinitis pigmentosa. Distinguishing alternative mechanisms of degeneration. Investigative Ophthalmology and Visual Science 1995;36:1027-1037. 7. Pepperberg DR, Birch DG, Hofmann KP, Hood DC. Recovery kinetics of human rod phototransduction inferred from the two-branched a-wave saturation function. Journal of the Optical Society of America 1996a;13:586-600. 8. Pepperberg DR, Birch DG, Hood DC. Photoresponses of human rods in vivo derived from paired flash electroretinograms. Visual Neuroscience 1997;14:73-82. 9. Kroll KL, Amaya E. Transgenic Xenopus embryos from sperm nuclear transplantations reveal FGF signaling requirements during gastrulation. Development 1996;122:3173-3183. 10. Umino Y, Everhart D, Solessio E, et al. Hypoglycemia leads to age-related loss of vision. Proceedings of the National Academy of Science 2006;103:19541-19545. 11. Weleber RG. The effect of age on human cone and rod ganzfeld electroretinograms. Investigative Ophthalmology and Vision Science 1981;20:392-399. 12. Lim S-H, Ohn Y-H. Study of blue and red flash in dark adapted electroretinogram. Korean Journal of Ophthalmology 2005;19:106-111. 13. Wachtmeister L, Dowling JE. The oscillatory potentials of the mudpuppy retina. Investigative Ophthalmology and Vision Science 1978;17:1176-1188. 60 14. Makhankov YV, Rinner O, Neuhauss SCF. An inexpensive device for non- invasive electroretinography in small aquatic vertebrates. Journal of Neuroscience Methods 2004;135:205-210. 15. Nieuwkoop PD, Faber J. Normal table of Xenopus laevis. Amsterdam: North Holland Publishing Company; 1956. 16. Birch DG, Hood DC, Nusinowitzet S, Pepperberg DR. Abnormal activation and inactivation mechanisms of rod transduction in patients with autosomal dominant retinitis pigmentosa and the pro-23-his mutation. Investigative Ophthalmology and Vision Science 1995;36:1603-1614. 17. Lyubarsky AL, Pugh ENJ. Recovery phase of the murine rod photoresponse reconstructed from electroretinographic recordings. The Journal of Neuroscience 1996;16:563-571. 18. Goto Y, Peachey NS, Ripps H, Naash MI. Functional Abnormalities in transgenic mice expressing a mutant rhodopsin gene. Investigative Ophthalmology and Vision Science 1995;36:62-71. 19. Pennesi ME, Howes KA, Baehr W, Wu SM. Guanylate cyclase-activating protein (GCAP) 1 rescues cone recovery kinetics in GCAP1/GCAP2 knockout mice. Proceedings of the National Academy of Science 2003;100:6783-6788. 20. Nowak LM, Green DG. Flicker fusion characteristics of rod photoreceptors in the toad. Vision Research 1983;23:845-849. 21. Kefalov V, Estevez ME, Kono M, Goletz PW, Yau K-W. Breaking the covalent bond - A pigment property that contributes to desensitization. Neuron 2005;46:879-890. 22. Yoshii C, Ueda Y, Okamoto M, Araki M. Neural retinal regeneration in the anuran amphibian Xenopus laevis post-metamorphosis: Transdifferentiation of retinal pigmented epithelium regenerates the neural retina. Developmental Biology 2006;1-12. 61 CHAPTER 4. Discussion 4.1 OBJECTIVES In this study our goals were to: 1) induce rod death by apoptosis in primary tadpoles, 2) establish an F1 generation with consistent rod apoptosis, 3) induce rod death in transgenic post-metamorphic frogs, 4) refine methods with which to assay retinal function longitudinally, and 5) gain insight into the implications of rod death on the retina. 4.1.1 Goal 1) Induce rod cell death by apoptosis in primary tadpoles We successfully completed this goal early in the project. After selection of iCasp9 out of the group of constructs provided from Dr David Spencer, we were able to ablate rods as consistently as differing expression would allow. We were fortunate to avoid extensive struggles with method of drug exposure, adequate crossing of the blood retinal barrier, and identifying an appropriate drug concentration. Figure 2.1 displays typical results of multiple experiments executed in primary transgenic animals. We attempted a few refinements, such as isolating and inserting X. laevis caspase 9 in place of the human gene provided by Dr Spencer, as well as including insulators which flanked the transgene within the genomic DNA to protect from silencing. Both met only limited success (data from double insulators is presented in Appendix I), and were not pursued further as they did not significantly improve on the efficiency of the system in the original form. 4.1.2 Goal 2) Establish an F1 generation with consistent rod apoptosis We mated four founders, two male, and two female, each with a wild type frog. We found varying expression levels, number of integration sites, and efficiency of AP20187 induced apoptosis in their F1 offspring. The females did not produce useful offspring, but both males did. Male one had two independently segregating integration sites for iCasp9, and one integration of eGFP which co-segregated with an iCasp9 integration site. Therefore all eGFP positive animals expressed iCasp9. These