LENTIVIRAL-MEDIATED RESTORATIO N OF SIGHT IN THE GUCY1*B CHICKEN By MELISSA WILLIAMS A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2006
Copyright 2006 by Melissa Williams
iii ACKNOWLEDGMENTS First, I would like to thank Sue Semple-Row land for the opportunity to be a part of such an amazing project. I was truly blessed to come into her laboratory when I did. Many years of diligent hard work on her part and the part of previous members of her laboratory laid the groundwork for the exciting discovery of rescue of vision at the behavioral level and I have had the awesome pleasure of being a part of all this work coming to fruition. Further, I would like to thank the memb ers of my committee, Clay Smith, Tom Foster and Greg Schultz for invaluable guidance, support and understanding throughout my graduate career. I would also like to thank past and pr esent members of the Semple-Rowland lab. First, I thank Shannon Haire for her assist ance with injections, her perseverance, understanding, scientific guida nce and comic relief. I also thank Amy Robinson and Elizabeth Humberstone for e xpert assistance with behavi oral studies and molecular biology, and for their friendship. I thank Kris Eccles for assistance with tissue processing and computer angst. Importantly, I thank Jason Coleman for laying so much of the groundwork that made the rescue experiment a reality.
iv TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iii LIST OF FIGURES...........................................................................................................vi ABSTRACT......................................................................................................................v ii CHAPTER 1 INTRODUCTION........................................................................................................1 Vertebrate Phototransduction.......................................................................................1 Retinal GC1..................................................................................................................2 Leber Congenital Amaurosis Type 1..........................................................................3 GUCY1*B Chicken: An An imal Model for LCA-1.....................................................3 Lentiviral Vector Mediated Retinal Gene Therapy......................................................5 2 DESIGN OF METHODS FOR SUCCESSFUL INJECTION AND HATCH OF TREATED GUCY1*B CHICKENS............................................................................8 Introduction................................................................................................................... 8 Methods......................................................................................................................10 Analysis of Egg Orient ation During Incubation..................................................10 Sanitizing Eggs and Controlling Relative Humidity in Hatching Environment.11 New Method of Injection L eaving the Air Sac Intact.........................................12 Results........................................................................................................................ .13 Egg Orientation During Incubation.....................................................................13 Egg Sanitation and Relative Humidity in Hatching Environment......................14 Injection Leaving the Air Sac Intact....................................................................14 Discussion...................................................................................................................15 3 LENTIVIRAL EXPRESSION OF RETINAL GUANYLATE CYCLASE-1 (RetGC1) IN RETINA RESTORES VISION IN AN AVIAN MODEL OF CHILDHOOD BLINDNESS......................................................................................19 Note........................................................................................................................... ..19 Introduction.................................................................................................................19 Methods......................................................................................................................21 Vector Design......................................................................................................21
v Experimental Animals.........................................................................................21 Embryonic Injections...........................................................................................21 Incubation and Hatching......................................................................................23 Behavioral Analyses............................................................................................23 Electroretinographic (ERG) Analyses.................................................................25 Sacrifice and Tissue Collection...........................................................................26 Retinal Immunohistochemistry and Light Microscopy.......................................27 Genomic and Reverse Tran scription PCR (RT-PCR).........................................29 Statistical Analyses..............................................................................................31 Results........................................................................................................................ .31 In vivo Lentiviral Treatment Restores Optokinetic and Volitional Visual Behaviors.........................................................................................................31 Retinal Electrophysiology also Indica tes Treatment Restores Function.............32 The Number of Integrated Transgenes per Genome Is Related to Treatment Efficacy............................................................................................................36 RT-PCR Analyses and GFP Immunostai ning Confirm Transgene Expression..37 Morphometric Analyses Suggest that Tr eatment Slows Retinal Degeneration..41 Discussion...................................................................................................................44 4 CURRENT AND FUTURE STUDIES......................................................................47 Apoptotic Cell Death in the Degenerating Retina......................................................48 Effects of Transgene Expression On Photoreceptor Viablility...................................50 Non-Autonomous Cell Death and Viral Titer............................................................52 Clinical Relevance to LCA1.......................................................................................53 LIST OF REFERENCES...................................................................................................55 BIOGRAPHICAL SKETCH.............................................................................................61
vi LIST OF FIGURES Figure page 2-1 Embryos are injected at Hamburger Hamilton stages 10-12......................................9 2-2 Egg, prior to injection, with a 5-7 mm opening made in the eggshell overlying the embryo without disturbing the me mbrane adjacent to the shell.........................13 2-3 The effect of egg orientation dur ing incubation on embryo survival.......................14 2-4 Increase in percent hatch of treated embryos over time...........................................18 3-1 In vitro analyses of the function of the pTYF-EF1 -IRES-eGFP vector and virus.......................................................................................................................... 22 3-2 Optokinetic and volitional behavior tests indicate treatment efficacy.....................33 3-3 Retinal electrophysiology rescued in treated eyes...................................................35 3-4 Quantitative genomic PCR estimate of integrated viral trangenes per genome.......37 3-5 Immunohistochemical analyses of GFP expression in treated retinas.....................40 3-6 Comparison of retinal morphol ogy of treated GUCY1*B, wild-type RIR, and untreated GUCY1*B................................................................................................43
vii Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science LENTIVIRAL-MEDIATED RESTORATIO N OF SIGHT IN THE GUCY1*B CHICKEN By Melissa Williams May 2006 Chair: Susan L. Semple-Rowland Major Department: Neuroscience Leber congenital amaurosis (LCA) is a fam ily of autosomal recessive retinopathies that cause congenital blindness in infants a nd progressive retinal degeneration. The first of several genes linked to this disease group was GUCY2D (LCA type 1), the gene encoding retinal guanylate cyclase-1 (GC1). GC1 is localized in the photoreceptor cells of the retina and is an im portant player in the phototr ansduction cascade. The GUCY1*B chicken is an animal model for LCA-1 which also has a mutation in the gene encoding GC1 (GUCY1). These animals are blind at hatch and their retinas are morphologically intact at birth with retinal degeneration b ecoming evident after a pproximately two weeks and progressing with age. The primary goal of this study was to use lentiviral-mediated gene delivery of a transgene encodi ng normal functioning GC1 to restore phototransduction and vision in these animals. Lentiviral vectors were delivered to re tinal progenitor cells of an embryonic GUCY1*B chick. Several variables determine whether treated embryos survive and hatch
viii as healthy chicks. Opening the egg for tr eatment increases the stringency of these requirements. Our efforts to research and te st these variables resulted in the first successful series of hatc hes of treated embryos. Lentiviral-delivery of the GC1 transgene and subsequent expression of functioning GC1 in photoreceptor cells was sufficient to rest ore vision to treated animals as indicated by photoreceptor cell response to light and robust sighted behavior. The treatment, however, did not prevent retinal degeneration. Fo r gene therapy treatm ent of this disease model to be complete in its effectiveness, expression of the therapeutic gene must produce protein which restores function to disabled cells and prevents retinal degeneration. Current and future studies are focused on enhancing treatment to prevent retinal degeneration.
1 CHAPTER 1 INTRODUCTION Vertebrate Phototransduction Two types of photoreceptors in the vertebra te retina respond to light entering the eye, rods and cones. Rods are important for vision in low lighting conditions and cones are responsible for color vision in well-lit environments. Verteb rate phototransduction, a signaling cascade that is triggere d by light and leads to ion ch annel closure, occurs in the outer segments of these cells (reviewed in Koutalos and Yau 1996). In the dark, cation channels located in photoreceptor outer segment membranes are held open by second messenger cGMP molecules allowing ions (pri marily Na+ and Ca2+) to flow into the cell. As a result of this influx, photoreceptors ar e maintained in a depolarized state in the dark. In the light, photons trigger isomeriza tion of the visual pigment chromophore, 11cis retinal, inducing a co nformational change in the visual pigment protein, opsin. This conformational change activates the pigment a nd triggers the phototra nsduction cascade. Once activated, each visual pigment molecule binds and activates several hundred photoreceptor specific G protein transducin molecules by catalyzing the exchange of GDP for GTP. The GTP bound subunit of transd ucin then binds and activates cGMP phosphodiesterase, which subsequently hydrolyz es cGMP to GMP. The decrease in cGMP concentration leads to closure of the cation channels and hype rpolarization of the cell (Koutalos and Yau, 1996; Polans et al., 1996; Arshavsky et al., 2002).
2 Retinal GC1 Guanylate cyclases (GCs) are enzymes th at catalyze the conversion of GTP to cGMP and are present in many cell types. GCs are divided into soluble and membranebound forms (Shyjan et al., 1992; Pugh et al., 1997). Membrane-bound GCs are single transmembrane proteins with the amino-te rminus positioned within the extracellular domain (Shyjan et al., 1992). Two memb rane-bound GCs are expressed in retinal photoreceptors, GC1 and GC2 (Shyjan et al., 1992; Lowe et al., 1995; Yang et al., 1995). GC1 is present in photoreceptor outer segments (Dizhoor et al., 1994) and is important in phototransduction. In order for photoreceptors to be restored to the pre-stimulus state after being exposed to light, cG MP concentrations within the cell must be replenished. In photoreceptor outer segments, GC1, modulat ed by guanylate cyclas e activating protein (GCAP), catalyzes the conversion of GTP to cGMP when calcium levels are low. Increased amounts of cGMP lead to opening of the cation channels and, in conjunction with inactivation of the phot otransduction cascade, return the photoreceptor to a depolarized dark state (Polans et al., 1996; Koutalos and Yau, 1996). In the absence of retGC1, photorecepto r cell function is compromised. In GC1 knockout mice, rod function is abnormal a nd cone cells do not respond to light as measured by electroretinography (ERG) (Yang et al., 1999). By 6 months of age, the cone cells in the retinas of these mice have degenerated (Coleman et al., 2004). In GUCY1*B chickens which carry a null muta tion in retGC1 (Semple-Rowland et al., 1998), both rod and cone cells do not respond to light as measured by ERG, and both cell types degenerate by 6 months of age (Ulsha fer et al., 1984; Ulshafer and Allen 1985).The first signs of retinal degenera tion are evident in the retina s of these animals 7-10 days post hatch (Ulshafer et al., 1984; Ulshafer and Allen, 1985).