animals showed consistent response to AP20187 from day seven to day 14, (Figure 2.3). We were therefore able to characterize the process of apoptosis with these animals (Figure 2.4). Progeny from this founder 62 underwent slow decrease in iCasp9 expression starting at dfp15, to the point where AP20187 no longer induced apoptosis in adulthood. (Potential causes of this will be discussed in section 4.2.1.) Therefore progeny from this founder are most useful for experiments within the first two weeks after fertilization. The progeny from male two were genetically more complicated, as the founder had two independently segregating integrations of eGFP, and one integration of iCasp9, which segregated with eGFP. This means we were forced to identify iCasp9 positive animals from the type of eGFP expression (visual inspection and rating as a weak or strong expresser), which is less reliable. Despite this minor limitation, we were able to do similar progression experiments to that executed in F1 from male one, (Appendix II), and reproducibly induce rod death in these animals in a variety of stages. To date, we have measured iCasp9 expression up to stage 66. 4.1.3 Goal 3) Induce rod cell death in transgenic post-metamorphic frogs Our first attempt to initiate rod cell apoptosis in post-metamorphic animal was a success, and is summarized in Figure 2.2. However, upon examining larger cohorts of animals, the percentage of transgenic frogs responsive to AP20187 has been lower than expected. A significant amount of time and energy has gone into questioning why this might be the case, and these ideas are presented in section 4.2. Despite this set back, we know it is possible to induce rod death in mature animals, and are working on establishing an F1 generation in which we meet this goal more reliably. 4.1.4 Goal 4) Refine methods with which to assay retinal function longitudinally Chapter three is devoted to the appropriate use of electrophysiology in X. laevis. We have been able to adapt standard ERG equipment to accommodate the differences intrinsic in our animals at different stages of development. We are able to measure rod and cone function independently through a variety of stimuli. We have also established expected amplitude of response, and variation for several standard stimuli. Therefore, we are now able to use electroretinography to assay rod and cone initiated function over time safely and reliably. 63 4.1.5 Goal 5) Gain insight into the implications of rod cell death on the retina The discussion in Chapter two and section 4.3 provide some insight into the implications of rod death on the retina. As will be discussed, this system will allow us to probe these issues in more depth, but even now, we can speculate about the implications of rod cell death in the X. laevis retina. We know there is a resulting dysfunction of cone cells within the first three months of rod death, which improves with time (Figure 2.5). We also know that five months after the initial AP20187 injection, there are more cone cells than in control animals, centrally located rod cell blebs, and a thickened inner nuclear layer (Figure 2.6). Since we do not know if cones remained healthy throughout this time, or if they died and regenerated, we must present at least two hypotheses to explain these observations. Scenario one is that cone cells sustained massed rod ablation, with only a delayed, temporary loss of function. Referring to hypotheses of secondary cone death described in sections 1.1.4, and 2.4, this includes exposure to toxins released by dying rods (potentially a reduction in ATP, or increase in calcium through gap junctions between rods and cones1) and a sudden excess of oxygen in the absence of metabolically active rod cells2-4. However, the reduction in cone function was gradual, and therefore this type of direct interaction is not likely to have caused the functional compromise. The cones were also able to survive without rod cells, and are therefore not dependent on a rod derived viability factor5, that could be the cause of gradual death. If ectopic synaptogenesis occurred after rod ablation between bipolar and horizontal cells and cones, as we expect in mammals6, 7, cones were able to cope with this, and subsequently establish an effective circuitry. This remodeling is supported by changes in the INL. Therefore, the most likely scenario is that our initial “induced” remodeling of the retina (i.e. elimination of rods) is incompatible with cone viability and/or function, but subsequent remodeling over a more prolonged period created a retinal environment that is compatible with both cone viability and function. 64 It is possible that cone cells did in fact die during the time of reduced function. If this was the case, the same arguments could be made suggesting rewiring is the most likely mechanism of secondary cone death. Additionally, this also necessitates that cone death was followed by regeneration, and appropriate wiring, as evident by morphologically and functionally normal cone cells at five months after initial injection. We also noticed blebbing rod cells in the central retina two days after the fifth month of AP20187 injections, which suggests some rod cells also regenerated during the month between injections. Together, these data address questions posed recently about the extent, origin, and applications of post-metamorphic retinal regeneration in X. laevis88, 9. Two potentially very interesting ideas that arise from this are 1) cone function is able to improve with time, which suggests a neural plasticity that would be valuable to understand, and 2) photoreceptor regeneration appears to occur in X. laevis, possibly in both rods and cones. The fact that we measured changes in the INL gives us a place to start looking for the mechanisms of these types of changes. 4.2 CHALLENGES There have been two challenges during these experiments that must be addressed. One is the extent to which this system works consistently in post-metamorphic animals. The second is the ability of electroretinography to reliably measure small variations in retinal function. Understanding these two challenges in more detail, will allow us to use these resources more effectively. The following sections provide clues, and hypotheses we have generated to help us better understand these challenges. 4.2.1 Difficulties inducing rod cell death in post-metamorphic X. laevis. We had difficulty inducing rod cell death consistently in a large cohort of animals, as was the hope after the interesting results from a small cohort discussed in Chapter two. This could be due to a number of non-mutually exclusive reasons. Here I have presented two possibilities. 65 Expression of iCasp9 As animals grow, we noticed a reduction in eGFP expression, and within eGFP-positive animals, less response to AP20187 in post metamorphic animals. It is important to note that we notice this to varying degrees in different groups of animals, with no specific pattern emerging at this point. If expression of iCasp9 is decreasing, either specifically during the changes associated with metamorphosis, or as a steady decline from larval stages, there are three hypotheses that could account for it. 1) The rhodopsin promoter activity decreases at later developmental stages. This seems unlikely, as the amount of endogenous rhodopsin does not appear to be reduced with age. 2) The nature of transgenic expression accounts for a unique type of reduction, which does not occur for endogenous rhodopsin. Specifically, that certain factors are capable of identifying abnormal gene insertions in the X. laevis genome and act on the promoter to decrease expression over time. This is possible, and therefore pursuing the insulated trangene designed to minimize gene silencing mentioned in Appendix I is an option. 3) ICasp9 activates spontaneously, and over time provides a retina skewed towards low expression of iCasp9. Although not observed to a significant extent in our experiments, it is theoretically possible for expressed iCasp9 to autoactivate and initiate the apoptotic cascade in the absence of AP20187, causing (possibly very slow) death of the rods. The outcome of this hypothesis would be adult transgenic frogs with uninduced rod ablation, no rods, and poor ERGs. This is not what we find. Rather, we see healthy rod cells and good ERGs, but in some cases, no response to AP20187 and/or low, mosaic, or non- expression of iCasp9. If regeneration of photoreceptors is possible in the X. laevis retina, as we suspect, every rod cell that dies will eventually be replaced. Through mitosis and cell division regenerating cells could somehow have the capacity to completely or partially silence transgene expression. Regenerated cells with a silenced or reduced- expression iCasp9 transgene would have a survival advantage over those expressing auto- activating levels of iCasp9. The longer this process continues, the more likely the outer nuclear layer will be dominated by rod cells not expressing iCasp9, as transgene expression would be ‘selected against’. 