3 Leber Congenital Amaurosis Type 1 Leber congenital amaurosis (LCA) is a fam ily of autosomal recessive retinopathies that are responsible for approximately five percent of all cases of human congenital blindness. LCA, first described in 1869 by The odore Leber, presents early in infancy as vision loss, absent or severely reduced ERG response, and a normal appearing fundus. Mutations in several genes have been li nked to LCA, including: RPE65, CRX, GUCY2D, RPGRIP, CRB1 and AIPL1 (Hanein et al., 200 4; Koenekoop et al., 2004). The first gene linked to LCA was GUCY2D, the gene enc oding GC1. GUCY2D maps to chromosome 17p13.1 and null mutations and mutations that disa ble protein function in this gene cause LCA-1 (Perrault et al., 1996). Lack of GC1 function leads to severely decreased cGMP levels in photoreceptor cells (Semple-Rowland et al., 1998; Rozet et al., 2001). Thus, the LCA1 phenotype may be attributed to chroni c closure of the cGMP gated cation channels thereby trapping the photoreceptors in a hype rpolarized state (Perrault et al., 1996; Semple-Rowland et al., 1998). GUCY1*B Chicken: An Animal Model for LCA-1 The GUCY1*B chicken was first described by Cheng et al. (1980) as an autosomal recessive disorder in Rhode Island Red chickens presenting with blindness upon hatching. Since its discovery, the biochemi stry, morphology and genetics underlying retinal degeneration in this animal have been investigated and characterized (Ulshafer et al., 1984; Ulshafer and Allen, 1985; Semple-R owland et al., 1998; Semple-Rowland and Cheng, 1999). These chickens have a deletion/ insertion mutation in the gene encoding GC1 (GUCY1). The deletion removes the re gion of the GUCY1 gene corresponding to exons 4 through 7 and an 81bp DNA fragment is inserted in this region (Semple-Rowland et al., 1998). This dele tion/insertion does not affect the reading frame of the transcript;
4 however, the translated protei n is predicted to lack the transmembrane spanning region and flanking regions. The dest abilizing effects of this de letion are indicated by the absence of GC1 protein in extracts from GUCY1*B retinas (Sempl e-Rowland et al., 1998). Retinal degeneration in this animal m odel is morphologically evident beginning around two weeks of age and is marked by decreased numbers of rod and cone photoreceptor outer segments and gaps innerv ating inner segments and cell bodies of the outer nuclear layer. By three weeks of ag e, pycnotic nuclei are present in the outer nuclear, inner nuclear and ganglion cell la yers; pycnotic nuclei are evidenced by chromatin condensation characteristic of ce lls undergoing apoptotic, or programmed, cell death (Ulshafer et al., 1984). Degeneration is most evident in the central retina at early stages of degeneration and spreads periphera lly with very few photoreceptors remaining intact across the retina by 6-8 months pos t-hatch (Ulshafer and Allen, 1985). In an additional animal model of LCA-1, the GC1 knockout mouse, the pattern of photoreceptor degeneration differs significantl y from that of the GUCY1*B chicken. The retinas of GC1 knockout mouse, like those of the chicken model, are morphologically normal at birth with reduced ERG responses to light stimulation; however, photoreceptor degeneration in the mouse model is mark ed by specific loss of cones while rod photoreceptor cells remain int act (Coleman et al., 2004). Many similarities exist between the GUC Y1*B chicken and LCA1 patients making it an excellent animal model for studying this disease. In both cases, the disease phenotype is linked to a null mutation in the GC1 gene. Further, GUCY1*B chickens are blind at hatch and LCA1 patie nts are blind at birth, both e xhibit severely reduced or
5 absent retinal activity as indicated by ERGs, and both exhibit delayed retinal degeneration relative to visi on loss (Ulshafer et al., 1984; Semple-Rowland et al., 1998; Milam et al., 2003; Koenekoop, 2004). However, there is one report of prenatal retinal degeneration in this genotype (Porto et al ., 2003). Retinal degene ration in the GUCY1*B chicken is morphologically evident beginning around two weeks of age and is marked by decreased numbers of photoreceptor outer segm ents and gaps innervating inner segments and cell bodies of the ou ter nuclear layer. Lentiviral Vector Mediated Retinal Gene Therapy The goal of gene therapy is to deliver therapeutic genetic material to target cells to reverse or slow disease processes. There are multiple vectors, viral and non-viral, used to achieve delivery of therapeutic genes includ ing, but not limited to adenoviral vectors, adeno-associated viral vector s (AAV), lentiviral vectors, and cationic lipids or polymers (Chaum and Hatton, 2002; Goverdhana et al., 2005). Diseases that cause retinal degeneration like LCA and retinitis pigmentosa (RP) have been the primary targets of ocular gene therapy. In autosomal recessive di seases, like LCA, the goal of gene therapy is to rescue the disease phenotype by intr oducing a normal copy of the mutated gene into affected cells. Gene therapy has proven prom ising in many studies of autosomal recessive retinal degeneration (Delyfer et al., 2004). For example, the rd mouse has presented with improved photoreceptor viability and slowed degeneration after viral-mediated delivery of the cGMP phosphodiesterase-beta gene (B ennett et al., 1996; Jomary et al., 1997; Takahashi et al., 1999). A large animal mode l of LCA2, the RPE65 mutant dog (Acland et al., 2001), and a rodent model, the RPE 65 knockout mouse (Dejneka et al., 2004) have shown improved visual function post AAV de livery of the wild type RPE65 gene.
6 Lentiviral vectors, which offer a large transgene carrying cap acity of up to 18 kb (Kumar et al., 2001), are ideal for carrying bicistronic transgenes expressing both therapeutic and reporter genes or two therapeutic genes. Len tivirus is a retrovirus which, upon entry into cells, integrates its genetic ma terial into the host cell genome. However, unlike other retroviruses that primarily infect only dividing cells, len tivirus is capable of crossing the intact nuclear membrane, enabli ng infection of, and long-term expression in, dividing and non-dividing cells (Amado and Chen, 1999; Chaum and Hatton, 2002). Lentiviral vectors are synthesized by tran siently transfecting a cell line with three plasmids. The HIV-1 based, self-inactivating len tiviral vector system consists of a pTYF vector plasmid, which contains the transgene and cis-acting elements vital for integration into the host genome and infection of the ta rget cell, a packaging plasmid pNHP, which contains necessary accessory proteins, and the pHEF-VSVG envelope plasmid (Amado and Chen, 1999; Coleman et al., 2003). The envel ope gene from vesicular stomatitis virus (VSV) is used because the protein coat of HIV-1 binds to the cell membranes of only Tlymphocyte cells and VSV infects all cell type s. Thus, using the VS V-G envelope protein allows packaged lentivirus to infect a variety of cell types (Chaum and Hatton, 2002). Injection of lentiviral vector into the ne ural tube of developi ng chick embryos has resulted in transgene expression in photorecep tor cells (Coleman et al., 2002; Coleman et al., 2003). Specifically, expression of reporte r genes driven by both ubiquitous and photoreceptor-specific promoters, hum an elongation factor-1a (EF1a) and interphotoreceptor retinoid-binding protein (IRBP), respectively, was evident in whole mount and cross sections of embryonic chick retina as early as 5 days post injection.
7 The GUCY1*B chicken, which carries an autosomal recessive GC1 null mutation, is an excellent candidate for gene therapy rescue of vision. Our current understanding of this model has lead to the hypothesis th at introduction of functional GC1 into photoreceptors of this retina would restore pho totransduction and visi on in these animals. In order to test this hypothes is, the initial portion of my thesis research focused on design of successful methods for delivery of lentivir al therapeutic transgenes to the embryo and incubation and hatching of treated embryos. Once the goal of successful hatch was attained, GUCY1*B chickens treated with GC1 lentivirus were tested to determine if phototransduction and vision were restored in these animals.
8 CHAPTER 2 DESIGN OF METHODS FOR SUCCESSFUL INJECTION AND HATCH OF TREATED GUCY1*B CHICKENS Introduction Interest in hatching chicks with inse rted genetic material is highest among research groups interested in the produc tion of transgenic chickens. Liposomes, DNA microinjection, and retroviral infection have been used to introduce genes into developing embryos. Two techniques have been used to breach the exterior shell and membrane of the egg to deliver the genetic material to the embryos. One method that has grown in popularity because of its ease of use it to create a small â€œwi ndowâ€ either in the large end of the egg or on the equatorial plane in th e shell through which the genetic material is delivered. The second method is an ex ovo technique involving transfer of the treated embryos to surrogate egg shells during and post-treatment (Pettite and Mozdziak, 2002). For our experiments, we chose to use the windowing technique to deliver our therapeutic transgene expressing GC1 to phot oreceptor cells of GU CY1*B chick. Vectors were delivered to embryonic day 2 (E2), Hamburger-Hamilton (1951) stage 10-12 embryos. Lentiviral vectors were chosen for in itial rescue studies be cause of their large carrying capacity, allowing for a bicistr onic transgene expressing both GC1 and a reporter to be delivered to transduced cells. Embryos were injected at 33-49 hours of development (Hamburger Hamilton stages 10-12) because during this time the three primary brain vesicles and the optic vesicles are visible on the de veloping neural tube (Figure 2-1) and the embryo has not yet unde rgone torsion and flexion (Hamburger and
9 Hamilton, 1951). The windowing method was used because the surrogate method requires the acquisition of turk ey eggs and mechanical equipment in addition to the standard incubator and hatching environmen ts. However, if the windowing method had proven unsuccessful, the surrogate met hod would have been considered. Figure 2-1. Embryos are inject ed at Hamburger Hamilton st ages 10-12. A.Stained chick embryo at Hamburger Hamilton stage 11, 40-45 hours of age. B.Live chick embryo injected with dye at Hamburger Hamilton stage 11. Positioning of the injection needle is indicated by the blue pointer, injected dye fills ventricular space in brain and optic vesicles. Normally, chicken eggs, once laid, incubate for twenty-one days before hatching. Within this twenty-one day period, multiple variables determine whether the embryos survive and hatch as healthy chicks. Thes e variables include e gg cleanliness, humidity and temperature of the incubating and hatc hing environments, and egg turning during incubation (North, 1978). The windowing tech nique exacerbates the specificity and stringency of these requirements and consid erably reduces hatchability. The purpose of this series of experiments was to design met hods for injection, incubation and hatching of virally-treated embryos that w ould increase the percent of tr eated chickens surviving to hatch; at the onset of these experiment s <<1% of treated embryos were hatching.
10 Methods Analysis of Egg Orientat ion During Incubation In the first experiments to hatch injected embryos, post injection eggs were set on their sides with the window facing up and were incubated without movement. This approach differed from the normal procedur es used in commercial chicken production wherein eggs are incubated with their large e nds up and are rotated back and forth along their vertical axes. This experiment was aime d at determining the most effective position for incubation of treated eggs that woul d both allow movement and extend embryo survival. Prior to injection, on day 0, eggs were set on their side and incubated without rotation at 37.5C and 60% humidity. On embryonic day 2 (E2) (Hamburger-Hamilton (1951) stage 10-12), the position of the embryo along the equatorial plane was determined using an egg-candling light. A pi nhole puncture was made at the large end of the egg shell overlying the air sac facilitati ng settling of the albumen and yolk and a 5-7 mm opening was made in the eggshell and me mbrane overlying the embryo. Viral vector was delivered to the neural tube using a pulled glass capillary needle held by a micromanipulator and connected to a manual microinjector (Sutter Instrument Company, Novato, CA). Approximately 0.5 l vector containing 0.03% fast green dye was slowly injected into the ventricular space of the developing neural tube with the aid of a dissecting microscope (2.5 x mag) (Figure 21). The opening in the egg shell was then sealed with Parafilm M (American National Ca n) using a warmed spatula to adhere the film to the shell and small amount of hot glue served to close the pin hole in the large end of the egg.
11 Of fifteen injected eggs, five were se t on their sides with the window facing up, five were set on their sides with the window facing down and five were set upright, with the large end up. All eggs were incubated at 37.5C and 60% humidity and were slowly turned three times per day (GFQ Sportsman Incubator). On E20 the fifteen incubating eggs were candled and embryo survival was assessed. Sanitizing Eggs and Controlling Relati ve Humidity in Hatching Environment Washing eggs is not reco mmended under normal circumstances and in former experiments we were not washing eggs prior to treatment because it removes some of the protective cuticle coa ting the egg, increasing the likelihood of bacterial and microbial entry into the egg (North, 1978) . This experiment aimed, in part, at determining if washing eggs prior to treatment would have a positive influence on the number of treated embryos surviving to hatch. Additionally, relative humidity levels are esp ecially critical just before hatch, days E18-E21. This experiment also set out to de termine if more fine control of humidity levels in the hatching envir onment would increase survivability of treated embryos. Fifty-five eggs were injected and in cubated under the following conditions over a four week period. Prior to injection, on day 0, eggs were washed with a blend of quaternary detergents (Biosentry RCL) in a water bath preheated to 43C, if the temperature of the water becomes cooler than the egg itself, the egg contents will contract drawing microorganisms in the wash water th rough the pores of th e shell (Cartwright, 2000), and allowed to dry at room temperatur e. Eggs were then set, injected and incubated upright, large end up as described in the prev ious section. At E18, eggs containing surviving embryos, as determined by candling, were moved to a separate hatching environment. From E18 through hatch, eggs were placed on their sides, were
12 stationary, and were maintain ed at 37.5C and 68-70% humid ity. Humidity levels were precisely maintained using an H22 Digita l Humidity Management Module (Brinsea Products Inc.). New Method of Injection Le aving the Air Sac Intact In previous experiments, our injectio n method involved creating a small pinpoint hole into the egg shell at the large end of the egg that breached the air sac prior to opening the shell above the embryo. This pr ocedure produced a temporary displacement of the position of the air sac in the egg, which is normally located at the large end, to a position over the embryo. Benjamin Scott, a graduate student in the Department of Biology at MIT, suggested a method of inject ion that does not requ ire puncturing the air sac. This experiment was designed to determin e if a method of injection that eliminated puncturing the shell overlying the air sac would increase embryo survivability and subsequent hatch rate. Fifty eggs were handled, treated and in cubated over a four week period using methods described in previous sections for washing, setting, incubating and hatching. The following changes were made to the procedur e of injection: On embryonic day 2 (E2) (Hamburger-Hamilton (1951) stage 10-12), the position of the embryo was determined using an egg-candling light and a 5-7 mm opening was made in the eggshell overlying the embryo without disturbing the membrane ad jacent to the shell (F igure 2-2). Sterile PBS (50-100 L) was applied to the exposed membra ne which was then removed prior to injection.
13 Figure 2-2. Egg, prior to injection, with a 57 mm opening made in the eggshell overlying the embryo without disturbing the membrane adjacent to the sh ell. Sterile PBS (50-100 L) is applied to the exposed membrane which will be removed to reveal the developing embryo for injecti on. No puncture was made in the shell overlying the air sac of this egg before the opening pictured here was made. Results Egg Orientation During Incubation This experiment revealed that upright, w ith the large end up is the most effective position for incubation of treated eggs a llowing for movement and extending embryo survival. Fifteen eggs incubated in three different orientations: five set on their side with the window facing up, five set on their side with the window facing down and five set upright, with the large end up, were candled at E20 to determine treated embryo survival. Three of the five treated embryos developing in eggs incubated upr ight, with the large end up survived to E20. None of the embryos in cubated in eggs set on their sides with the window facing up or down surviv ed to E20 (Figure 2-3).
14 Figure 2-3. The effect of egg orientation during incubation on embryo survival. This graph shows the number of treated embryos surviving to E20 (Y-axis) in eggs incubated in three different orientations (X-axis). Of five eggs in each group, three treated embryos survived to E20 in eggs that were se t upright, large end up. No embryos survived in eggs incu bated on their side with the window facing up or down. Egg Sanitation and Relative Humi dity in Hatching Environment Decreasing the risk of contamination by wa shing the eggs prior to injection and generating a fine control of re lative humidity in the hatc hing environment increased the number of embryos surviving to hatch. In cluding egg washing and more stringent humidity controls increased th e percent hatch rate from <<1% to 5.45%. Of the fifty-five embryos undergoing treatment under these c onditions over the four week period, three hatched. Injection Leaving the Air Sac Intact Application of a method of in jection that eliminated puncturing the shell overlying the air sac increased hatch percentage and em bryo survivability. Six of the fifty embryos
15 treated over the four week peri od utilizing the new method of in jection survived to hatch. This alteration in methodology further increased the percent hatch from 5.45% to 12%. Discussion Hatching manipulated chicken embryos requires careful control of several variables. Critical variab les include egg positioning and movement during incubation, egg cleanliness, the viral injection procedure, and the temperature and humidity of the incubator from day 18 to hatch. Movement of treated eggs during incubation is important for survivability. If eggs are not rotated, the two layers of thick albumen within the egg, usually separated by a layer of thin album in, come in contact with each other, significantly increasing embryo mo rtality (North, 1978). This is the likely explanation for why we were able to decrease embryo mort ality and increase chances for successful hatch by changing egg positioning and introducing rocking into the incubation environment. The GUCY1*B chicken colony is housed in litter floor pens and the eggs become very soiled posing a risk for increased em bryo mortality. When these eggs are windowed for injection, the interior of the egg is direc tly exposed to any debris on the exterior of the egg greatly increasing the risk of contamina tion of the embryo. Introducing egg washing using a safe and appropriate detergent at the recommended te mperature, 43 C which is higher than the temperature of the egg interior, reduced the amount of dirt and debris on the egg exterior subsequently decreasing th e chance for contamination and increasing embryo survival rate and hatch percentage. Improving humidity control in the hatc hing environment was perhaps the most important procedural change affecting hatcha bility. If the humidity is too high, the space taken by the egg air sac will decrease in size as water is absorbed into the egg. When this
16 occurs, the full-term embryos can drown within the egg. If humidity levels are too low, the air sac enlarges and the full-term embr yos become sticky and unable to turn to position for hatching (Cartw right, 2000). Movement of the air sac simplified the procedure of positioning the injection needle for delivery of the vector. However, we found that this method, while simplifying injection, increased the like lihood that the membrane lining the egg shell, to which the umbilical cord is attached, would detach from the eggshell compromising the process of yolk sac internalization prior to hatching. T hus, implementing a new injection procedure that eliminated this punctur e had positive affects on embryo survival and hatch rate. Through the series of experiments desc ribed above, carried out from midDecember 2004 to late-January 2005, improvements in egg positioning, humidity control in the hatching environment, egg cleanlin ess and injection procedure significantly improved hatchability (Figure 2-4). An additional increase in the percent hatch rate was achieved when the conventional incubator (ROLL-X incubator, Ly on Electric) and H22 Humidity Modulator combination used as a hatching environmen t was upgraded to an AV-2 Precision Parrot Incubator (Avey Incubator LLC) in Apr il 2005 (Figure 2-4). The H22 Humidity Modulator was accurate in measuring humidity but unable to maintain suitable levels in the large hatching environmen t (this equipment was designed for smaller incubators) without the researcher regularly adding water to the incubator reservoir. The AV-2 allows for precise digital control of temperature and humidity, within 0.2F and 1% relative humidity, respectively, and is equipped with a built in humidifier unit for large scale and
17 rapid maintenance of humidity levels. The decr ease in percent hatch in March was likely due to inaccurate maintenance of humidity levels in the hatching environment. Variables related to animal husbandry also affect hatchability and survival of treated embryos. We found that it is of vital importance for laying females to be on a nutrient rich diet. In May, laying females and males began receiving a diet rich in vitamins and nutrients appropriate for breeding colonies (Breeder 1 Diet, Hillandale Farms, LLC) (Figure 2-4). When appropriate feed diets are provi ded, the health and viability of hatched chicks increase significan tly. Before laying hens began receiving the nutrient rich diet, many chicks were hatching with curly toe paralysis and splayed legs, deformities resulting in part from nutrient deficiencies. Our efforts to research and test multiple variables resulted in the first successful series of hatches of treated embryos. If thes e experimental parameters are followed it is possible to obtain a 12-18% hatc h rate of manipulated embryos.
18 Figure 2-4. Increase in percent ha tch of treated embryos over time. Percent hatch rate (Yaxis) of treated embryos injected each month increasing over time, November 2004 through July 2005 (X-axis). The accompanying timeline shows improvements in variables related to injection, incubation and hatching procedures as they correspond to change s in percent hatch of treated embryos.
19 CHAPTER 3 LENTIVIRAL EXPRESSION OF RETINAL GUANYLATE CYCLASE-1 (RETGC1) IN RETINA RESTORES VISION IN AN AVIAN MODEL OF CHILDHOOD BLINDNESS Note The work presented in this chapter was accepted for publication in PLoS Medicine (Williams et al., in press). My significant contribution to these rescue experiments was to refine the embryonic injection procedure a nd develop methods to successfully hatch treated embryos; design scoring system for, conduct and analyze behavioral assessments; and contribute to interpreta tion of all experimental resu lts. Dr. Jason Coleman designed and tested the therapeutic transgene and optimized the protocol for packaging it into lentivirus. Shannon Haire assisted with inj ections. Drs. Samuel G. Jacobson, Tomas S. Aleman, and Artur V. Cideciyan performe d and analyzed ERG measures. Dr. Susan Semple-Rowland, assisted by Kris Eccles and Amy Robinson, performed and analyzed immunhistochemistry, light microscopy and genomic and reverse-transcription PCR experiments. Amy Robinson also as sisted with behavior testing. Introduction Leber congenital amaurosis (LCA) is a fam ily of autosomal recessive retinopathies that cause congenital blindness in infants. Muta tions in several genes are associated with LCA (Hanein et al., 2004; Koenekoop et al., 20 04). The first gene linked to this disease group was GUCY2D (LCA type 1), the gene encoding retGC1 (Pe rrault et al., 1996). Lack of retGC1 function leads to severely decreased cGMP levels in photoreceptor cells (Semple-Rowland et al., 1998; Rozet et al ., 2001). Normally, cGMP molecules hold
20 cation channels in photoreceptor outer segments open in the dark allowing ions to flow into and depolarize the cell. Light triggers the phototransduction cascade which decreases cGMP concentration and leads to closure of ion channels and hyperpolarization of the cell (Koutalos and Yau, 1996; Polans et al., 1996; Arshavsky et al., 2002). The GUCY1*B chicken, an animal model for LCA-1, has a deletion/insertion mutation in the gene encoding GC1 (GUCY1 ) (Semple-Rowland et al., 1998). These animals are blind at hatch with absent ERGs . Their retinas are mor phologically intact at birth with retinal degeneration becoming ev ident beginning around two weeks of age and progressing with age (Ulshafer et al., 1984). Our current understanding of this model is that in the absence of functional GC1, photore ceptors fail to synthesize sufficient cGMP for normal function and, as a result, are tra pped in a light adapted state (Semple-Rowland et al., 1998), leading us to hypothesize that introduction of func tional GC1 into the photoreceptors of this retina would restore pho totransduction and vision in these animals. To address this hypothesis, we have chosen to use lentiviral v ectors. These vectors provide an attractive gene delivery system because they are capable of carrying large transgenes, a feature that permits use of bicistronic transgenes that express both therapeutic GC1 and a reporter gene. We chose to use EF1 , a ubiquitously expressed promoter, to drive expression of our transgene in our initial experime nts. We chose this promoter because previous analyses of its e xpression characteristics showed that it would produce sufficient GC1 protein to induce a change in photoreceptor function. The nonspecific targeting of expression of the transgene was deemed a potential drawback of the EF1 ; however, this characteristic of the promoter seemed less significant to the overall goal of the experiment.
21 Methods Vector Design The lentiviral transducing vector constructed for use in this study was pTYF-EF1aGC1-IRES-eGFP. This plasmid is the transduc ing vector plasmid backbone of the HIV-1 based, self inactivating lentivir al system (Coleman et al., 2003). The transgene itself is bicistronic, encoding bovine retGC1 and enhanced green fl uorescent protein (eGFP) both of which are driven by an EF1 promoter (Figure 3-1 A). This transgene was packaged into lentivirus using methods developed a nd previously described by Coleman et al. (2003). Previous analyses of the viral vector showed that the GC1 enzyme encoded by the transgene is active and its activity in creases under low calcium conditions in the presence of guanylate cyclase activating protein -1 (GCAP1 ) (Coleman, 2003) (Figure 31 C and D). Analyses of transduced DF 1 cells also showed that GC1 and eGFP proteins are both translated from the viral transgene; this vector was furthe r tested and shown to be capable of transducing embryonic chicken re tina (Coleman, 2003) (Figure 3-1 B). Experimental Animals A breeding colony of GUCY1*B Rhode Isla nd Red (RIR) chickens is maintained at the University of Florida Racing Lab a nd is cared for in accordance with National Institutes of Health guidelines. Experime ntal groups included untreated GUCY1*B chickens (n = 3), wild-type (wt), untreated RIR chickens (n = 5), and treated GUCY1*B chickens (n = 7). Embryonic Injections Prior to injection, on day 0, eggs were washed with a blend of quaternary detergents (Biosentry RCL) in a water bath preheated to 43C. Dried eggs were then set
22 on their sides and incubated without rotati on at 37.5C and 60% humidity. On embryonic day 2 (E2) (Hamburger-Hamilton (1951) stage 10-12), the position of the embryo was Figure 3-1. In vitro analyses of the function of the pTYF-EF1 -IRES-eGFP vector and virus. A. Diagram of the bicistroni c vector, designed by Jason Coleman, indicating production of bovine GC1 and GFP proteins from a single transcript. B. DF1 cells transiently transfected with the vector and subsequently analyzed for expression of bovine GC1 and GFP. Cells immunostained with antibody recognizing bovine GC1 also expressed GFP. C. Comparison of GC1 activity measur ed in bovine rod outer segments, TE671 cells transduced with EF1 -IRES-eGFP virus, and TE671 cells transduced with the control virus, EF1 -eGFP. Activity was assayed in the presence or absence of bovine GCAP1 under both high and low calcium conditions and expressed as nmole cGMP produced/minute. D. Examination of ability of chicken GCAP1 to ac tivate bovine GC1 under physiological conditions. GC1 activity was measured in preparations of bovine rod outer segments in the presence and absence of either bovine GCAP1 or chicken GCAP1 under high and low calcium condi tions. All assays were conducted in triplicate. Bars represent mean SEM.