66 However, prior to pursuing the mechanism of reduced iCasp9 expression for the purpose of eliminating this problem, we should continue to pursue identification of animals with consistent iCasp9 expression in late stages of development, since we know this reduction in sensitivity to AP20187 does not always occur. The primary transgenic animals described in Chapter two responded well to AP20187 in post-metamorphic stages, and a subset of primary trasngenics similarly identified in later experiments also reacted to the drug in post metamorphic stages. Furthermore, we have recently identified a subset (four out of ten) of the F1 generation from male two that express iCasp9 consistently past metamorphosis, and show responsiveness to AP20187. AP20187 The properties and administration of AP20187 could be relevant to our difficulty inducing rod cell death in post-metamorphic animals. As we recall from section 1.1.4, some relevant points regarding the intrinsic apoptotic pathway in the context of iCasp9-induced activation are as follows: 1) Although proteolytic activation is sufficient for the activity of most caspases, it has been speculated that apoptosome formation is necessary for the downstream effects of caspase 910, and there is no a priori reason to suspect that an apoptosome containing mitochondrial cytochrome c would be formed on administration of AP20187. However, we (and others11, 12) have demonstrated that caspase 9 dimerization is sufficient to induce apoptosis in systems such as cultured cells, tadpoles, and post-metamorphic frogs. 2) FK506 (the precursor to AP20187) inhibits calcineurin through interaction with endogenous FKBP12 (an immunophilin), which inhibits Bad dephosphorylation, and therefore release of cytochrome-c13. This means FK506 derivatives could have anti- apoptotic effects upstream of caspase 9. However, specific chemical modifications were made to the FK506 derivatives FK1012 and AP20187 to prevent interaction with calcineurin14, and only in very high doses were effects on calcineurin observed with FK101214. 3) There are ten to 50 times more immunophilins in the brain than in the 67 immune system, and the binding levels are up to 50 times higher. Therefore, conceivably at ‘normal’ drug levels, an anti-apoptotic effect of AP20187 may be possible in neurons. However, even if this were the case, inhibition of reactions upstream of caspase 9 are unlikely to effect iCasp9 activation by dimerization or the resulting cascade. For either inefficient apoptosis induction, or increase apoptosis inhibition to be the cause of our observed reduction of induced apoptosis in later developmental stages, these factors must change significantly with age. Other considerations about the drug include proper handling, reconstituting, and lot number, which we have been careful to control for, but can not eliminated as candidate problems. 4.2.2 Measuring small changes in retinal function with electroretinography Chapter three summarizes what we know about the origin of certain ERG components, and the average values and variability we see from control animals. We have used this information to identify abnormal retinas, as seen in Chapter two. We are able to identify complete rod ablation (Figure 2.2), and changes in cone function over time (Figure 2.5). Our analysis has been limited to a- to b- wave amplitude, projected a- wave asymptote, measures of sensitivity based on the Naka Rushton equation, and photopic amplitude. This information has been extremely valuable, but as we can see from the work done in humans and rodents, this is only a portion of the rich information that can be derived from electroretinography experiments. We have not yet characterized all the components of the X. laevis ERG, or identified what their parameters tell us about the retina of this animal. We have not conducted detailed analysis on more advanced parameters of the a- wave (particularly within the paired flash paradigm), probed oscillatory potentials, or the x-wave, or even the latencies of each of the components, all of which have provided useful information about the retinal health in humans and mice. Each of these things will enrich the information we can collect from the animals without sacrificing them. In the cases where we have tested more subtle changes in waveforms, our results have not been conclusive. We cannot yet fully account for the variability between animals, and 68 between days on which we collect electroretinography data, therefore, in some cases it is difficult to differentiate minor pathology from normal variation. To use analyses based on more subtle changes effectively, we need to continue to work on our technique to make our measurements more consistent. As we make progress in this area, it will also be useful to sacrifice animals upon ERG analysis to correlate minor changes in retinal function to changes in retinal morphology. With continued research, this tool will develop into a more diverse measure of retinal function which will be useful to test this model, and also any other models of retinal abnormality in X. laevis. We can also use this technique to investigate specific hypotheses regarding defects in phototransduction or neural pathways in the retina of transgenic X. laevis. The amount of research correlating changes in retinal function to minor variations in ERG parameters in humans and rodents suggest the possibilities are endless15-18. We have the benefit of learning from these studies, and replicating them in X. leavis, as we have begun to do in Chapter three. Our knowledge of X. laevis retinal function will no doubt increase exponentially in the coming years. 