23 determined using an egg-candling light a nd a 5-7 mm opening was made in the eggshell overlying the embryo without disturbing the memb rane adjacent to the shell. Sterile PBS (50-100 L) was applied to the exposed membrane which was then removed. Viral vector was delivered to the neural tube using a pulled glass capillary needle held by a micromanipulator and connected to a manual microinjector (Sutter Instrument Company, Novato, CA). Approximately 0.5 l vector containing 0.03% fast green dye was slowly injected into the ventricular space of the developing neural tube with the aid of a dissecting microscope (2.5 x mag). The opening in the egg shell was then sealed with Parafilm M (American National Can) using a warmed spatula to adhere the film to the shell. Incubation and Hatching Post-injection, eggs were incubated up right, large end up, at 37.5C and 60% humidity and were slowly rocked three tim es per day (GFQ Sportsman Incubator). At E18, eggs containing surviving embryos, as determined by candling, were moved to a second incubator for hatching that allowed pr ecise control of temp erature and humidity (AV-2 Precision Parrot Incubator, Avey Inc ubator LLC). From E18 through hatch, eggs were placed on their sides, were stationar y, and were maintained at 37.5C and 68-70% humidity. Post-hatch chicks were reared for three to fi ve days in an Octagon TLC-4 brooder (Brinsea Products Inc.) and then transfer red to small heated pens located at the University of Florida Racing Lab. Behavioral Analyses Animals were tested for the presence of both reflexive and volitional visual behaviors every 3 to 10 days for six weeks. Reflexive visual responses were assessed using an optokinetic nystagmus (OKN) paradi gm. The OKN reflex is driven primarily by
24 visual stimuli processed by the pe ripheral regions of the retina. The reflex is manifest as a compensatory head movement of the animal in an attempt to fixate a moving stimulus. The stimuli were two high contrast vertical s quare wave gratings with spatial frequencies of 0.065 or 0.26 cycles degree-1 (bar widths of either 5 cm or 1.25 cm, respectively). Stimuli were presented in the form of a ro tating drum (average speed 14.6 rpm) and the animals were held stationary in the center of the drum while it was rotated in both clockwise and counterclockwis e directions. Movement during periods when the drum was not rotating served as a reference point for evaluation of the behaviors elicited by the moving stimuli. A positive OKN response was characterized by a smooth head movement in the direction and at the speed of stimulus rotation followed by a rapid head movement in the opposite direction. Behavi or was recorded (Nikon Coolpix Digital Camera and Video Recorder) and analyzed fo r the presence of OKN responses using a 0 to 3 scoring system: 0 â€“ no OKN response; 1 â€“i nconsistent or unidirectional responses to the lower spatial frequency grating; 2 â€“ cons istent bidirectional responses to the lower spatial frequency grating; 3 â€“ consistent bi directional responses to the higher spatial frequency grating. Volitional visual responses were assessed by placing animals in a testing environment that contained novel visual st imuli including colored candies, aluminum foil, and paper with printed black dots. Volitional visual behavior is driven primarily by stimuli processed by the central/foveal regions of the retina. The behavior is driven by interest in the environment. Animals were e xposed to the testing environment for periods of approximately three to five minutes and behavior was video r ecorded and analyzed and using a 0 to 3 scoring system: 0 â€“ no visu al behavior with random head drift; 1 â€“ no
25 random head drift detected, evidence of orie ntation to surroundings; 2 â€“ exhibits head movements coordinated with the presence and/or movement of visual stimuli; 3 exhibits pecking of visual ta rgets and stimuli. Electroretinographic (ERG) Analyses ERGs were recorded by Drs. SG Jacobson, TS Aleman and AV Cideciyan from the University of Pennsylvania with the assi stance of members of the Semple-Rowland laboratory. The pupils of dark-adapted (>12 hr ) animals were dilated (repeated topical administration of vecuronium bromide, pr oparacaine HCl and benzalkonium chloride; and tropicamide) over a 30 minute period prior to the recordings. The animals were then anesthetized using a mixtur e of ketamine HCl (10 mg/ kg) and xylazine (2.5 mg/kg) delivered intramuscularly. A quarter of the initial dose was given as needed during the recording session to maintain anesthesia. Anesthetized animals were placed in a supine position with their heads resting on a head holder. Full field ERGs were recorded from the right eye of each animal using custom made contact lens electrodes (Hansen Ophthalmics, Iowa City, IA). An eyelid specu lum was used and the el ectrode was held in place by a stereotaxic apparatus. A platinum needle placed in the skin above the eye served as reference. ERGs were recorded using a commercially available ganzfeld and computer-based system (ColorDome and Espion Console, Diagnosys LLC, Littleton, MA). Recordings began with dark-adapted ERG luminance-response functions elicited with increasing intensities of wh ite flashes (-3.2 to +0.8 log cd.s.m-2; 0.5 log unit steps; 2 sec interstimulus interval; digital filter di sabled). For low intensity stimuli, 4-10 responses were recorded and averaged. For th e highest intensity s timuli, 2-6 responses (>15 sec interstimulus interval) were reco rded. Upon completion of the dark-adapted stimulus series, the animals we re light-adapted to a 30 cd.m-2 white background and the
26 ERGs elicited by 29Hz flicker s timulation (white; +0.8 log cd.s.m-2) were recorded and averaged (20 responses). In some cases, 250750 responses were recorded to detect submicrovolt amplitude flicker ERGs. The amplitudes of the ERG waveforms were measured conventionally: a-waves were measur ed from baseline to the trough; b-waves were measured from baseline or from th e a-wave trough to the positive peak; 29Hz flicker amplitudes were measured from trough to peak. Sacrifice and Tissue Collection The animals were sacrificed within one w eek of the last behavioral testing period and ERG recordings. Animals were anesthetiz ed with an intramuscular injection of ketamine (16 mg/kg) and euthanized using a protocol approved by the University of Florida Institutional Animal Care and Use Committee. The eyes were then enucleated and the anterior segment and vitreous body of each was removed. The posterior eye cup of the right eye was fixed overnight in 4% paraformaldehyde (PFA) at 4 C. Once fixed, the eye was bisected along the inferior / supe rior midline axis and one half was processed for semi-thin plastic histology and the other ha lf was processed for frozen sectioning and genomic DNA extraction. The left eye cup wa s bisected in the same way and equal portions of the retina, retinal pigment ep ithelium and choroids were removed, were placed in sterile, sealed tubes, flash frozen in liquid nitrogen and stored at -70 C in preparation for RNA analyses. To permit use of the right and left retina s of treated animals in our analyses, we determined if delivery of the virus via th e neural tube produced similar transduction percentages in both eyes. Three embryos were injected with 0.5 l of pTYF-EF1 -PLAP lentivirus (5x109 TU/ml) to compare the pattern and pe rcent transduction of left and right retinas. On embryonic day 10, the embryos were sacrificed an d the retinas were
27 processed for PLAP staining and analyzed as described previously (Coleman et al., 2002; Coleman et al., 2003). The patterns and percent transduction values for the left and right retinas of individual animals were si milar (41%, 39%; 18%, 13%; 9%, 15%). The variability observed in the percent trans duction between animals was dependent on the quality of the embryonic injection and decrea sed as percent transduction increased. No striking interocular asymmetry was observed in the percent viral transduction within individual animals. Retinal Immunohistochemistry and Light Microscopy Following fixation in 4% paraformaldehyde , the right eye of each animal was bisected along the superior / inferior mid line axis: the temporal portion of each eyecup was processed for immunohistoc hemical analyses while the nasal portion was processed for semi-thin plastic analyses. Tissues de signated for immunohist ochemical analyses were cryoprotected by soaking overnight in a 30% sucrose (wt/vol)-PBS solution, sectioned (14 m), and stored at -30C un til stained. Prior to immunostaining, the tissues were dried overnight, rinsed in PBS, a nd permeabilized and blocked for 1 hour in PBS containing 0.3% Triton X-100, 1% BSA, and 10% goat serum. GFP was detected using a polyclonal antibody (generously provided by W. Clay Smith, University of Florida, Gainesville, FL) diluted 1:500 in primary dilutio n buffer (PBS containing 1% BSA and 10% goat serum). Sections were incubate d with the primary antibody overnight at 4C.The primary antibody was visualized by labeling with a goat anti-rabbit IgG secondary antibody tagged with the Alex a-488 fluorophore (Molecula r Probes) diluted 1:500 in primary dilution buffer. Sections we re counterstained with DAPI, mounted in Gel/Mount aqueous media (Biomeda Corp., Fost er City, CA), coverslipped, and sealed with Permount resin. Tissues designated for semi-thin plastic analyses were dehydrated
28 through a graded series of ethanol solutions (50%, 70% and 80%) and embedded in JB-4 Plus (Electron Microscopy Sciences, Hatfie ld, PA) using the manufacturerâ€™s protocol. Plastic embedded tissues were sectioned (1.5 m ), stained with 1% to luidine blue in 1% (wt/vol) sodium borate, and coverslipped using Permount resin. Immunohistochemical and plastic sections were examined and photographed using a Zeiss Axioskop 2 plus fitted with a Spot image acquisition system (Diagnostic Instruments, Inc., Sterling Heights, MI). The immunostained and semi-thin plasti c sections were analyzed to obtain information regarding the extent of viral transduction and th e effects of the treatment on retinal morphology, respectively. We were unable to estimate the percent viral transduction of the retina by examining flat-m ounted retinas becaus e GFP expression was too low to allow direct visualization of th e transduced cells. Low protein expression is frequently observed from the second cist ron of IRES-based bicistronic expression cassettes in vivo (Mizuguchi et al., 2000). Info rmation about the distribution of transduced cells was obtained by analyses of GFP immunofluorescence of serial sections cut along the superior-to-inferi or axis of the right eyes of the treated animals. Immunostained serial retinal sections were each divided into four regions relative to the optic nerve and were designated far superior (FAR SUP), superior (S UP), superior optic nerve (SON) and inferior optic nerve (I ON). Each region was assigned a percent transduction score by an independent obser ver that was based on the number of GFP positive cells within the region. These scor es were plotted to yield 3-D graphic representations of tr ansduction across the 500 m retinal expanse using SigmaPlot v8.0. To assess the effects of treatment on reti nal morphology, the total width of the retina
29 extending from the outer limiting membrane to the ganglion cell layer was measured at 100 m intervals within each of the four reti nal regions. Measurements were made from digital images of representative semi-thi n retinal sections us ing Adobe Photoshop v7.0. For each region, the retinal widths of the tr eated and wild-type retinas expressed as percent change relative to the average widt h of the GUCY1*B untr eated retina in the corresponding region. Genomic and Reverse Transcription PCR (RT-PCR) Genomic DNA was extracted from 25 mg of retina / pigment epithelial tissue taken from the right eye that had been fixed a nd cryoprotected but not sectioned. The tissues were soaked in PBS to remove the sucrose and DNA was extracted from the tissue using a DNAeasy kit (Qiagen, Valencia, CA). Total RNA was extracted from retina / pigment epithelium tissue that had been removed from th e left eye and stored at -70C. The frozen tissue was pulverized under liquid nitrogen and RNA was extracted using an RNeasy kit (Qiagen, Valencia, CA) according to the ma nufacturers recommended protocol that included treatment of the RNA samples with RNA-free DNAse to remove trace quantities of genomic DNA. Known copy numbers of pTYF-EF1a-GC1-IRES-eGFP plasmid DNA ranging from 3,000 to 300,000 copies were prepared and amplified in parallel with the genomic DNA samples. The plasmid DNA standards, genomic DNA, and total RNA were amplified using primers that spanned the IRES-eGFP elements present in the lentiviral transgene (sense: 5â€™-TTT CCC CGG TGA TGT CG T; antisense: 5â€™-GCC GGT GGT GCA GAT GAA). The RT-PCR analyses include d amplification of chicken -actin mRNA (sense: 5â€™-TGC TGC GCT CGT TGT TG; antisen se: 5â€™-GTC ACG GCC AGC CAG AT) to control for the quality and quantity of RNA in each sample and the efficiency of the RT reaction.
30 PCR amplification of standard and genomic (0.5 g DNA template) DNA was carried out in 50 l reactions. RT-PCR reactions were carried out in two steps. Total RNA (1 g) was reverse transcribed in a 50 l reaction volume. Aliquots of the RT reaction were then amplified for GC1 (15 l) and -actin (3 l) transcripts in se parate reactions (50 l final volume). Components for the genomic and RT -PCR reactions were obtained from an Ampli Taq Gold RT-PCR kit (Applied Biosystems, Foster City, CA). The reaction parameters used to amplify the GC1 transgene and its transcript were 95C â€“ 2 min; 95C â€“ 1min, 61C â€“ 1 min, 72C â€“ 1 min (x35 cycles ); 72C â€“ 10 min; 4 C soak. The reaction parameters used to amplify the -actin transcript were the same as above except that amplification was carried out for 30 cycles. Aliquots (20 l) of the PCR reactions were separated on a 1% agarose gel containing 32 nM ethidium bromide and photographed and analyzed using a Gel Doc 1000 system and Quantity One software (BioRad). The RT-PCR analyses were repeated three times . For each RT-PCR trial, the quantity of transgene mRNA in each sample was normalized to the average amount of -actin mRNA present in the samples. The number of integrated vector transgenes in the genomic DNA extracted from the treated tissues was estimated by compari ng the amount of product obtained in these reactions to that obtained from PCR am plification of known copy numbers of the respective plasmid DNA. The genomic DNA and standard reactions were amplified and analyzed under identical condi tions and were repeated thr ee times. When imaging the PCR gels, care was taken to insure that the signals were below satu ration. The intensity values for the standards obtained from three i ndependent trials were plotted and analyzed using SigmaPlot 8.0 (SPSS, Inc.). The number of vector transgene copies present in the
31 genomic DNA samples was calculated using th e equation for the best-fit sigmoid curve. These values were converted to vect or transgene copies per genome copy (transgenes/genome) using the value of 380,000 genome copies / 0.5 g chicken DNA (Gregory, 2005). Statistical Analyses The morphological data obtai ned from the semi-thin plas tic sections was analyzed using a one sample t-test to determine if th e mean percent change in retinal thickness of treated animals relative to untreated GUCY1*B controls was signifi cant in the four retinal regions examined. Results In vivo Lentiviral Treatment Restores Optoki netic and Volitional Visual Behaviors Seven GUCY1*B embryos treated with EF1 -GC1-IRES-eGFP lentivirus survived to hatch and were used in this study. Vision of treated animals was assessed behaviorally through examination of optokinetic reflexes a nd volitional visual behaviors. These tests allow assessment of visual function associat ed with different retinal regions: the OKN test requires function of the pe ripheral retina while the volit ional test measures function of the central/foveal retina (C onley and Fite, 1980; Komenda and Fite, 1983; Schmid and Wildsoet, 1998). Of 7 treated animals, 6 exhi bited varying degrees of sighted behavior over the course of the 4-5 week testing pe riod. One treated animal failed to exhibit optokinetic or volitional visu al behaviors; electrophysiolo gical tests were conducted on this animal, but its retinas were not proce ssed for molecular or morphological analysis. The other six treated animals exhibited robust OKN responses to the two different spatial frequencies tested with an overall group m ean score of 2.23. Volitional, sight-directed pecking behavior was observed in treated an imals as early as 3 days post-hatching. They
32 also exhibited high levels of e xploratory behavior and were ab le to peck at a variety of objects within their visual fields at later tim e points. The mean volitional behavior score for the 6 treated animals exhibiting rescue d sight over the entire study period was 2.05. On the final day of testing, 5 of the 6 tr eated animals received OKN scores of 3 and volitional behavior scores of e ither 2 or 3. Visual behavior sc ores of the remaining animal dropped during the last week of the study. At th e time of sacrifice, this animal did not show any evidence of volit ional sight. Wild-type and untreated GUCY1*B animals received mean scores of 3.0 and 0, respectively, on both the OKN and volitional behavior tests. A summary of the visi on test results for the 6 treate d animals exhibiting sight is shown in Figure 3-2. Retinal Electrophysiology also Indica tes Treatment Restores Function Electroretinography (ERG) conducted unde r both darkand light-adapted conditions was used to assess photoreceptor-med iated retinal function in wild-type, and untreated and treated GUCY1*B chickens 3-4 da ys prior to sacrifice. All animals were 31-37 days of age at the time of testing. ERGs in wild-type a nd untreated GUCY1*B chickens differed dramatically (Figure 3-3 A and B). Dark-adapted untreated GUCY1*B chickens presented with no measurable ERG re sponses regardless of stimulus intensity. Conversely, dark-adapted wild-type animal s had normal ERG responses with a-wave (photoreceptor origin) and b-wave (bipolar cell origin) components that increased in amplitude with increasing stimulus intensit y (Figure 3-3 A). Sim ilarly, light-adapted, untreated GUCY1*B animals had no detectab le flicker ERG responses while lightadapted wild-type chickens had large amp litude cone-mediated ERG responses when presented a 29 Hz flicker stimulus in th e presence of rod-desensitizing background illumination (Figure 3-3 B). Treatment of GUCY1*B animals with EF1-GC1-IRES-eGFP
33 Figure 3-2. Optokinetic and vol itional behavior tests indicate treatment efficacy. A. Optokinetic reflex (OKN) exhibited by 21-day-old, wild-type RIR chicken. Two frames of video (4.3 sec between frames) are shown illustrating the head movement observed in birds in response to counterclockwi se rotation of the 0.26 cycles degree-1 (bar width =1.25 cm) stimulus. B. Volitional visual behavior exhibited by treated GUCY1*B animal 2 on day 7. Two frames of video are shown illustrating the animalâ€™s abilities to perceive and peck at objects within its visual field. C and D. Graphic summaries of the behavioral test results for the 6 treated animals th at exhibited sighted behavior. The left series of graphs show the optokinetic sc ores (0-3) for the 6 treated animals as a function of age. The right series of graphs shows the corresponding scores obtained by these animals for the volitional behavioral tests.
Figure 3-3. Retinal electrophysio logy rescued in treated ey es. A. Comparison of darkadapted ERGs in response to incr easing intensities of light in a GUCY1*B/GUCY1*B animal injected with EF1 -bGC1-IRES-eGFP (Treated) compared to a control (Untr eated). ERGs in untreated animal are non-detectable in contrast to the sizeable ERGs evoke d in the treated animal. Results from a wild-type animal are s hown in the left column for comparison. B. Light-adapted 29 Hz flicker ERGs in the same animals as shown in (A) demonstrate restoration of respons es after treatment. C. Overlapping waveforms are ERGs elicited by 0.8 log cd.s.m-2 white flashes presented in dark-adapted (DA 1Hz) and light-adapted (LA Flicker) states in all treated animals compared to wild-type controls . Functional rescue was observed in five (top waveforms) of the seven tr eated animals, whereas two (bottom waveforms) showed responses indis tinguishable from noise. D. Summary parameters of dark-adapted photorecep tor (a-wave) and post-photoreceptor (bwave) function, as well as light-adapted f licker amplitude in treated animals as compared to untreated and wild-type animals. Five of 7 treated animals showed amplitudes substantially larger than untreated animals but smaller than wild-type controls (gray symbols; mean SD).
36 lentivirus restored retinal function (Figure 33 A and B right). Five of 7 treated animals presented with ERG responses to single flas hes under dark-adapted conditions and to flickering stimuli under light-adapted conditions . The shapes of these responses were similar to those generated by wild-type anim als but with lower amplitudes. Two of the treated animals had no detectable responses (Figure 3-3 C). The amplitudes of the ERG a-waves in the five responding animals (6.6 1.3 V; mean SD) were 6% of the wildtype response (105.8 36.0 V; mean SD) suggesting that phototransduction had been restored in a subset of photoreceptors. The tw o treated animals that did not exhibit visual behavior failed to produce ERGs that were distinguishable from noise (Figure 3-3 D). The Number of Integrated Transgenes per Genome Is Related to Treatment Efficacy Quantitative genomic PCR was carried out on DNA extracted from the right eyes of the 6 treated animals that exhibited vi sual behavior to obtain a measure of the efficiency of the viral treatment. The retina s of the seventh animal , that did not exhibit evidence of sighted behavior following treatm ent, were not processed for molecular or morphological analyses. The DNA samples were analyzed on 3 different days and each reaction set included amplification of DNA standards containing known copy numbers of the transducing vector. The primers were designed to amplify a 638 bp product that spanned the IRES-eGFP elements within the transgene (Figure 3-4 A). The lentiviral transgene was detected in all 6 samples obt ained from the treated animals; no product was amplified from untreated GUCY1*B, wild-type RIR, or water control reactions (Figure 3-4 B). The amount of PCR product obtained in a ll experimental samples fell within the linear porti on of the standard amplification curve (Figure 3-4 C) used to calculate the number of copies of the viral transgene within each sample. The estimated
37 number of integrated viral transgenes per ge nome in the retinas of the 6 treated animals ranged from 0.12 to 0.02 with a mean value of 0.07 0.01(mean SEM). A noteworthy finding is that the animal w ith the lowest transgene copy number was the animal that scored poorly on the visual behavior test s toward the end of the study and had no measurable ERG responses. Figure 3-4. Quantitative genomic PCR estimate of integrated viral trangenes per genome. A. Diagram of integrated viral transgene and lo cation of PCR primers spanning the IRES-eGFP junction that were used to amplify the transgene. B. Ethidium bromide stained gel sh owing 638 bp product amplified from the experimental and standard reactions. Lanes 1-6 contain samples amplified from the six treated animals exhibi ting restored sight. No product was observed in untreated GUCY1*B (lanes 7, 8) or wild-type chickens (lane 9). A blank water control is shown in lane 10. Lanes 11-14 contain product amplified from 300000, 150000, 30000, 15000 copies of the transducing vector. DNA size ladders, 1kb and 123bp, were run to the left of the samples and standards. C. Plot of amount of product obtained from the standard PCR reactions (black) best fit with a si gmoid curve (f=y0+a/(1+exp(-(x-x0)/b))^c with r2=0.999 . The products obtained from each experimental sample (red) all fell within the linear portion of the am plification curve and the number of transgenes present in each sample wa s calculated using th e equation for the best-fit curve. Individual values plotted are mean SEM (n=3). RT-PCR Analyses and GFP Immunostaini ng Confirm Transgene Expression RT-PCR analyses were carried out to examine transgene expression in the retinas of treated animals. Amplificat ion of aliquots of reverse tran scribed RNA revealed that the
38 transcript derived from the viral transgene wa s present in all retinal samples from the six treated animals with rescued vision. The 638 bp PCR pr oduct for the transgene was not detected when the reverse transcription step was omitted from the procedure, nor was it detected in GUCY1*B untreat ed or wild-type retina samples. All samples contained approximately equal amounts of -actin mRNA as determined by the staining intensity of the 538 bp -actin product. The amount of transgene and -actin transcript in each sample was quantified and the amount of transgene mRNA was normalized to the average amount of -actin mRNA detected ac ross all samples run in th at particular assay. The relative number of transgene mRNA copies presen t in each of the experimental samples, as determined from three independent assays , ranged from a high of 144 to a low of 19 with a mean value of 94 12 (mean SEM) . In general, transgene mRNA levels correlated with estimates of the number of integrated tran sgenes per genome within animals. Variations in these measures were likely due to sampling variability arising from the random distribution of transduced cells a nd use of different tissue samples for these analyses. The distribution of transdu ced cells within treated retinas was determined by examining serial sections of the right ey ecup taken along the vertical meridian. Each section was divided into four contiguous regi ons and the percent of retinal cells positive for GFP was estimated for each region (Figure 3-5 A). In all retinas examined, greater
Figure 3-5. Immunohistochemical analyses of GFP expression in treated retinas. A. Schematic of the right eye cup that shows how the retina was apportioned for analyses. The right eye of each anim al was bisected along the vertical meridian. One half was processed for immunohistochemical analyses and the other for detailed histologi cal analyses. The retinal se ctions were divided into four regions (FAR SUP, SUP, S ON, ION) to simplify analyses. B. Flat-mount retinas stained to reveal pattern of tran sduction obtained foll owing neural tube delivery of 0.5 l EF1a-PLAP virus (109 TU/ml). The percent transduction for the left (left panel) and ri ght (right panel) eyes of 2 of the 3 animals analyzed is shown. C. Far superior region of a treated retina immunostained for GFP. Arrows indicate staining in photorecep tor cell bodies. GFP staining is also visible in the IS and OS of these cells. D. Topographical distribution of GFP expressing cells in treated right eyes from treated animals 2 (left) and 1 (right). The percent transduction of each of the four retinal regions was plotted on the z-axis as a function of location along the superior-inf erior axis of the eye (x-axis) and distance from the mid line (y-axis). These analyses represent the results obtained from serial secti ons over 500 m beginning at the midline axis and moving laterally.