4.3 FUTURE WORK The previous sections suggested multiple experiments that could be done in order to tackle some of the challenges posed by the work done so far. These experiments will be important contributions to people who work with X. laevis, allowing this type of inducible apoptosis, and electroretinography to be employed in the most efficient way possible. However, if we refocus to the primary question of what occurs after rod cell death, this project poses some interesting ideas that could be pursued by researchers in two fields: mechanisms of secondary cone cell dysfunction, and retinal regeneration. Currently, we are working with Robert Marc’s lab in Utah. This lab studies retinal remodeling resulting from various types of retinal damage. Marc and colleagues have developed a technique called computational molecular phenotyping19, which allows the 69 visualization of small molecule targets, and identification of cells based on analyses of their distribution. They are able to analyze processes such as Muller cell activation, ectopic synapses, and stress responses within neurons. They currently have eyes from the progeny of male two which were treated with AP20187 at day seven, and sacrificed up to day 16 (contralateral eyes presented in Appendix II). The first objective of our collaboration is to characterize Muller cell activity in these retinas. Muller cells are support cells within the retina with diverse functions. When photoreceptors undergo a stress response, initially Muller cells have a protective effect by releasing growth factors20, however, continual activation leads to a hypertrophic state characterized by an upregulation of glial fibrillary acidic protein (GFAP), and changes in osmotic regulation. Hypertrophic Muller cells are associated with a thickening of the retina21, and retinal remodeling22. Muller cells eventually form seals between the inner and outer nuclear layers, contributing to neuroma formation, and overall retinal dysfunction7. This type of remodeling poses problems for therapies directed at replacing photoreceptors through stem cell therapy or transplantation, and even more so for therapies directed at replacing genes within photoreceptors, as the retinal circuitry is no longer able to accommodate healthy photoreceptors. Muller cell activity is at the beginning of this process, and understanding the events that occur upon initial insult is valuable. Studying these processes relies on replicating to a reasonable extent the initial insults occurring in human disease. As mentioned previously, rod cell apoptosis is a common initiating insult that leads to retinal degeneration in human diseases, therefore the ability to induce rod apoptosis without causing damage to other retinal cells provides an extremely valuable tool. Furthermore, in our animals, these early steps can be studied in detail, as rod death occurs between 24 and 48 hours after AP20187 exposure, within which we can easily generate as many time points as is useful, as numbers are not a limiting factor when we are using F1 animals. Our tadpoles provide an easy way to not only characterize Muller cell activation, but also to test compounds that may inhibit it. 70 These would be valuable experiments that could be done in relatively short two week time periods, allowing us to test multiple agents with ease. The Marc lab is also testing our transgenic X. laevis retinas with a gamut of small molecules allowing assessment of photoreceptor stress, as well as the integrity of synaptic connections between photoreceptors and bipolar cells. Analysis of these results will help determine the mechanism of secondary effects on the retina, and allow us to pool this data to the growing field of secondary cone cell death, which has such significant implications for people with retinitis pigmentosa. Similar work could be done in lines with mosaic expression patterns to examine the impact of various levels of rod cell death in tadpoles. Regeneration is another field this project could contribute to. Recently there has been much interest in the regenerative ability of anuran amphibians, specifically X. laevis. This has been initiated by an interesting finding by Yoshi and colleagues9, and a subsequent review by the same group8. We have observed interesting phenomenon, described in Chapter two, that suggest our animals also undergo regeneration in adulthood. We have not yet probed the extent to which this occurs. It will be interesting to induce rod death, remove the death stimulus, and characterize the extent of any regeneration. If this occurs, the interesting questions will be: where do the new cells arise from? The data presented in Chapter two suggests they are originating from the central retina, possibly from the INL. This is very interesting, as regeneration at later stages has only been documented to occur via transdifferentiation of the RPE9 or proliferation from the ciliary marginal zone. From this point, many significant hypotheses can be tested. For example, what transcription factors are produced after rod death that initiate regeneration? Do these factors initiate proliferation of precursors, or transdifferentiation of mature cells? The ultimate goal will be to take the answers to these questions, and use them to develop techniques that will stimulate this type of activity in human patients that do not have a regenerative capacity. 