41 than 90% of the GFP staining was localized to the photoreceptor cell layers (Figure 3-5 C). Occasionally, a complete cell column wa s stained. These cell columns, which were observed more frequently in treated retinas analyzed prior to or just after hatching (Coleman, 2003), reflect passage of integrat ed transgenes from transduced retinal progenitor cells to their daughter cells during development (data not shown). The estimates of percent GFP staining observed in each 14 m section were used to create a topographical map of the percent G FP staining observed within the 500 m region for each animal. The results obtained for 2 of the 6 treated animals (Figure 3-5 D, animals 1 and 2) illustrate the variation observed in th e distribution of trans duced cells in the 500m region selected for these analyses. The estimates for the number of integrated transgenes per genome for animals 1 and 2 were 0.08 and 0.07, respectively. The spatial distribution of transduced cells observed in th e retinas of these animals was variable over the extent of the retinal area; similar to th at observed in our analyses of flat-mounted EF1 -PLAP treated retinas (Figur e 3-5 B). Based on the percent transduction estimates obtained for retinas treated with the EF1 -PLAP virus, a conservative estimate for the percent transduction of the re tinas treated with the EF1 -GC1-IRES-eGFP virus would be 15-40%, the actual value depending on the quality of the embryonic injection. Morphometric Analyses Suggest that Treatment Slows Retinal Degeneration Retinal degeneration in this animal m odel is morphologically evident beginning around two weeks of age and is marked by decreased numbers of rod and cone photoreceptor outer segments and gaps innerv ating inner segments and cell bodies of the outer nuclear layer. By three weeks of ag e, pycnotic nuclei are present in the outer nuclear, inner nuclear and ganglion cell la yers; pycnotic nuclei are evidenced by chromatin condensation characteristic of ce lls undergoing apoptotic, or programmed, cell
42 death (Ulshafer et al., 1984). Degeneration is most evident in the central retina at early stages of degeneration and spreads periphera lly with very few photoreceptors remaining intact across the retina by 6-8 months posthatch (Ulshafer and Alle n, 1985). Sections of retinas of the right eyes of treated animals were compared to those of age-matched wildtype and GUCY1*B untreated retinas to dete rmine if the viral treatment altered the histopathology. The results suggested that treatment may have slowed retinal degeneration in some animals but it did not pr event degeneration (Fig ure 3-6 A). The best preservation of the retina was observed in an imal 1 as evidenced by increased numbers of nuclei within the outer nuclear layer and incr eased thickness of the inner plexiform layer relative to untreated GUCY1*B retina. In contra st, animal 6 had severe loss of cells from the outer nuclear layer across all regions examined and the retinal pigment epithelium had changes characteristic of the late-stage degeneration usually observed in untreated animals (Ulshafer et al., 1984). The retinal changes were quantified by measuring the distances between the outer lim iting membrane and the gangli on cell layer within each of the 4 regions sampled. These measures were expressed as percent change in retinal thickness relative to that observed in untre ated GUCY1*B retinas (Figure 3-6 B). The results of these analyses revealed that the th ickness of all of the treated retinas relative to untreated retinas was significantly great er in the SON [t(5)=2.9, p<0.05] and ION [t(5)=3.5, p<0.05], the two regions of the reti na that normally undergo the most rapid degeneration in untreated GUCY1*B retinas (U lshafer et al., 1984) . The SUP region of the retinas of four of the si x treated animals also showed signs of slowed degeneration but these changes were not found to be si gnificant. Analyses of the FAR SUP region revealed that one treated animal showed signs of unusually severe retinal degeneration in
43 Figure 3-6. Comparison of retin al morphology of treated GU CY1*B, wild-type RIR, and untreated GUCY1*B. A. The morphology of the retinas of treated animals was examined to determine if the viral treatment had affect ed the course of retinal degeneration. Regions were analyzed along the vertical meridian of the right eyes of experimental and contro l animals; representative micrographs from the locus superior to the optic nerve (SON) are shown from wild-type, untreated GUCY1*B and 3 treated retinas (animals 1, 5 and 6). RPE â€“ retinal pigment epithelium; OS â€“ outer segments; IS â€“ inner segments; ONL â€“ outer nuclear layer; INL â€“ inner nuclear laye r; IPL â€“ inner plex iform layer; GCL â€“ ganglion cell layer. B. Relative percent change in retinal thickness of treated animals. Retinal width (m) from the outer limiting membrane to the ganglion cell layer was measured at 6 loci (100 m apart) within the regions examined. These widths were expressed as percent change relative to average widths of the untreated GUCY1*B retinas in thes e regions which were set to zero for graphing purposes. A simple one group ttest (null hypothesis that treatment groups are not different from untreated GUCY1*B) was used to analyze the results. The results of this test show ed that retinal thickness of the treated animals in the SON and ION regions we re greater than those of untreated animals (p<0.05). Four of 6 treated anim als also showed evidence of slowing of degeneration in the SUP region but th is was not statistically significant. The key for animal number is shown in the upper left of the graph.
44 this region; this region in the remaining five treated animals was not significantly different from untreated retinas. Discussion This study is the first to demonstrate rest oration of sight in an animal model of LCA1. We successfully used a lentiviral vector to deliver the therapeutic transgene EF1aGC1-eGFP to retinal progenitor cells. Expr ession of normal copies of GC1 in photoreceptor cells was sufficient to restor e cell function, measured by ERG, and, most importantly, generate robust sighted behavior . Overall treatment efficacy was indicated by a combination of sighted behavior, ER G measures, retinal morphology, estimate of integrated viral transgenes per genome and transgene expression in photoreceptor cells. High visual performance scores and ERG measures correlat ed with better preserved retinal morphology and estimates of integr ated transgenes. Similarly, low visual performance scores and ERG measures correlat ed with more degenerated retinas and less integrated transgenes. For gene therapy treatment of an autoso mal recessive retinal degenerative disease to be complete in its effectiveness, expr ession of the therapeutic gene must produce functional protein which rest ores function to disabled cells and prevents retinal degeneration. Our treatment restored function at cellular and behavi oral levels, but was insufficient to prevent retinal degenerati on, though in some cases degeneration was slowed. Of particular interest is treated animal 6 which presented with a decline in open field response coincident with complete loss of OKN reflex behavior during the last set of behavior tests (Figure 3-2 C and D). This an imal also had the most severely degenerated retina of the six treated animals (Figure 3-6 A). A similar decline in visual behavior was observed in four treated animals (data not show n) which were sacrificed at three to five
45 months of age. Each of these four animals exhibited a decline in visual performance, similar to that observed for animal 6, presenting as early as six to eight weeks of age. By four to five months of age, both open fiel d and OKN responses were absent in each of these animals. Hence, it is vital to prevent or limit retinal degenerati on in order to sustain restored vision in treated animals. The patchy transduction pattern generated by our treatment (Figure 3-5 B) leads to an intermingling of non-treated dysfunctional photoreceptor cells wi th treated functioning cells. Thus, a â€œbystanderâ€ eff ect or non-autonomous cell degeneration may be at play. This â€œbystanderâ€ phenomenon suggests that cross talk between phot oreceptor cells or environmental cues play a significant role in the progressive nature of photoreceptor degeneration (Travis, 1998; Ripps, 2002). This is evident in many i nherited photoreceptor diseases linked to genes whose e xpression is limited to rod cells (http://www.sph.uth.tmc.edu/Retnet/) but lead s to degeneration of both rod and cone photoreceptor cells. Further, chimeric retinas of wild type and tran sgenic mice expressing mutated rhodopsin exhibit uniform cell deat h, however retinas of chimeric animals degenerated more slowly than those of transgenic mice (Huang, 1993; Travis, 1998) suggesting a relationship between the number of cells expressing the mutated gene and retinal morphology over time. Degeneration of photoreceptors expressing abnormal proteins had a negative impact on the ability of otherwise healthy cells to survive, a trend that was tempered in retinas containing higher percentages of normal cells. If the ratio of non-treated to treated cells is high in our treated animals, then degeneration of the nontreated cells could be adversely affec ting survival of th e treated cells.
46 The simplest approach to achieve long-term restoration of sight in our paradigm may be to increase the total number and/or density of photoreceptor cells transduced by the viral vector. By increasing the ti ter of the injected virus from 109 to 1010 TU/ml, for example, we can effectively increase the pe rcent transduction of the photoreceptors from approximately 40% to 85% (Coleman et al., 2003). Another modification to our therapeutic strategy that may improve treatm ent efficacy is use of photoreceptor-specific promoters to drive transgene expression in our viral vector . Limiting expression of GC1 to photoreceptors could improve treatment effectiveness by eliminating any untoward effects induced by ubiquitous GC1 expressi on. A future strategy might also include bicistronic therapeutic transgenes that not onl y encode GC1 but also encode factors that improve photoreceptor viability or prevent ce ll death. Several member s of the fibroblast growth factor family (Lau et al, 2000; Green et al, 2001), ciliary neurotrophic factor (Liang et al, 2001) and the apoptotic inhibito r, bcl-2 (Chen et al , 1996; Bennet et al, 1998) have been shown to delay photoreceptor degeneration in models of inherited retinal disease. Recently, a factor secreted by rod cells has been identified that appears to support cone cell viability (Leveillard et al., 2004).
47 47 CHAPTER 4 CURRENT AND FUTURE STUDIES We have shown that lentiviral deliv ery of the therapeutic transgene EF1 -GC1IRES-eGFP is sufficient to rescue vision in the GUCY1*B chicken, an animal model for LCA-1, but not effective in preventing photore ceptor degeneration. In order for delivery of the GC1 transgene to ultimately rescue vi sion in the GUCY1*B chicken, the treatment must preserve photoreceptors as well as restore function. Furt her, before attempting to preserve these cells in treated animals, it is necessary to understand factors contributing to cell death in both the untreat ed and treated retinas. There are several possible explanations for the persistence of cell death in treated retinas. First, the therapeutic protein, bovG C1, could be dimerizing with or otherwise upregulating the mutated GC1 protein in GUCY1*B retinas, subsequently causing photoreceptor cell death. However, this is an unlikely explanation. The mutated GC1 protein is degraded soon after translation a nd cannot be detected in a Western blot and Rhode Island Red chickens heterozygous for the mutation present with a normal phenotype (Semple-Rowland et al., 1998). Another possibility is that GC1 overexpression and expression in cells other than photoreceptors is contributing to retinal degeneration. This possibility can be tested by treating retina s of wild type Rhode Island Red chickens. If overexpression or non-targeted expression of GC1 c ontributes to retinal degeneration, then degeneration would occur in otherwise normal retinas of these animals.
48 The most likely explanation for persistent degeneration in treated animals is the non-autonomous nature of photor eceptor cell survival. It is now well documented that the death of subsets of photoreceptor cells has a negative impact on the survival of neighboring cells (Travis, 1998; Ripps, 2002). Our curren t treatment transduces a subpopulation of the total number of photoreceptor cells, with less than half of the target cell group expressing bovGC1, more of these cells remain non-functioning and die having a negative impact on the entire photoreceptor cell population causing degeneration. While the critical number of transduced cells required to overcome the effects of neighboring cell death is not known, it is likely that our current treatment does not reach this threshold. The goals of current and future studie s include an examination of variables associated with retinal degeneration in bot h treated and untreated GUCY1*B chickens and ultimately extension of cell survival in treated animals with rescued vision. Initial studies have focused on delineating the mechan ism of cell death and analyzing the effects of factors associated with transgene deliver y and expression in treated animals. Future studies include examination of the relations hip between the amount of viral transduction in the retina and rescued vi sion and developing a treatment upon delivery of which retinal degeneration in treated animals is de layed and rescued vision is extended. Apoptotic Cell Death in the Degenerating Retina There are two types of cell death: necrosis , which results from injury or physical insult, and apoptosis, programmed cell death. Ne crosis is characterized by cell membrane damage, swelling of the cell and mitochondria and an inflammatory response (Alberts et al., 2002). Apoptosis plays an important physio logical role in tissue development and maintenance of adult tissues. Morphol ogically, it is characterized by chromatin
49 condensation, DNA fragmentation, plasma me mbrane blebbing and cell shrinkage (Gavrieli et al., 1992; Reed, 2000). These morphological changes are induced by activation of a protease signaling cascade. The primary proteases involved are members of the cysteine aspartyl-specific protea se family or caspases. Recently, caspase involvement in apoptosis has come under sc rutiny as examples of apoptosis have been identified that do not require caspase activa tion (Doonan, 2003). It is likely that multiple pathways are capable of initiating the apoptotic signaling cascade. Apoptosis appears to be the cause of photoreceptor cell death in many forms of congenital retinopathy (Chang et al., 1993; Lolle y et al., 1994; Portera-Cailliau et al., 1994; Tso et al., 1994). For example, retinas of multiple mouse models of retinitis pigmentosa exhibit DNA fragmentation a nd positive TUNEL labeling indicating that apoptosis is ultimately responsible for cell deat h in these models (Portera-Cailliau et al., 1994). The mechanism underlying photoreceptor ce ll death in the un treated GUCY1*B chicken has not been examined using modern technologies that in clude TUNEL staining, immunohistochemistry and the DNA ladder assa y. However, the presence of pycnotic nuclei in the degenerating retinas of this an imal model are indicative of apoptotic cell death (Ulshafer et al., 1984). We hypothesize, based on the results of studies of cell death in several other animal models of inherite d retinal disease (Chang et al., 1993; Lolley et al., 1994; Portera-Cailliau et al ., 1994; Tso et al., 1994) and th e studies of Ulshafer et al. (1984), that photoreceptor loss in the GUC Y1*B chicken occurs via apoptotic mechanisms. Many laboratories are examini ng the possible use of anti-apoptotic gene therapy, or gene therapy deliv ering apoptotic blockers, in treating retinal degeneration.
50 For example, adenoviral delivery of bcl-2, a known apoptotic blocker, was shown to slow retinal degeneration in the rd mouse (Bennett et al., 1998). Pr eliminary studies of gene transfer of neurotrophic f actors such as, CNTF, GDNF and BDNF have also proven promising in slowing retinal degeneration (Chaum and Hatton, 2002; Delyfer et al., 2004). In preliminary studies, we have used TUNEL assay, DNA ladder assay and an antibody for activated caspase 3 to determine if these assays can detect apoptotic cells in degenerating retinas of GUCY1*B chickens. Multiple stages of degeneration have been examined in animals ranging in age from 3 to 84 days. To date, we have been unable to detect apoptotic cells in these retinas. Noneth eless, we still suspect that apoptosis is the mechanism of retinal cell death in the dege nerating retinas of GUCY1*B chickens. The retinas of these animals do not exhibit mor phological signs of infl ammation that include cell and mitochondrial swelling, both of which ar e characteristic of necrotic cell death. A possible explanation for our inab ility to detect apoptosis in preliminary studies is the relatively slow rate of cell loss from these retinas over time. Without multiple cells undergoing apoptosis concurrently it could be difficult to detect individual cells undergoing apoptosis. If apoptosis is determined to be the mechanism of cell death in the retinas of treated GUCY1*B chickens, lentiv iral delivery of apopt otic blockers or neurotrophic factors may ex tend cell survival in chic kens with rescued vision Effects of Transgene Expression On Photoreceptor Viablility In addition to determining the mechanism of cell death in the untreated GUCY1*B retina, it is also important to determine if expression of the len tiviral transgene, EF1 GC1-eGFP, has any negative effects on photor eceptor viability and retinal integrity. Initial experiments are focused on wild-type Rhode Island Red chickens whose retinas do
51 not undergo degeneration. To date, two wild-t ype Rhode Island Reds, now eleven weeks of age, treated with the lentiviral transgene, EF1 -GC1-eGFP, have retained vision as evidenced by high scores on behavioral tests. One treated wild-type animal, which also scored high on visual assessments, was sacrifi ced at six weeks of age. Retinal tissue is being processed for this animal. Preliminary ex amination suggests that the retinas of this animal are morphologically intact . Behavior testing of treate d wild-type animals at later time points, ERG measurements and further pr ocessing of retinal tissue will reveal any negative effects that our treatment may have on retinal morphology and function. It may also be desirable to replace the EF1 promoter that we used in our first rescue study with one that limits expressi on of the transgene to photoreceptor cells. We chose to use EF1 , a ubiquitously expressed promoter, to drive expression of GC1 and the eGFP reporter in our initial experiments because previous analyses of its expression characteristics showed that it would produce sufficient GC1 protein to induce a change in photoreceptor function. The non-sp ecific targeting of expre ssion of the transgene was deemed a potential drawback of EF1 at that time; however, th is characteristic of the promoter seemed an acceptable compromise relative to the overall goal of the experiment. Changing the promoter driving tr ansgene expression in treated animals from EF1 to IRBP, a photoreceptor-specific promoter, will not only restrict expression of the transgene to photoreceptors but will also bring levels of GC1 expression more in line with those found in wild-type retina. Recentl y, two GUCY1*B chickens treated with the lentiviral transgene, IRBP-GC1-eGFP, have presented with high behavioral vision scores. Treating more GUCY1*B embryos with IRBP -GC1-eGFP and subsequent analysis of behavior, ERG measures and retinal morphology will ultimatel y determine if expression
52 of GC1 driven by a photoreceptor specific prom oter is a viable and more attractive treatment option. Non-Autonomous Cell Death and Viral Titer Apoptotic cell death in degenerating retina s does not appear to be cell autonomous. This â€œbystanderâ€ phenomenon suggests that cross talk between phot oreceptor cells or environmental cues play a significant role in the progressive nature of photoreceptor degeneration (Travis, 1998; Ripps, 2002). The percent of the total photoreceptor population slated for death seems to influe nce non-autonomous retinal degeneration. For example, in the GC1 knockout mouse, degenera tion of the cones, which represent 3-5% of the photoreceptor population and are distri buted evenly across the retina, does not adversely affect survival of the rod ce lls (Yang et al., 1999; Coleman et al., 2004). Further, in humans with rod-specific re tinal degeneration caused by rhodopsin gene mutations, cone cell function a nd survival becomes compromi sed when greater than 75% of the affected rod cells dege nerate (Cideciyan et al., 1998). In accordance with this idea that the density of normal functioning photoreceptor cells is directly related to non-autonomous retinal degenera tion, photoreceptor viability in treated GUCY1*B chickens may be influen ced by the number of transduced cells expressing the GC1 transgene. By increasing the titer of the injected vector from 10e9 to 10e10 TU/ml, we will be able to increase the percent of cells transduced from approximately 60% to 85% (Coleman et al., 2003). However, it is unclear what percentage of the total phot oreceptor cell population must be transduced to overcome the phenomenon of non-autonomous, density-depen dent photoreceptor degeneration. We are currently focused on further optimizing our viral packaging method (Coleman et al., 2003) to facilitate consistent generation of viral titers of 10e10 TU/ml or higher because
53 although we were able to restore vision by tr ansducing a relatively low percent of the photoreceptor population, transducti on of a larger percent of these cells may be the most attractive modification leading to significant improvement in the long-term effectiveness of our gene therapy. Clinical Relevance to LCA1 Success in restoration of vision in the GUCY1*B chicken following our treatment is promising. The ultimate goal would be to treat humans with LCA1. There are multiple avenues for approaching clinical treatment in cluding sub-retinal inje ctions both post natal and in utero and through targeting germ cells. An in utero treatment, viral gene delivery to mouse fetuses with a different subtype of LCA, was successful in restoration of visual function and did not cause any evident impair ment of development and growth (Dejenka et al., 2004). However post natal treatment may also be effective, postmortem studies of retinal histopathology in an 11-year-old pa tient with LCA1 revealed regions with preserved retinal cells despite profound visual disturbance at this age (Milam et al., 2003). This discovery paired with the observa tion that the photore ceptor transduction efficiencies achieved in our study support near normal visual behavior in the treated animals is encouraging and consistent with what is well known clinically about human retinal degenerations: patients can display se rviceable vision even if small islands of functioning retina are retained (Geller and Sieving, 1993; Seip le et al., 1995; Carroll et al., 2004). Early treatment of pediatric LC A1 patients, however, would seem worth considering because of a report of prenatal re tinal degeneration in this genotype (Porto et al., 2003). The earliest possible treatment would o ccur at the level of germ cells. In vitro fertilization provides the opportuni ty to use viral gene delivery to treat germ cells before
54 fertilization eliminating any effects of the GUCY2D mutation on development and prenatal retinal degeneration.
55 55 LIST OF REFERENCES Acland GM, Aguirre GD, Ray J, Zhang Q, Al eman TS, Cideciyan AV, Pearce-Kelling SE, Anand V, Zeng Y, Maguire AM, J acobson SG, Hauswirth WW, Bennett J (2001) Gene therapy restores vision in a canine model of childhood blindness. Nat Genet 28: 92-95. Alberts B, Johnson A, Lewis J, Raff M, Robe rts K, Walter P (2002) Molecular biology of the cell. 4th Ed. New York, NY: Garland Science. Amado GR, Chen IS (1999) Lentiviral vectorsthe promise of gene therapy within reach? Science 285: 674-676. Arshavsky VY, Lamb TD, Pugh EN Jr. (2002) G Proteins and Phototransduction. Annu Rev Physiol 64: 153-187. Bennett J, Tanabe T, Sun D, Zeng Y, Kj eldbye H, Gouras P, Maguire AM (1996) Photoreceptor cell rescue in retinal degene ration (rd) mice by in vivo gene therapy. Nat Med 2:649-654. Bennett J, Zeng Y, Bajwa R, Klatt L, Li Y, Maguire AM (1998) Adenovirus-mediated delivery of rhodopsin-promoted bcl-2 results in a delay in photoreceptor cell death in the rd/rd mouse. Gene Ther 5:1156-1164. Carroll J, Neitz M, Hofer H, Neitz J, W illiams DR (2004) Functional photoreceptor loss revealed with adaptive optics: An alternate cause on color blindness. PNAS 101:8461-8466. Cartwright AL (2000) Incuba ting and hatching eggs. Texa s Agricultural Extension Service. http://texaserc.tamu.edu/ , last accessed March, 2006. Chang GQ, Hao Y, Wong F (1993) Apoptosis : final common pathway of photoreceptor cell death in rd, rds, and rhodops in mutant mice. Neuron 11:595-605. Chaum E, Hatton MP (2002) Gene therapy for genetic and acquired re tinal diseases. Surv Opthalmol 47:449-469. Chen J, Flannery JG, La Vail MM, Steinberg RH, Xu J, Simon MI. (1996) bcl-2 overexpression reduces apoptotic photoreceptor cell death in thr ee different retinal degenerations. Proc Natl Acad Sci U S A 93:7042-7047.