71 We have been able to set up a model system in which inducible rod cell death is specific, consistent, and relevant to human disease processes. We have also developed an assay capable of measuring functional outcomes non-invasively in these animals. This provides a framework to understand, and eventually interfere with mechanisms leading to vision impairment. I am proud to have made this small contribution towards a very valuable goal. 72 4.4 REFERENCES 1. Ripps H. Cell death in retinitis pigmentosa: Gap junctions and the 'bystander' effect. Experimental Eye Research 2002;74:327-336. 2. Cingolani C, Rogers B, Lu L, Kachi S, Shen J, Campochiaro PA. Retinal degeneration from oxidative damage. Free Radical Biology and Medicine 2006;20:660- 669. 3. Shen J, Yang X, Dong A, et al. Oxidative damage is a potential cause of cone cell death in retinitis pigmentosa. Journal of Cellular Physiology 2005;203:457-464. 4. Rogers B, Symons RCA, Komeima K, et al. Differential sensitivity of cones to iron-mediated oxidative damage. Investigative Ophthalmology and Visual Science 2007;48:438-445. 5. Leveillard T, Mohand-Said S, Lorentz O, et al. Identification and characterization of rod-derived cone viability factor. Nature Genetics 2004;36:744-759. 6. Banin E, Cideciyan AV, Alema TS, et al. Retinal Rod Photoreceptor–Specific Gene Mutation Perturbs Cone Pathway Development. Neuron 1999;23:549–557. 7. Jones BW, Watt CB, Marc RE. Retinal remodelling. Clinical and Experimental Optometry 2005;88:282-291. 8. Araki M. Regeneration of hte amphibian retina: Role of tissue interaction and related signalling molecules on RPE transdifferentiation. Developmental Growth and Differentiation 2007;49:109-120. 9. Yoshii C, Ueda Y, Okamoto M, Araki M. Neural retinal regeneration in the anuran amphibian Xenopus laevis post-metamorphosis: Transdifferentiation of retinal pigmented epithelium regenerates the neural retina. Developmental Biology 2006;1-12. 10. Hengartner M, O. The biochemistry of apoptosis. Nature 2000;407:770-776. 11. MacCorkle RA, Freeman KW, Spencer DM. Synthetic activation of caspases: Artificial death switches. Proceedings of the National Academy of Science, USA 1998;95:3655-3660. 12. Fan L, Freeman KW, Khan T, Pham E, Spencer DM. Improved death switches based on caspases and FADD. Human Genetic Therapy 1999;10:2273-2285. 13. Grosskreutz CL, Hanninen V, A., Pantcheva M, B., Huang W, R. PN, Dobberfuhl A, P. FK506 blocks activation of the intrinsic caspase cascade after optic nerve crush. Experimental Eye Research 2005;80:681-686. 73 14. Spencer DM, Wandless TJ, Schreiber SL, Crabtree GR. Controlling signal transduction with synthetic ligands. Science 1993;262:1019-1024. 15. Birch DG, Hood DC, Nusinowitzet S, Pepperberg DR. Abnormal activation and inactivation mechanisms of rod transduction in patients with autosomal dominant retinitis pigmentosa and the pro-23-his mutation. Investigative Ophthalmology and Vision Science 1995;36:1603-1614. 16. Pepperberg DR, Birch DG, Hood DC. Photoresponses of human rods in vivo derived from paired flash electroretinograms. Visual Neuroscience 1997;14:73-82. 17. Gargini C, Terzibasi E, Mazzoni F, Strettoi E. Retinal organization in the retinal degeneration 10 (rd10) mutant mouse: A morphological and ERG study. The Journal of Comparative Neurology 2007;500:222-238. 18. Hood DC, Birch DG. The a-wave of the human ERG and rod photoreceptor function. Investigative Ophthalmology and Vision Science 1990;31:2070-2081. 19. Marc RE, Kalloniatis M, Jones Bw. Excitation mapping with the organic cation AGB2+. Vision Research 2005;45:3454-3468. 20. Frasson M, Picaud S, Leveillard T, et al. Glial cell line-derived neurotrophic factor induces histologic and functional protection of rod photoreceptors in the rd/rd mouse. Investigative Ophthalmology and Visual Science 1999;40:2724-2734. 21. Jacobson SG, Cideciyan AV, Sumaroka A, et al. Remodeling of the human retina in chorioderemia: rab escort protein (REP-1) mutations. Investigative Ophthalmology and Vision Science 2006;47:4113-4120. 22. Jones BW, Marc RE. Retinal remodeling during retinal degeneration. Experimental Eye Research 2005;1-15. 