56 Cheng KM, Shoffner RN, Gelatt KN, Gum GG, Otis JS, Bitgood JJ (1980) An autosomal recessive blind mutant in the chicken. Poult Sci 59:2179-81. Cideciyan AV, Hood DC, Huang Y, Banin E, Li ZY, Stone EM, Milam AH, Jacobson SG (1998) Disease sequence from mutant rhodopsin allele to rod and cone photoreceptor degeneration in man. Proc Natl Acad Sci U S A 95: 7103-7108. Coleman JE (2003) Efficient transduction and targeted expression of lentiviral vector transgenes in the developi ng retina. UF Online Doctoral Dissertations, University of Florida, Gainesville, FL USA. Coleman JE, Fuchs GE, Semple-Rowland SL (2002) Analyses of the guanylate cyclase activating protein-1 gene promoter in the developing retina. Invest Ophthalmol Vis Sci 43:1335-1343. Coleman JE, Huentelman MJ, KasparovS, Metcalfe BL, Paton JFR, Katovich MJ, Semple-Rowland SL, Raizada MK (2003) Efficient large-scale production and concentration of HIV-1 base d lentiviral vectors for us e in vivo. Physiol Genomics 12:221-228. Coleman JE, Zhang Y, Brown GAJ, Semple-R owland SL (2004) Cone cell survival and downregulation of GCAP1 protein in the retinas of GC1 knockout mice. Invest Ophthalmol and Vis Sci 45:3397-3403. Conley M, Fite KV (1980) Optokinetic nysta gmus in the domestic pigeon. Effects of foveal lesions. Brain Behav Evol 17: 89-102. Dejneka NS, Surace EM, Aleman TS, Cideciyan AV, Lyubarsky A, Savchenko A, Redmond TM, Tang W, Wei Z, Rex TS, Glover E. Maguire AM, Pugh EN Jr, Jacobson SG, Bennet J (2004) In utero gene therapy rescues vi sion in a murine model of congenital blin dness. Mol Ther 9:182-188. Delyfer MN, Leveillard T, Mohand-Said S, Hi cks D, Picaud S, Sahel JA (2004) Inherited retinal degenerations: therapeutic prospects. Biol Cell 96:261-269. Dizhoor AM, Lowe DG, Olshevskaya EV, Laura RP, Hurley JB (1994) The human photoreceptor membrane guanylyl cyclase, Re tGC, is present in outer segments and is regulated by calcium and a so luble activator. Neuron 12:1345-1352. Doonan F, Donovan M, Cotter TG (2003) Casp ase-independent photoreceptor apoptosis in mouse models of retinal degeneration. J Neurosc 23:5723-5731. Gavrieli Y, Sherman Y, Ben-Sasson SA ( 1992) Identification of programmed cell death in situ via specific labeling of nucle ar DNA fragmentation. J Cell Bio 119: 493501. Geller AM, Sieving PA (1993) Assessment of fov eal cone photoreceptors in Startgardtâ€™s macular dystrophy using a small dot detection task. Vision Res 33:1509-1524.
57 Goverdhana S, Puntel M, Xi ong W, Zirger JM, Barcia C, Curtin JF, Soffer EB, Mondkar S, King GD, Hu J, Sciascia SA, Candolfi M, Greengold DS, Lowenstein PR, Castro MG (2005) Regulatable gene expression systems for gene therapy applications: progress and future challe nges. Mol Ther 12:189-211. Green ES, Rendahl KG, Zhou S, Ladner M, Coyne M, Srivastava R, Manning WC, Flannery JG (2001) Two animal models of retinal degeneration are rescued by recombinant adeno-associated virus-me diated production of FGF-5 and FGF-18. Mol Ther 3: 507-515. Gregory TR (2005) Animal Genome Size Database. http://www.genomesize.com/ , last accessed April 2006. Hamburger V, Hamilton HL (1951) A series of normal stages in the development of chick embryos. J Morphol 88:49-92. Hanein S, Perrault I, Gerber S, Tanguy G, Barbet F, Ducroq D, Calvas P, Dollfus H, Hamel C, Lopponen T, Munier F, Santos L, Shalev S, Zafeiriou D, Dufier JL, Munnich A, Rozet JM, Kaplan J (2004) Leber congenital amaurosis: comprehensive survey of the genetic hete rogeneity, refinement of the clinical definition, and genotype-phenotype correla tions as a strategy for molecular diagnosis. Hum Mutat 23:306-317. Huang PC, Gaitan AE, Hao Y, Petters RM , Wong F (1993) Cellular interactions implicated in the mechanism of photor eceptor degeneration in transgenic mice expressing mutant rhodopsin gene . Proc Natl Acad Sci USA 90:8484-8488. Jomary C, Vincent KA, Grist J, Neal MJ, Jones SE (1997) Rescue of photoreceptor function by AAV-mediated gene transfer in a mouse model of inherited retinal degeneration. Gene Ther 4:683-690. Koenkoop RK (2004) An overview of Leber c ongenital amaurosis: a model to understand human retinal development. Su rv of Ophthalmol 49:379-398. Komenda JK, Fite KV (1983) Optokinetic nysta gmus in progressive retinal degeneration. Behav Neurosci 97: 928-936. Koutalos Y, Yau KW (1996) Regulation of sens itivity in vertebrate rod photoreceptors by calcium. Trends Neurosci 19:73-81. Kumar M, Keller B, Makalou N, Sutton RE (2001) Systematic determination of the packaging limit of lentiviral vectors. Hum Gen Ther 12:1893-1905. Lau D, McGee LH, Zhou S, Rendahl KG, Manning WC, Escobedo JA, Flannery JG (2000) Retinal degeneration is slowed in transgenic rats by AAV-mediated delivery of FGF-2. Invest Ophtha lmol Vis Sci 41: 3622-3633.
58 Leveillard T, Mohand-Said S, Lorentz O, Hicks D, Fintz AC, Clerin E, Simonutti M, Forster V, Cavusoglu N, Chalmel F, Dolle P, Poch O, Lambrou G, Sahel JA (2004) Identification and characterization of rodderived cone viability factor. Nat Genet 36: 755-759. Liang FQ, Aleman TS, Dejneka NS, Dudus L, Fisher KJ, Maguire AM, Jacobson SG, Bennett J (2001) Long-term protection of re tinal structure but not function using RAAV.CNTF in animal models of retin itis pigmentosa. Mol Ther 4: 461-472. Lolley RN, Rong H, Craft CM (1994) Linkage of photoreceptor degeneration by apoptosis with inherited defect in ph ototransduction. Invest Opthalmol Vis Sci 35:358-62. Lowe DG, Dizhoor AM, Lui K, Gu Q, Spencer M, Laura R, Lu L, Hurley JB (1995) Cloning and expression of a second phot oreceptor specific membrane retina guanylyl cyclase (RetGC), RetGC-2. Neurobiology 92:5535-5539. Milam AH, Barakat MR, Gupta N, Rose L, Aleman TS, Pianta MJ, Cideciyan AV, Sheffield VC, Stone EM, Jacobson SG ( 2003) Clinicopathologic effects of mutant GUCY2D in Leber congenital amau rosis. Ophthalmology 110: 549-558. Mizuguchi H, Xu Z, Ishii-Watabe A, Uchi da E, Hayakawa T (2000) IRES-dependent second gene expression is significantly lower than cap-dependent first gene expression in a bicistronic vector. Mol Ther 1: 376-382. North MO (1978) Commercial chicken pr oduction manual, 2nd Ed. Westport, CT: AVI Publishing Company, Inc. Perrault I, Rozet JM, Calvas P, Gerber S, Ca muzat A, Dollfus H, Chatelin S, Souied E, Ghazi I, Leowski C, Bonnemaison M, Le Pa isler D, Frezal J, Dufier JL, Pittler S, Munnich A, Kaplan J (1996) Retinal-speci fic guanylate cylclase gene mutations in Leberâ€™s congenital amaurosis. Nat Genet 14:461-464. Pettite JN, Mozdziak PE (2002) Production of transgenic poultry. In: Transgenic animal technology: a laboratory hand book. 2nd Ed. Elsevier Science (USA). Polans A, Baehr W, Palczewski K (1996) Turned on by Ca2+! The physiology and pathology of Ca2+-binding proteins in the retina. Trends Neurosci 19:547-554. Portera-Cailliau C, Sung CH, Nathans J, Adler R (1994) Apoptotic photoreceptor cell death in mouse models of retinitis pigmentosa. Proc Natl Acad Sci USA 91:974978. Porto FB, Perrault I, Hicks D, Rozet JM, Ha noteau N, Hanein S, Kaplan J, Sahel JA (2003) Prenatal human ocular degeneration occurs in Leber's Congenital Amaurosis (LCA1 and 2). Adv Exp Med Biol 533: 59-68.
59 Pugh EN Jr, Duda T, Sitaramayya A, Sh arma RK (1997) Photoreceptor guanylate cyclases: A review. Biosci Rep 17:429-473. Reed JC (2000) Mechanisms of apoptosis. Ame J Pathol 157:1415-1430. Ripps H (2002) Cell death in retinitis pigm entosa: gap junctions and the â€˜bystanderâ€™ effect. Exp Eye Res 74:327-336. Rozet JM, Perrault I, Gerber S, Hanein S, Barbet F, Ducroq D, Souied E, Munnich A, Kaplan J (2001) Complete abolition of the retinal-specific guanylyl cyclase (retGC1) catalytic ability consistently leads to Leber congenital amaurosis(LCA). Invest Ophthalmol Vis Sci 42:1190-1192. Schmid KL, Wildsoet CF (1998) Assessment of visual acuity and contrast sensitivity in the chick using an optokinetic nyst agmus paradigm. Vision Res 38: 2629-2634. Semple-Rowland SL, Lee NR, Van Hooser JP, Palczewski K, Baehr W (1998) A null mutation in the photoreceptor guanylate cyclase gene causes the retinal degeneration chicken phenotype. Proc Natl Acad Sci USA 95:1271-1276. Semple-Rowland SL, Cheng KM (1999) rd a nd rc carry the same GC1 null allele (GUCY1*). Exp Eye Res 69:579-581. Shyjan AW, de Sauvage FJ, Gillett NA, Goeddel DV, Lowe DG (1992) Molecular cloning of a retina-specific membrane guanylyl cyclase. Neuron 9:727-737. Sieple W, Holopigian K, Szlyk JP, Greenst ein VC (1995) The effects of random element loss on letter identification â€“ Implications for visual-acuity loss in patients with retinitis-pigmentosa. Vision Res 35:2057-2066. Takahashi M, Miyoshi H, Verma IM, Gage FH (1999) Rescue from photoreceptor degeneration in the rd m ouse by human immunodeficiency virus vector-mediated gene transfer. J Virol 73:7812-7816. Travis GH (1998) Mechanisms of cell death in the inherited retinal degenerations. Am J Hum Genet 62:503-508. Tso MO, Zhang C, Abler AS, Chang CJ, W ong F, Chang GQ, Lam TT (1994) Apoptosis leads to photoreceptor degeneration in i nherited retinal dystrophy of RCS rats. Invest Opthalmol Vis Sci 35:2693-2699. Ulshafer RJ, Allen C, Dawson WW, Wolf ED (1984) Hereditary retinal degeneration in the Rhode Island Red chicken. I. Histology and ERG. Exp Eye Res 39:125-135. Ulshafer RJ and Allen CB (1985) Hereditary retinal degeneration in the Rhode Island Red chicken: ultrastructural analysis. Exp Eye Res 40:865-877.
60 Williams ML, Coleman JE, Haire SE, Aleman TS, Cideciyan AV, Sokal I, Palczewski K, Jacobson SG, Semple-Rowland SL. (2006) Lentiviral expression of retinal guanylate cyclases-1 (retGC1) restores vision in a avian model of childhood blindness. PLoS Med, In press. Yang RB, Foster DC, Garbers DL and Fulle HJ (1995) Two membrane forms of guanylyl cyclase found in the eye. Pr oc Natl Acad Sci USA 92:602-606. Yang RB, Robinson SW, Xiong WH, Yau KW, Bi rch DG, Garbers DL (1999) Disruption of a retinal guanylyl cyclase gene lead s to cone-specific dystrophy and paradoxical rod behavior. J Ne urosci 19:5889-5897.
61 61 BIOGRAPHICAL SKETCH Melissa Williams grew up in Charlotte, North Carolina. She earned a Bachelor of Science in biology at the Univ ersity of North Carolina at Charlotte where she conducted research on the behavior of honeybees. Melissa entered the Interdis ciplinary Program in Biomedical Sciences at the University of Florida in August of 2003. She joined the Semple-Rowland lab in the Department of Neuroscience in May of 2004 and began pursuing research goals in gene therapy rescue of vision.