74 APPENDICIES 5.1 APPENDIX I. Optimizing inducible apoptosis system Figure 5.1.1. AP20187 has no effect in wild type tadpoles. In wildtype animals AP20187 did not have a significant effect on the quantity of rhodopsin as measured by western blot. There was slightly more rhodopsin measured in treated eyes than non- treated eyes. 75 Figure 5.1.2. A dose response analysis of AP20187 in primary transgenic tadpoles shows optimal results at 10nM. EGFP-positive primary transgenic tadpoles were raised until seven days, and then place in ringer with varying AP20187 concentrations. At 14 days we sacrificed all animals, and measured anti-HA signal on a western blot. The anti- HA signal was lowest in animals treated with 10nM. No anti-HA signal was detected in microscopy at this concentration, indicating AP20187 at concentrations higher than 10nM would not enhance the effect. 76 Figure 5.1.3. ICasp9 subcloned into a plasmid with double insulators improved the expression of iCasp9 in primary transgenics. In untreated primary transgenic retinas, iCasp9 expression was higher when the gene was flanked by insulator (2DI), compared to the uninsulated counterpart. The 2DI iCasp9 was similarly broken down on administration of AP20187, as was the case with the original construct. 77 5.2 APPENDIX II. Progeny from male one and male two Figure 5.2.1. Male 1. Consistent iCasp9 expression at 14 days in eGFP positive tadpoles. Primary transgenic tadpoles treated with 10nM AP20187 at day seven, and sacrificed at day 14 showed consistent expression of HA as measured by regular anti-HA signal in eGFP-positive retinas. The progeny from this founder had consistent expression, activation of iCasp9, and induction of apoptosis, as documented in Chapter 1. This was consistent in several matings with many tadpoles at this stage. It is apparent by the iCasp9 present in the GFP negative animals that there are at least two integration sites of iCasp9, only one of which is associated with eGFP. 78 Figure 5.2.2. Male 1. After 14 days, the expression of iCasp9 decreases such that it interfered with AP20187-induced apoptosis. Approximately half of the eGFP positive animals no longer expressed iCasp9 at a sufficiently high level for AP20187 administration to induce apoptosis. This poses limitations for the progeny of this frog. Although valuable in early stages of development, apoptosis can not reliably be initiated after this point. In post-metamorphic F1 frogs from this founder, ERG response is not compromised after AP20187 injection. 79 Figure 5.2.3. Male 2. Expression of iCasp9 differed with the type of eGFP expression, but in the cases where iCasp9 expression was apparent, AP20187 induced apoptosis was dramatic. This highlights two important attributes of progeny from Male 2. 1)There are likely two eGFP integration sites, and only one iCasp9 integration site, therefore eGFP expression does not guarantee iCasp9 expression. However, the phenotype of the eGFP integration site associated with iCasp9 was distinguishable from the other. No eGFP expression, or dim uniform eGFP expression, are both associated with the absence of iCasp9, while mosaic bright eGFP expression was associate with high uniform iCasp9 expression. However, prior to sectioning, distinguishing the more uniform dim expression from bright, mosaic expression is a somewhat unreliable assay. Here the column labels indicate categorization by looking at eGFP expression in vivo, in the uniform eGFP category, I displayed animals in which it is clear that all except that in the lower right quadrant are in this category erroneously. We have been able to overcome this limitation to a certain degree with experience, therefore, these animals show promise for a good model, as apotosis is complete, and occurs at 14dfp, as well as 21dfp, and a response has been recently observed by ERG in some post- metamorphic animals. 80 Figure 5.2.4. Male 2. The progeny from Male 2 show high iCasp9 expression, and AP20187 induced retinal apoptosis. Retinal sections at various time points show consistent iCasp9 expression in untreated F1 animals and characteristic iCasp9 activation, apoptosis and clearing of rod cells in AP20187 treated animals. In this case these animals were selected (best one from group of three in each category), and the contralateral eye sent to Robert Marc’s lab to be analyzed using computational molecular phenotyping. 81 5.3 APPENDIX III. University of British Columbia Animal Care Certificates 82 83 84 85 86 "@en ; edm:hasType "Thesis/Dissertation"@en ; vivo:dateIssued "2008-05"@en ; edm:isShownAt "10.14288/1.0066186"@en ; dcterms:language "eng"@en ; ns0:degreeDiscipline "Neuroscience"@en ; edm:provider "Vancouver : University of British Columbia Library"@en ; dcterms:publisher "University of British Columbia"@en ; dcterms:rights "Attribution-NonCommercial-NoDerivatives 4.0 International"@en ; ns0:rightsURI "http://creativecommons.org/licenses/by-nc-nd/4.0/"@en ; ns0:scholarLevel "Graduate"@en ; dcterms:title "Controlled ablation of rod photoreceptors in transgenic Xenopus laevis"@en ; dcterms:type "Text"@en ; ns0:identifierURI "http://hdl.handle.net/2429/236"@en .